Volcaniclastic rocks

Several landcover classifications based on Landsat images from 1966 to 2018

Since the report of the RNA aptamer for theophylline, theophylline has become a key molecule in chemical biology for designing RNA switches and riboswitches. In addition, theophylline is an important drug for treating airway diseases including asthma. The classic RNA aptamer with excellent selectivity for theophylline has been used to design biosensors, although DNA aptamers are more desirable for stability and cost considerations. In this work, we selected DNA aptamers for theophylline, and all the top sequences shared the same binding motifs. Binding was confirmed using isothermal titration calorimetry and a nuclease digestion assay, showing a dissociation constant (Kd) around 0.5 μM theophylline. The Theo2201 aptamer can be truncated down to 23-mer while still has a Kd of 9.8 μM. The selectivity for theophylline over caffeine is around 250,000-fold based on a strand-displacement assay, which was more than 20-fold higher compared to the classic RNA aptamer. For other tested analogs, the DNA aptamer also showed better selectivity. Using the structure-switching aptamer sensor design method, a detection limit of 17 nM theophylline was achieved in the selection buffer, and a detection limit of 31 nM was obtained in 10% serum.

A DNA Aptamer for Theophylline with Ultrahigh Selectivity Reminiscent of the Classic RNA Aptamer

This dataset provides two 30-m resolution time series products of annual land cover classifications over the Arctic Boreal Vulnerability Experiment (ABoVE) core domain for each year of the period 1984-2014. The data are the annual dominant plant functional type in a given 30-m pixel derived from Landsat surface reflectance, landcover training data mapped across the ABoVE domain (using Random Forests modeling, with clustering and interpretation of field photography) and very high resolution imagery to assign land cover classifications. One product has a 15-class land cover classification that breaks out forest and shrub types into several additional classes; the other product provides a simplified, 10-class approach. Classification accuracy assessment results are provided per year. Assessments were based on a probability-based random sample of reference data that supported statistically robust estimation of areas and uncertainties in mapped areas.

This dataset is a synthesis of field plot characterization data, derived above-ground and below-ground combusted carbon, and acquired Fire Weather Index (FWI) System components for burned boreal forest sites across Alaska, USA, the Northwest Territories, and Saskatchewan, Canada from 1983-2016. Unburned plot data are also included. Compiled plot-level characterization data include stand age, disturbance history, tree density, and tree biophysical measurements for calculation of the above-ground (ag) and below-ground (bg) biomass/carbon pools, pre-fire and residual post-fire soil organic layer (SOL) depths and estimates of combustion of tree structural classes. The measured slope and aspect for each site and an assigned moisture class based on topography are also provided. Data from 1019 burned and 152 unburned sites are included. From the estimates of combusted ag and bg carbon pools and SOL losses, the total carbon combusted, the proportion of pre-fire carbon combusted, and the proportion of total carbon combusted were calculated for each plot. FWI System components including moisture and drought codes and indices of fire danger were obtained for each plot from existing data sources based on the plot location, year of burn, and a dynamic start-up date (day of burn, DOB) from the global fire weather database. Data for soil characteristics are included in a separate file.

1. Description of methods used for collection/generation of data: A. Soil Landscapes of Canada (SLC) Data Version 2.2 (Centre for Land and Biological Resources Research, Agriculture and Agri-Food Canada, 1996): This dataset was published in Dec 1996 by Agriculture and Agri-Food Canada based on soil survey mapping done over the years and updated regularly. The data has a resolution of 1:1 million and covers entire Canada. The dataset is structured as below: a. It divides whole of Canada into several ecodistricts which have been made available as a polygon shapefile where each polygon corresponds to an ecodistrict. b. Each polygon is further divided into a number of soil texture components. c. The percentage area covered by the components in an ecodistrict are given in tables but their locations are not known i.e. there is no shapefile defining coverage of each component inside of each ecodistrict. d. Each of these components have been allocated a soil type: CL - clay loam, KCL - clay loam with >35% coarse fragments, CY - clay, LM - loam, KLM - loam with >35% coarse fragments, SD - sand, KSD - sand with >35% coarse fragments, SL - sandy loam, KSL - sandy loam with >35% coarse fragments, KSP - cobbly sand, O - organic, # - not applicable (rock, ice, urban). Centre for Land and Biological Resources Research. 1996. Soil Landscapes of Canada, v.2.2, Research Branch, Agriculture and Agri-Food Canada. Ottawa. B. STASTSGO2 USA data (United States Department of Agriculture, 2015): The dataset was published by National Cooperative Soil Survey in December 2015 succeeding the STATSGO dataset. The dataset was developed by using detailed soil survey maps, geology, vegetation, climate data and Landsat images. The dataset has a resolution of 1:250,000 in the continental U.S., Hawaii, Puerto Rico, and the Virgin Islands and 1:1,000,000 in Alaska. The structure of dataset is very similar to the SLC dataset with a few differences: a. The dataset divides the entire area into map units similar to the ecodistricts in SLC dataset. Unlike the SLC dataset, a single map unit can correspond to several polygons in the shapefile. b. Each map unit is further divided into a number of components. Like the Canadian dataset, the location of these components are not known (i.e. is it not fully distributed). c. For each component, there are several vertical layers. % sand, % silt, % clay and % organic for each vertical horizon is present. Soil Survey Staff, Natural Resources Conservation Service, United States Department of Agriculture (2015). U.S. General Soil Map (STATSGO2). Available online at http://websoilsurvey.nrcs.usda.gov/. Accessed [06/21/2017]. 2. Methods for processing the data: The processing of the data was done separately for the Canadian part of the river basins (entire Mackenzie River Basin and majority of Nelson-Churchill River Basin) and for the US part of the river basin (a small part of Nelson-Churchill River Basin covering parts of Minnesota, Montana, North Dakota, South Dakota and Washington and Wisconsin). A. Processing the SLC 2.2 dataset: I. Conversion of soil texture classes to percentage soil texture: The soil type for each component was converted to % sand, % clay and % organic using Canadian System of Soil Classification (CSSC) manual, 3rd edition (Soil Classification Working Group, Agriculture and Agri-Food Canada, 1998). a. For this, the minimum and maximum of percentage values of sand and clay were extracted from the soil triangle in the CSSC manual and the mean value was calculated by averaging max and min values for each of texture class. b. Organic were given a minimum value of 5%. For organic soil, the organic percent was increased to 100% setting each of %Sand and %Clay to zero. c. It was ensured that % organic + % sand + % clay stays less than or equal to 100. To ensure this, the minimum percentage organic in sandy soils (SD and KSD) was changed from 5% to 2.5%. d. Some classes were not present in the soil triangle like KSD (sand with greater than 35% coarse fragments) and KSL (sandy loam with greater than 35 % coarse fragments). Assuming coarse fragments as sand, the percentages were changed by increasing max and min values of sand by 35 (capped at 100 %) and finding the mean thereafter. These soil classes were however negligibly present in our areas of interest. e. Several components had the value '#' suggesting rock, urban or ice cover. The ecodistricts where these components had an area greater than 50% was considered as '#' or NA for its entirety while for the ecodistricts where this component area was less than 50%, this component was ignored altogether in further computations. It should be noted that out of 683 components in Mackenzie River Basin, only 68 had NA or '#' as their texture class and most of these had areas less than 25%. Similar findings were observed in Nelson-Churchill River Basin also. Soil Classification Working Group. 1998. The Canadian System of Soil Classification, 3rd ed. Agriculture and Agri-Food Canada Publication 1646, 187 pp. II. Aggregating the soil percentage to 0.125 degree resolution: Following were the steps followed in aggregation of soil texture percentage to create gridded dataset: a. First, inland water was “differenced” from the ecodistrict polygon shapefile using QGIS as the area percent of each component in a polygon corresponds to the land area of the polygon rather than the total area. b. Sand, clay and organic percentage was calcualted for each ecodistrict polygon by following formula: Min % of Polygon = Minimum of (Min % of each component in the polygon) Max % of Polygon = Maximum of (Max % of each component in the polygon) Mean % of Polygon = Summation of (%area X %mean) for each component/100 c. Each 0.125 degree resolution grid cell was intersected with the shapefile layer of ecodistricts thus getting the polygons and their respective area inside each grid cell. The percentage values of sand, clay and organic matter were then calculated using similar formulae as above. B. Processing STATSGO2 dataset: I. Conversion of soil class to percentage soil texture: Since the dataset already has the percentage values of sand, silt, clay and organic, there was no need to map the soil classes to soil texture percentages as in SLC 2.2 dataset. However, a few adjustments were made to the percentage values so that the processed data from the two sources are coherent: a. The STATSGO2 dataset has organic percent for each vertical layer of soil. For all the soil layers which have organic percent greater than 30%, the %sand, %clay, %silt are all zero suggesting a fully organic soil for that component. Therefore, the organic soils (having organic soil percent greater than 30%) were made a 100 % organic similar to SLC 2.2 dataset assumption. b. %sand, %silt and %clay add up to 100 % while organic is an addition which makes %sand, %silt, %clay, %organic greater than 100. So, the percentage sand, silt, clay and organic were normalized to 100. II. Aggregating the soil percentage to 0.125 degree resolution: The aggregation methodology followed for STATSGO2 dataset is very similar to the SLC 2.2 dataset. An extra step involved in processing this dataset was the calculation of soil texture data per component by averaging soil texture percentage of vertical horizons weighted over their depth. After getting soil texture per component, the rest of the steps were similar to the SLC 2.2 dataset to calculate the soil texture percent per grid cell. The two datasets were then appended to create gridded soil texture data for Nelson-Churchill River Basin and Mackenzie River Basin. 3. Instrument- or software-specific information needed to interpret the data: The dataset is comprise of two file typess: the "*.csv" data table file containing the minimum, maximum and average percentage of sand, clay and organic for each 0.125 degree grid cell (identified by Latitude and Longitude in the table); and the shapefiles of gridded river basins, which can be viewed in any GIS software (e.g. QGIS). The data table file can be joined with the grid cells shapefile using "ID" variable to view spatial distribution of % values of different soil texture variables (%sand, %clay, and %organic), either the average, minimum or maximum can be selected.

Data was collected using autonomous recording units (ARU), and in-person point count surveys, during the 2018 field season. Variables such as: dominant vegetation cover, area of peatland, elevation, Natural Subregion classification, distance to closest road, ambient noise, wind speed, date and time of year, were collected along with the point counts and recorded files. Spectrograms were used to analyze recorded files since each bird species have unique songs and calls. Once files were processed a list of bird species present at each site was used for statistical analysis. This data collection method will also be repeated for the 2019 field season.

Data was collected from three rivers in the Greater Toronto Area of Southern Ontario: Ganatsekiagon Creek (City of Pickering), Wilket Creek (City of Toronto), and Morningside Creek (City of Toronto). The grain size distribution at each site was calculated using a Wolman Pebble count with a 200-stone sample size. Bedload transport was monitored over three years using Radio Frequency Identification (RFID) tracer stones, and periodic topographic surveys were conducted. A total of 300 tracers in 3 size classes were seeded in each site in August 2015. Tracer positions were recorded after each major rainfall event during the active field season each year, resulting in a total of 10, 12, and 13 recoveries in Ganatsekiagon Creek, Wilket Creek, and Morningside Creek, respectively. With each recovery, the travel distance of each tracer since its last known position is calculated. Detailed topographic surveys of the channel bed were conducted in the summers of 2016 and 2018 using a total station. Surveys were used to create DEM of Difference (DOD) at each site after a Triangular Irregular Network interpolation.

Shapefile detailing classified prairie watersheds (n = 4175) according to physiographic characteristics. These characteristics were assembled from a variety of sources, including remote sensed data and government databases. Variables included climatic (annual precipitation, potential evapotranspiration), physical (slope, elevation), surficial geology, wetland (density, size distribution), and land cover/use data. Watersheds were classified using a hierarchical clustering on principal components analysis. As a result, seven distinct classes of watersheds were identified. The dataset defines two classifications schemes: (1) Integrated Watershed Classification, and (2) Land Cover Watershed Classification. The schemes differ as the latter was performed without climatic variables. As such, the land cover approach is suited for applications where local climate is forced using other data sources (e.g., hydrological modelling). The integrated classification is suited for general applications. The associated manuscript, which includes methods and data sources, can be found here: https://doi.org/10.5194/hess-23-3945-2019

Wolfe, J., Whitfield, C. J., Shook, K. R., Spence, C. (2019). Canadian Prairie Watershed Classification [Dataset]. Federated Research Data Repository. https://doi.org/10.20383/101.0197

Canadian Prairie Watershed Classification

Prairies watershed classification geography

Develop a systematic classification of Prairie watersheds based on similar geographic characteristics. The classification serves as a foundation for virtual watershed modelling within the project to investigate how watershed hydrology and biogeochemistry respond to environmental change.

MESH 1.4 CLASS 3.6

Model name: Coupled hydrology land-surface model (MESH), using the Canadian Land Surface Scheme (CLASS) Model version number: MESH (1.4) using CLASS (3.6) Model source/webpage: https://wiki.usask.ca/pages/viewpage.action?pageId=220332269 Model output pre-processing script: TBA. Scripts will be shared on github. Model output post-processing script: TBA. Scripts will be shared on github. Model setup: Physically based with no calibration. Time step: Hourly Initial condition: Physically based, set based on the understanding of the hydrological system Boundary condition: Physically based, set based on the understanding of the hydrological system

surface radiative and turbulent fluxes snow energetics snow hydrology alpine hydrology glacier energetics and hydrology diagnostic variables MESH CLASS Canadian Rockies

Snow and ice processes in high mountain environments are controlled by precipitation, blowing snow redistribution, sublimation, and the exchange of radiative and turbulent fluxes. Snow interception in forests is important in many alpine regions. The current dataset is the collection of simulated data at different alpine and glacierized research sites in the Canadian Rockies (e.g. alpine ridges, glacier, alpine forests and clear cuts, and montane sites). The coupled hydrology land-surface model (MESH), using the Canadian Land Surface Scheme (CLASS), is run using a physically based modeling approach based on the understanding of the hydrological system. For the different sites, the model was run in single column mode and forced by 30-min meteorological observations collected as part of the Canadian Rockies Hydrological Observatory. The forcing data of shortwave and longwave irradiance were adjusted to slopes; meanwhile, air temperature, humidity, pressure and precipitation were adjusted for elevation to the single column model site.

Land surface schemes can be applied to simulate evapotranspiration. This dataset contains the driving meteorological data, and various diagnostic data, from one of the Boreal Ecosystem Research and Monitoring Sites in central Saskatchewan, known as the Old Jack Pine site. In Nazarbakhsh et al. (2019, Hydrological Processes, https://doi.org/10.1002/hyp.13674) we used these data to drive two Canadian land surface schemes (CLASS and CLASS–CTEM). We used half–hourly values of shortwave radiation, longwave radiation, precipitation, air temperature, specific humidity, wind speed, and atmospheric pressure to drive the models. Flux tower estimates of evapotranspiration, with energy balance closure applied, were used to assess the performance of the models on daily and monthly timescales for years 2000 to 2010. We also used soil moisture (measured with Campbell Scientific CS615 probes, which measure liquid water content only) and soil temperature observations for years 2000 to 2010 to assess the models’ performance during the snowmelt and soil–thaw periods in the spring.

boreal forest BERMS Old Jack Pine site evapotranspiration land surface schemes CLASS CLASS-CTEM

Global climate change has had a significant effect on the permafrost landscape in Northern cold regions. Due to natural processes such as precipitation change and rising temperatures, in addition to anthropogenic intervention, we notice a steadily increasing, and systemic pattern of permafrost thaw. The fieldwork took place in the Sahtu region of the Northwest Territories, specifically within the Bogg Creek Watershed, located 30 kilometers South of the Town of Norman Wells. Our work seeks to use two geophysical methods, electrical resistivity tomography (ERT) and electromagnetic induction (EMI) to estimate the permafrost table depth in areas with local landscape features such as lakes and clear cuts. The purpose of this research is also to benchmark a non-ground coupled system (EMI) against the more classical ground-coupled methods (ERT). The ERT system we used is the Syscal Junior 48, and the EMI system was the Geonics Inc EM-31 and EM-34. The field mission lasted four days, and we collected permafrost probe, ERT, and EMI measurements along three survey lines. Two lines (Line 001 and Line 002) were located at a drill pad site, labeled as MW04T, and another line (Line 004) was located near a lake’s shoreline, labeled as “Marg Lake”. The permafrost probe showed a clear plunging of the permafrost table at the tree line at MW04T, and another plunge towards the lakeshore. The ERT data showed similar trends where the regions of high resistivity, representing ice/permafrost, plunged similar to our permafrost probe measurements. Our EMI measurements and interpretation also agreed with the permafrost probe measurements and ERT measurements, whereby a deepening of the permafrost table manifested itself in the data as increasing conductivity. This work successfully demonstrated that small ground-based EMI systems are suitable equipment for use in detecting permafrost thaw in Northern settings.

MESH model is used for calibration/ validation of the streamflow. MESH is a physics-based, land-surface hydro- logical modelling system developed by Environment and Climate Change Canada. MESH performs both water and energy balances and is best suited to cold-region, large-scale catchments due to its ability to simulate snow processes, such as snow accumulation, redistribution, and melt. MESH is used to simulate important hydrological processes (like runoff generation, evapotranspiration, and soil moisture). SED model is a physically based watershed sediment transport model which is developed based on empirical equations and the sediment mass balance equation (for overland and instream flow). It includes different sediment classes and is suitable for large scale cold regions catchments. The model simulates hourly sediment load and concentration. The following input data were used for hydrological modelling: - Meteorological data: For model calibration and validation, seven forcing data (precipitation, longwave radiation, shortwave radiation, wind speed, air temperature, barometric pressure and specific humidity) is taken from GEM-CaPA. (https://wiki.usask.ca/display/MESH/Forcing+Datasets+for+MESH#ForcingDatasetsforMESH-GEM-CaPA) - Drainage basin data: Digital Elevation data (Geobase database as Canadian Digital Elevation Data), Landcover data (http://cec.org/tools-and-resources/map-files/land-cover-2010-landsat-30m), soil data (Soil Landscapes of Canada (version 2.2), eco-regions shape files for Canada, and other relevant data was obtained from different sources. The drainage basin was created using GreenKenue. - Hydrological data: The ECCC hydat stations for Athabasca River Basin is used for the discharge data. - Sediment concentration data: Surface water quality data website/ Long Term River Data http://environment.alberta.ca/apps/EdwReportViewer/LongTermRiverStation.aspx

Most of the data are available free for analysis in the Fraser Basin. The analysis procedure is applied on different data types, including point (e.g., streamflow records), vector (e.g., river network), and raster images (e.g., DEM) as inputs for the setup of the MESH model in the Fraser Basin. Different processing steps can be applied over input datasets, such as clipping, merging, filtering, mosaicking. The input data are as follows: 1) Basin, subbasin, subsubbasin boundary shapefiles 2) River networks 3) Streamflow records 4) Digital Elevation Model (DEM) 5) Land Cover Classes 6) Meteorological forcing 7) Soil Dataset

Data collected for this project include a series of surface and subsurface variables characterizing the thermal and mass balance of observation points and the study basin as a whole. This includes both continuous (30 minute intervals) and discrete data sets collected between 2017-07-10 and 2019-09-10. Each data set was stratified across a series of land cover classes designated within the study basin: open water, mineral-cored uplands, riparian, ice-rich permafrost, and thermokarst features. Surface variables measured continuously include stream discharge, albedo, air temperature, relative humidity, net radiation, and rain. Surface variables measured discretely include snow depth, snow density, vegetation height, and vegetation density. Subsurface variables measured continuously include soil temperature, volumetric moisture content, and water table depth. Subsurface variables measured discretely include evapotranspiration rates, groundwater pressure head, and near surface soil thermal properties. Drone surveys allow for estimates of snowpack depletion rates, basin extent, and mean values of elevation, aspect, and relief for each land cover class. Field samples collected for later laboratory work include water from surface and subsurface sources (d2H and d18O isotope analysis), and soil samples from mineral and organic strata (density, porosity, and hydraulic conductivity).

Investigation of alpine land cover classes and their influence on basin water balance in the Mackenzie Mountains, Northwest Territories

The Parsivel is an optical disdrometer that measures the velocity and size of falling hydrometeors with the goal of classifying hydrometeor type and retrieve precipitation Particle Size Distribution (PSD). We use an OTT Parsivel² laser-optical disdrometer that functions with two sensor heads facing each other. One head is a transmitter that emits radiation (at 650 nm wavelength at the red band) in a horizontal plane and the other head is a receiver that senses how much of that radiation is received. The instrument measures the size of the hydrometeor by measuring the length of radiation that is blocked by the particle diameter. The velocity of the hydrometeor is estimated based on the time that a particular hydrometeor is blocking the radiation between the transmitter and receiver. The OTT Parsivel² retrieves particle velocity and size every minute, with a range in velocity from 0.2 to 20 m/s and a range in particle diameter from 0.2 to 25 mm (OTT Hydromet GmbH, 2018). OTT Hydromet GmbH. (2018). Operating instructions Present Weather Sensor OTT Parsivel 2. Kempten, Germany. Retrieved from https://www.ott.com/download/operating-instructions-present-weather-sensor-ott-parsivel2-without-screen-heating/

intensity of precipitation mm/h 1 min Parsivel precipitation since start mm 1 min Parsivel weather code SYNOP WaWa 1 min Parsivel weather code METAR/SPECI 1 min Parsivel weather code NWS 1 min Parsivel radar reflectivity dBz 1 min Parsivel (simulated by Parsivel software) MOR Visibility m 1 min Parsivel signal amplitude of laserband 1 min Parsivel number of detected particles 1 min Parsivel temperature in sensor °C 1 min Parsivel heating current A 1 min Parsivel sensor voltage V 1 min Parsivel kinetic energy J/(m2h) 1 min Parsivel snow intensity mm/h 1 min Parsivel size and speed spectrum number of particles per class 1 min Parsivel (spectrum for 32 classes of particle size and 32 classes of speed)

173 pesticide concentrations ug/L Spring/Summer Ca, CO3, Cl, F, Fe, Mg, Mn mg/L Spring/Summer conductivity us/cm Spring/Summer pH mg/L Spring/Summer nitrate dissolved mg/L Spring/Summer phosphorus – ortho & total mg/L Spring/Summer hardness, alkalinity mg/L Spring/Summer total dissolved solids mg/L Spring/Summer total nitrogen mg/L Spring/Summer ammonia-N mg/L Spring/Summer total organic carbon mg/L Spring/Summer wetland class classification Spring/Summer wetland vegetation percent Spring/Summer cyanobacteria bloom yes/no Spring/Summer Wetland area m2 Spring/Summer aquatic invertebrates abundance Spring/Summer waterfowl abundance Spring/Summer pesticide toxicity ug/L Collected from literature

permanence class Category Annual cover class Category Annual pH dimensionless Spring/summer YSI sonde conductivity us/cm Spring/summer YSI sonde alkalinity mg/L CaCO3 Spring/summer YSI sonde total nitrogen mg/L Spring/summer SmartChem analyzer 170 nitrate mg/L Spring/summer SmartChem analyzer 170 ammonia-N mg/L Spring/summer SmartChem analyzer 170 total phosphorus mg/L Spring/summer SmartChem analyzer 170 total dissolved phosphorus mg/L Spring/summer SmartChem analyzer 170 soluble reactive phosphorus mg/L Spring/summer SmartChem analyzer 170 sulfate mg/L Spring/summer SmartChem analyzer 170 Chlorophyll a ug/L Spring/summer Spectrometer

A high resolution, enhanced version of Environment and Climate Change Canada’s MESH (Modélisation Environnementale Communautaire - Surface Hydrology) land surface hydrological model was set up at a spatial resolution of approximately 4 km by 4 km to correspond to the resolution of dynamically downscaled Weather Research Forecast (WRF) atmospheric model outputs for current and future climates in the region. This convection-permitting WRF product used ERA-Interim reanalysis product boundary conditions over 2000 - 2015 to produce realistic, high resolution weather simulations. The pseudo global warming (PGW) approach to dynamical downscaling of future warming projection under RCP8.5 (2086 - 2100), used WRF bounded by ERA-Interim outputs that were perturbed by the mean outcomes of an ensemble of Coupled Model Intercomparison Project Phase 5 (CMIP5) climate model projections. Available land surface data consist of digital elevation models (DEMs), i.e. the hydrologically conditioned HydroSheds DEM that has a spatial resolution of approximately 90 m available at (https://www.mrlc.gov/downloads/sciweb1/shared/hydrosheds), and its derived products including flow direction and drainage density. Soil data was collected from a rasterized version of the Soil Landscapes of Canada (SLC) dataset (https://open.canada.ca/data/en/dataset). The dataset covers Canada at 90 m spatial resolution and is derived from original data at a scale of 1:1M. This dataset has some missing information for the Bow River Basin, for instance there is no information on the percentages of clay and sand of the first soil layer (0 – 5 cm depth). Landcover data was downloaded from the Commission for Environmental Cooperation (CEC) (http://www.cec.org/north-american-land-change-monitoring-system/) covering all of the North America at a resolution of 30 m with 19 land cover classes. The Randolph Glacier Inventory 6.0 data (https://www.glims.org/RGI/rgi60_dl.html), based on Landsat imagery from 2004–06, were used to delineate glacier coverage in the basin. The inventory was generated and manually checked in 2008 (Bolch et al., 2010). Prior to this project, MESH, did not consider the impact of slope and elevation on meteorological forcings below the resolution of the data, which is not a reasonable assumption in mountains. Here, incoming solar radiation was calculated as a function of terrain slope and aspect. Also, precipitation, temperature, pressure, humidity, and longwave radiation were corrected for elevation. The necessary cold regions processes (blowing snow, intercepted snow, sublimation, frozen soil infiltration, slope/aspect impacts on melt rates, glacier ice melt) and water management processes needed to simulate the natural and reservoir-managed streamflow hydrographs in the basin were modelled. Most model parameter values were set based on remote sensing, land surveys and the results and understandings from previous regional hydrological investigations, however forest root depth and stomatal resistance, and soil hydraulic conductivity and channel routing model parameters were calibrated using measured (2006 - 2015) streamflows on the Bow River at Banff, and evaluated (2000 - 2005) at the same stream gauging station.

The MESH version r1589_RTE_ts450s was used. This version of MESH has a hybrid functionality which allows users to define different grid specifications for routing scheme and land surface scheme. Here, in our model, routing scheme has a cell resolution of 0.0083 degree and land surface scheme has a resolution of 0.09 degree. The model includes 13 Ground Response Unit (GRU) classes. The Digital Elevation Models (DEM) was downloaded from USGS HydroSHEDS database with 30 arc-second resolution. Soil data was obtained from Global Soil Dataset for Earth System Models (Shangguan et al. 2014) (http://globalchange.bnu.edu.cn/research/soilw) and the landcover data was obtained from Climate Change Initiative (CCI) Land Cover 2015. The forcing data was extracted from the Global Environmental Multiscale atmospheric model and the Canadian Precipitation Analysis (GEM-CaPA) which a high-resolution gridded database.

Snowmelt contributions to streamflow in mid-latitude mountain basins typically dominate other runoff sources on annual and seasonal timescales. Future increases in temperature and changes in precipitation will affect both snow accumulation and seasonal runoff timing and magnitude, but the underlying and fundamental roles of mountain basin geometry and hypsometry on snowmelt sensitivity have received little attention. To investigate the role of basin geometry in snowmelt sensitivity, a linear snow accumulation model and the Cold Regions Hydrological Modeling (CRHM) platform driven are used to estimate how hypsometry affects basin-wide snow volumes and snowmelt runoff. Area-elevation distributions for fifty basins in western Canada were extracted, normalized according to their elevation statistics, and classified into three clusters that represent top-heavy, middle, and bottom-heavy basins. Prescribed changes in air temperature alter both the snow accumulation gradient and the total snowmelt energy, leading to snowpack volume reductions (10–40%), earlier melt onsets (1–4 weeks) and end of melt season (3 weeks), increases in early spring melt rates and reductions in seasonal areal melt rates (up to 50%). Basin hypsometry controls the magnitude of the basin response. The most sensitive basins are bottom-heavy, and have a greater proportion of their area at low elevations. The least sensitive basins are top-heavy, and have a greater proportion of their area at high elevations. Basins with similar proportional areas at high and low elevations fall in between the others in terms of sensitivity and other metrics. This work provides context for anticipating the impacts of ongoing hydrological change due to climate change, and provides guidance for both monitoring networks and distributed modeling efforts.

To investigate the role of basin geometry in snowmelt sensitivity, a linear snow accumulation model and the Cold Regions Hydrological Modeling (CRHM) platform driven are used to estimate how hypsometry affects basin-wide snow volumes and snowmelt runoff. Area-elevation distributions for fifty basins in western Canada were extracted, normalized according to their elevation statistics, and classified into three clusters that represent top-heavy, middle, and bottom-heavy basins. Snow pillow data (Government of BC - https://aqrt.nrs.gov.bc.ca/Data/) Manual Snow Course Stations (Government of BC - http://www.env.gov.bc.ca/wsd/data_searches/snow/asws/data/allmss_archive.csv) Climate Normals (ECCC - https://climate.weather.gc.ca/climate_normals/index_e.html)

Model calibration and validation are critical in hydrological model robustness assessment. Unfortunately, the commonly-used split-sample test (SST) framework for data splitting requires modelers to make subjective decisions without clear guidelines. This large-sample SST assessment study empirically assesses how different data splitting methods influence post-validation model testing period performance, thereby identifying optimal data splitting methods under different conditions. This study investigates the performance of two lumped conceptual hydrological models calibrated and tested in 463 catchments across the United States using 50 different data splitting schemes. These schemes are established regarding the data availability, length and data recentness of the continuous calibration sub-periods (CSPs). A full-period CSP is also included in the experiment, which skips model validation. The assessment approach is novel in multiple ways including how model building decisions are framed as a decision tree problem and viewing the model building process as a formal testing period classification problem, aiming to accurately predict model success/failure in the testing period. Results span different climate and catchment conditions across a 35-year period with available data, making conclusions quite generalizable. Calibrating to older data and then validating models on newer data produces inferior model testing period performance in every single analysis conducted and should be avoided. Calibrating to the full available data and skipping model validation entirely is the most robust split-sample decision. Experimental findings remain consistent no matter how model building factors (i.e., catchments, model types, data availability, and testing periods) are varied. Results strongly support revising the traditional split-sample approach in hydrological modeling

The dataset is comprised of inputs to and outputs from the Cold Regions Hydrological Model (CRHM) when it was run as a virtual model of two classes of Canadian Prairie watersheds, as defined by Wolfe et al. (2019). These classes are Pothole Till and High Elevation Grasslands. These watersheds represented typified prairie watersheds based on physiogeography and coherent response to environmental change.Model parameters were informed by the results of Wolfe et al. (2019). The .prj files necessary to run the virtual models are included in the dataset. Climate forcing data are from the Adjusted and Homogenized Canadian Climate Dataset from a cohort of stations contained within each watershed class and cover a period from 1960-2006. There are a series of climate sensitivity scenarios that include applying a delta method to the original climate data (i.e., 1°C increments of warming, and -20%, +10%, +20% and +30% of precipitation). The .prj and .obs files for the baseline and each sensitivity scenario are included in the dataset. Model output includes hourly catchment outflow, snow water equivalent, and surface storage for the baseline and each scenario. In addition, virtual model outputs where linked to biogeochemical and biological models to calclate nutrient loading and biodiversity indicators. Nutrient load information is calculated using stream concentrations from the Pothole Till watershed class. Using the stream concentration-flow relationship, the hydrological data from the CRHM simulations were fed into these equations to estimate changes in nutrient loading. Wetland bird abundance and bird species richness are calculated using calculated pond areas that are modified from the CRHM outputs.

The dataset is comprised of inputs to and outputs from the Cold Regions Hydrological Model (CRHM) when it was run as a virtual model of the High Elevation Grasslands class, as defined by Wolfe et al. (2019). These watersheds represented typified prairie watersheds based on physiogeography and coherent response to environmental change. Model parameters were informed by the results of Wolfe et al. (2019). The .prj files necessary to run the virtual models are included in the dataset. Climate forcing data are from the Adjusted and Homogenized Canadian Climate Dataset from a cohort of stations contained within each watershed class and cover a period from 1960-2006. There are a series of climate sensitivity scenarios that include applying a delta method to the original climate data (i.e., 1°C increments of warming, and -20%, +10%, +20% and +30% of precipitation). Model output includes hourly catchment outflow, rainfall, snowfall, snow sublimation and snow water equivalent for the baseline and each scenario.

He, Z., Spence, C., Shook, K., Whitfield, C., Pomeroy, J., Wolfe, J. (2021). Virtual Watershed Model Simulations for Typified Prairie Watersheds in High Elevation Grasslands [Dataset]. Federated Research Data Repository. https://doi.org/10.20383/102.0517 Spence, C., He, Z., Shook, K. R., Mekonnen, B. A., Pomeroy, J. W., Whitfield, C. J., and Wolfe, J. D.: Assessing hydrological sensitivity of grassland basins in the Canadian Prairies to climate using a basin classification–based virtual modelling approach, Hydrol. Earth Syst. Sci. Discuss. [preprint], https://doi.org/10.5194/hess-2021-186, in review, 2021.

The dataset is comprised of inputs to and outputs from the Cold Regions Hydrological Model (CRHM) when it was run as a virtual model of the seven prairie basin classes, as defined by He et al. (2023). These watersheds represented typified prairie watersheds based on physiogeography and coherent response to environmental change. Model parameters were informed by the results of He et al. (2023). The .prj files necessary to run the virtual models are included in the dataset. Climate forcing data are from the Adjusted and Homogenized Canadian Climate Dataset from a cohort of stations contained within each watershed class and cover a period from 1960-2006. There are a series of climate sensitivity scenarios that include applying a delta method to the original climate data (i.e., 1°C increments of warming, and -20%, +10%, +20% and +30% of precipitation). Model output includes hourly catchment outflow, and depression water storage in the HRUs for the baseline and each scenario. There are also a series of wetland drainage scenarios that progressively reduced the wetland depression area.

He, Z., Shook, K., Spence, C., Pomeroy, J., Whitfield, C. (2023). Virtual Watershed Model Simulations for Typified Prairie Watersheds in Seven Basin Classes. Federated Research Data Repository. https://doi.org/10.20383/103.0815

Virtual Watershed Model Simulations for Typified Prairie Watersheds in Seven Basin Classes

M~A~P Classification Subclassification Interest Exemplar Template LatLon , Shape zoom: 8^^begin: bbox^bigmap: no^linecolor: blue^fillcolor: white^lineopacity: 1.0^linewidth: 3^longitudes: -118.6 -121.1^latitudes: 48.6 50.6^end^^begin pinarray^bigmap: no^order: latlon^49.45384, -120.56396, "Test Location 1", www.gwfnet.net/ping^49.12422, -120.25635, "Test Location 2"^50.10649, -119.35547, "Test Location 3"^49.97949, -118.93799, "Test Location 4"^end^^pin: latlon, 49.2, -119.0, "A one-line, quick pin, Another Test Location", https://www.gwfnet.net/ping^^begin: region^bigmap: no^level: 0^linecolor: red^linewidth: 2^fillcolor: yellow^fillopacity: 0.4^order: lonlat^-120.74784, 48.65478^-120.93109, 49.10886^-121.09583, 49.39167^-120.76008, 49.90798^-120.21076, 49.85798^-119.79167, 50.11667^-119.29228, 50.54367^-119.05934, 50.19086^-118.67881, 50.08579^end

AAL-ML-1 Inductively coupled plasma-optical emission spectrometry (ICP-OES) (Thermo Scientific iCAP 6300) Water quality (either major cationic elements (Al, Ca, Fe, K, Mg, Mn, Na, P, S, and Si) or trace metals (As, Cr, Cu, Ni, Pb, Sr, Zn), other trace metals upon request (B, Ag, Cd, Co, Hg, Mo, Se Tl)) per sample $65 $75 $101 AAL-ML-2 Gallery™ Discrete Analyzer (Thermo Scientific) Water quality (nutrients) per sample $30 $35 $47 AAL-ML-3a Capillary Ion Chromatograph system (Dionex ICS-5000 with auto sampler) Water quality (inorganic anions (e.g., nitrate, sulfate, chloride, bromide), organic acids (e.g., acetate, butyrate, succinate)) per sample (for 7 inorganic anions only) $40 $46 $62 AAL-ML-3b " " " " per sample (for 7 inorganic ions plus organic acids) $50 $58 $78 AAL-ML-4 Total Organic C, Total N Analyzer (Shimadzu TOC-LCPH/CPN with auto injector) Water quality (organic and inorganic carbon and total nitrogen) per sample $40 $46 $62 AAL-ML-5 Gas Chromatograph (Shimadzu) Greenhouse and standard gas analysis per sample $65 $75 $101 AAL-ML-6 Water Isotope Analyzer (Picarro L2130-i Isotope and Gas Concentration Analyzer) Water quality (δ18O and δD for Isotopic H2O) per sample $40 $46 $62 AAL-ML-7 Carbon-13 CH4/CO2 Isotope Analyzer (Picarro G2201-i Isotopic Analyzer) Greenhouse gas analysis (δ13C- CH4 and CO2) per sample $50 $58 $78 AAL-ML-8 UV-Visible Spectrophotometer (Thermo Evolution 260) Water quality (nutrients, chlorophyll-a) per sample $35 $40 $54 AAL-ML-9 Voltammeter (Metrohm 797 VA Computrace voltammeter with autosampler) Trace analysis (metal speciation) per sample $65 $75 $101 AAL-ML-10 Atomic fluorescence (AFS) (PSAnalytical PSA 10.055 Millenium Excalibur with PSA 20.400 Autosampler) Trace analysis (measures total arsenic (10 ppt) and selenium (2 ppt) in aqueous samples) per sample $65 $75 $101 AAL-ML-11 HPLC coupled to Atomic fluorescence (HPLC-UV-HG-AFS) (Agilent HPLC used in conjunction with the Millenium Excalibur) Trace analysis (selenium speciation) per sample $65 $75 $101 AAL-ML-12 Multimeterprobes (ThermoScientific Orion VersaStar) Probes for:^• pH^• Eh/ORP (redox potential)^• Electrical conductivity^• Temperature^• Dissolved oxygen per day $100 $115 $155 AAL-ML-13 Multi-mode plate Reader, Flex Station 3, Molecular Devices Automated well plate reader, Fluorescence, Absorbance, Luminescence per day $125 $144 $194 AAL-L-14 Standard equipment (1) Autoclave (Tuttnauer model 3850E Electronic Tabletop)^Analytical balance (Mettler Toledo XS205 Dual Range)^Weigh scales (Mettler Toledo New Classic MF)^Dishwasher (Miele Professional) ^Sterile laminar flow hoods (AirClean)^Oven (ThermoScientific Heratherm)^Centrifuge (ThermoScientific Sorvall ST16R)^Temperature-controlled incubators per day $100 $115 $155 AAL-ML-15 Standard equipment (2) Temperature-controlled chambers Anaerobic chambers per day $125 $144 $194

SWSL-DS-1 Drone: DJI M600 Pro^^Sensors:^Riegl Mini Vux-1^Sony RGB^^Sensor Description:^Lidar Mapping unit with Integrated Sony 24mp camera^^Capabilities:^Class 1 laser product capable of producing 100,000 shots per second in a 360 degree field of view. IMU data is 200 hz. Data is post processed in Applanix and Riegl softwares. Products include .las files that are elevation colored, Intensity scale or RGB colorized. ^^Wavelength:^905nm, ^380nm-740nm per day $1,000 $1,200 $1,400 SWSL-DS-2 Drone: DJI M600 Pro^^Sensors:^Corning Hyperspectral Camera^^Sensor Description:^Push-broom Line Imaging Spectrometer^^Capabilities:^Produces continuous imaging of narrow spectral bands over a spectral range^^Wavelength:^400nm-1000nm per day $800 $960 $1,120 SWSL-DS-3 Drone: DJI Mavic 2 Pro^^Sensors:^RGB camera^^Sensor Description:^Hasselblad 20mp camera^^Capabilities:^Still imagery and video files are created using the 20MP camera. Products are jpeg and MP3 files^^Wavelength:^400nm-700nm per day $200 $240 $280 SWSL-DS-4 Drone: eBee RTK^^Sensors:^Soda^^Sensor Description:^Aerial RGB Photogrammetry camera^^Capabilities:^Aerial photos that can be used in the SfM softwares to produce data files capable of producing DEM through external software packages^^Wavelength:^380nm-740nm per day $300 $360 $420 SWSL-DS-5 Drone: eBee X^^Sensors:^3D Soda, ^Duet T, and ^S110 NIR^^Sensor Description:^Gimbal Aerial RGB Photogrammetry camera, Dual Aerial RGB Photgrammetry and thermal mapping camera / 12mp Canon near infrared (NIR) band camera^^Capabilities:^Thermal, NIR and RGB images are produced and avalable for post processing. Flight time up to 75 min.^^Wavelength:^380nm-740nm, 7.5µm-12.5µm, G-550nm R-625nm NIR-850nm per day $300 $360 $420 SWSL-DS-6 Drone: eBee X^^Sensors:^Duet T^^Sensor Description:^Dual Aerial RGB Photogrammetry and thermal mapping camera^^Capabilities:^Thermal and RGB Images are collected and are processed using the Pix4D software^^Wavelength:^7.5µm-12.5µm per day $300 $360 $420 SWSL-DS-7 Drone: ALTA X^^Sensors:^Riegl MiniVux-2UAV/ Sony Alpha 7RIII /TetraCam Multispectral with thermal Camera^^Sensor Description:^Lidar Mapping unit with Integrated Sony 42mp camera and a Multispectral 6 narrow band camera array with thermal camera option^^Capabilities:^Class 1 laser produce capable of producing 200,000 shots per second in a 360 degree field of view. IMU data is 200 hz. Data is post processed in Applanix and Riegl softwares. Products include .las files that are elevation colored, Intensity scale or RGB colorized^^Wavelength:^905nm, 380nm-740nm / 450nm-1000nm / 7.5µm-13.5µm per day $1,500 $1,800 $2,100 SWSL-DS-8 G16 Leica Base station with Rover per day $100 $120 $140 SWSL-DS-9 Technician required (x2) per day $1,000 $1,200 $1,400 SWSL-DS-10 Lidar data processing per day $750 $900 $1,050 SWSL-DS-11 Imagery/colorization of lidar /DEM generation per day $375 $450 $525 SWSL-DS-12 Hand held 3d scanner (in lab use) per day $25 $30 $35 SWSL-DS-13 Statasys F370 3D printer includes material per day $300 $360 $420

UTABFWU-ML-1 Culturing:^Labconco Logic Class 2 Type A2 Biosafety Cabinet Work station provides personnel, product and environmental protection from hazardous particulates such as agents that require Biosafety Level 1 or 2. per day $30 $35 $42 UTABFWU-ML-2 Culturing:^Millipore Synergy UV Water Purification System Water purification system providing ultrapure water for laboratory needs (e.g., buffers and bacterial media preparation). per day $10 $12 $14 UTABFWU-ML-3 Culturing:^Powers Scientific Light & Temperature Control Growth Chamber Temperature-controlled unit allowing the use of large-volume flasks. Ability to control light intensity and light cycles. per day $30 $35 $42 UTABFWU-ML-4 Culturing:^Thermo Revco Elite Plus -86°C Freezer Ultra-Low Freezer allowing for the safe storage and preservation of biological samples. per day $20 $23 $28 UTABFWU-ML-5 Culturing:^VWR ADV Model 5000 Shaker Shaker allowing the use of large-volume flasks. per day $5 $6 $7 UTABFWU-ML-6 Lyophilization:^Labconco FreeZone 12 Plus Lyophilizer Freeze drier used to preserve perishable materials, extend shelf life or make material more convenient for transport. per day $5 $6 $7 UTABFWU-ML-7 Lyophilization:^Labconco FreeZone 2.5 Lyophilizer Freeze drier used to preserve perishable materials, extend shelf life or make material more convenient for transport. per day $10 $12 $14 UTABFWU-ML-8 Spectrophotometry:^Thermo Genesys 10S UV-Vis Spectrophotometer Spectrophotometer which features a high resolution (1.8nm) dual-beam optics, allowing simultaneous measurement of the sample with real-time reference beam correction, providing highly accurate data and photometric accuracy. per day $30 $35 $42 UTABFWU-ML-9 Spectrophotometry:^Molecular Devices SpectraMax 340pc Microplate Reader Microplate reader which supports various absorbance assay categories, such as protein quantitation, endotoxin detection, microbial growth, ELISAs and immunoassays, reporter gene assays, phosphotases/kinases, enzyme assays, and cell viability, proliferation, and cytotoxicity. per day $30 $35 $42 UTABFWU-ML-10 Mass Spectrometry:^Agilent 7700 Series ICP-MS with Autosampler Uses an inductively coupled plasma to ionize samples and create atomic and small polyatomic ions, which are then detected. Used for its ability to detect metals and several non-metals in liquid samples at very low concentrations. It can detect different isotopes of the same element, which makes it a versatile tool in isotopic labeling. per day $30 $35 $42 UTABFWU-ML-11 Water Isotope Analysis:^LGR DLT-100 Water Isotope Analyzer Measures hydrogen and oxygen isotopic composition in liquid water samples. per day $30 $35 $42 UTABFWU-ML-12 Mercury Analysis:^LECO AMA254 Mercury Analyzer Specifically designed for fast, safe, and accurate determination of trace amounts of mercury in various materials. per day $30 $35 $42 UTABFWU-ML-13 Mercury Analysis:^NIC MA-3000 Mercury Analyzer Utilizes the technique of Direct Thermal Decomposition-Gold Amalgamation-CVAAS to measure total mercury in solid, liquid, or gaseous sample matrices. per day $30 $35 $42 UTABFWU-ML-14 AFM and Raman Microscopy:^NT-MDT Ntegra Spectra Atomic Force Microscopy and Raman Spectroscopy Integration of SPM and confocal microscopy/Raman scattering spectroscopy. Owing to Tip Enhances Raman Scattering it allows carrying out spectroscopy/microscopy with up to 10 nm resolution. per day $1,000 $1,150 $1,400

Phycological Community Analysis Portable AlgaeTorch (bbe Moldaenke)per day Handheld operational unit used for the quantification of cyanobacteria and total chlorophyll. University of Toronto 1 Phycological Community Analysis Portable PhycoLabAnalyser (bbe Moldaenke) Quick, simple chlorophyll measurement with algal class differentiation. It also provides phycocyanin measurements – an indicator for cyanotoxins as well as flavours and odorous substances. University of Toronto 1 Phycological Community Analysis Portable FluoroProbe III (bbe Moldaenke) Used for creating depth profiles of different phycological divisions. Individual profiles during the measurements are taken for green algae, blue-green algae/cyanobacteria, diatoms/dinoflagellates and cryptophytes (others can be added, e.g. Planktothrix rubescens). University of Toronto 1 Phycological Community Analysis Portable PHYTO-PAM-II^(Walz) Facilitates measurements of mixed algae populations for composition analysis and differences in the photosynthetic performance of up to four algae groups within one sample. University of Toronto 1 Phycological Community Analysis Portable FlowCam Cyano^(Fluid Imaging Tecnologies) ^ For detecting and verifying cyanobacteria presence and biovolume calculations. The system automatically identifies cyanobacteria from other algae using excitation wavelengths, phycocyanin fluorescence measurement, and image recognition software. University of Toronto 1 Field Equipment ThermoSafe 450 Dry Ice Storage Chest; 85 lb Pellet Capacity^^(Two units available) Chests contain extra-thick foam plastic insulation core. It takes six days for 85 lb of dry ice to sublimate to 5 lb. University of Toronto 2 Field Equipment Universal Percussion Corer/ Extruding Device ^^(Six units available) Used for paleoecological analyses. Instruments can be modified for lake or wetland coring. University of Toronto 6 Field Equipment Filter Holder Vacuum Manifolds (three positions) + Vaccum pressure pump (MilliporeSigma)^^(Three units available) Provides vacuum support for simultaneous filtration. University of Toronto 3 Boat (16’) GRIZZLY® 1648 SCw/ Mercury® FourStroke 20 ELPT FourStroke Boat, Motor, & Trailer For accessing lakes University of Toronto 1 Boat (14’) Scamper Alumnimum boats/ Mercury® FourStroke 20 ELPT FourStroke Boat, Motor, & Trailer^^(Two units available) For accessing lakes University of Toronto 2

UTHAAL-DS-1 Phycological Community Analysis:^AlgaeTorch (bbe Moldaenke) Portable^Description:^Handheld operational unit used for the quantification of cyanobacteria and total chlorophyll. per day $200 $230 $280 UTHAAL-DS-2 "" per week $1,000 $1,150 $1,400 UTHAAL-DS-3 Phycological Community Analysis:^PhycoLabAnalyser (bbe Moldaenke) Portable^Description:^Quick, simple chlorophyll measurement with algal class differentiation. It also provides phycocyanin measurements – an indicator for cyanotoxins as well as flavours and odorous substances. per day $300 $345 $420 UTHAAL-DS-4 "" per week $1,800 $2,070 $2,520 UTHAAL-DS-5 Phycological Community Analysis:^FluoroProbe III (bbe Moldaenke) Portable^Description:^Used for creating depth profiles of different phycological divisions. Individual profiles during the measurements are taken for green algae, blue-green algae/cyanobacteria, diatoms/dinoflagellates and cryptophytes (others can be added, e.g. Planktothrix rubescens). per day $300 $345 $420 UTHAAL-DS-6 "" per week $1,800 $2,070 $2,520 UTHAAL-DS-7 Phycological Community Analysis:^PHYTO-PAM-II (Walz) Portable^Description:^Facilitate measurements of mixed algae populations for composition analysis and differences in the photosynthetic performance of up to four algae groups within one sample. per day $250 $288 $350 UTHAAL-DS-8 "" per week $1,000 $1,150 $1,400 UTHAAL-DS-9 Phycological Community Analysis:^FlowCam Cyano (Fluid Imaging Technologies) Portable^Description:^For detecting and verifying cyanobacteria presence and biovolume calculations. The system automatically identifies cyanobacteria from other algae using excitation wavelengths, phycocyanin fluorescence measurement, and image recognition software. per day $400 $460 $560 UTHAAL-DS-10 "" per week $2,400 $2,760 $3,360 UTHAAL-DS-11 Field Equipment:^ThermoSafe 450 Dry Ice Storage Chest; 85 lb Pellet Capacity^(Two units available)^Description:^Chests contain extra-thick foam plastic insulation core. It takes six days for 85 lb of dry ice to sublimate to 5 lb. per week $100 $115 $140 UTHAAL-DS-12 "" per month $300 $345 $420 UTHAAL-DS-13 Field Equipment:^Universal Percussion Corer/ Extruding Device^(Six units available)^Description:^Used for paleoecological analyses. Instruments can be modified for lake or wetland coring. per day $50 $58 $70 UTHAAL-DS-14 "" per week $200 $230 $280 UTHAAL-DS-15 Field Equipment:^Filter Holder Vacuum Manifolds (three positions) + Vacuum pressure pump (MilliporeSigma)^(Three units available)^Description:^Provides vacuum support for simultaneous filtration. per day $100 $115 $140 UTHAAL-DS-16 "" per week $600 $690 $840 UTHAAL-DS-17 Boats:^(16’) GRIZZLY® 1648 SCw/ Mercury® FourStroke 20 ELPT FourStroke Boat, Motor, & Trailer^Description:^For accessing lakes per week $1,000 $1,150 $1,400 UTHAAL-DS-18 "" per month $3,000 $3,450 $4,200 UTHAAL-DS-19 Boats:^(14’) Scamper Alumnimum boats/ Mercury® FourStroke 20 ELPT FourStroke Boat, Motor, & Trailer^(Two units available)^Description:^For accessing lakes per week $500 $575 $700 UTHAAL-DS-20 "" per month $1,500 $1,725 $2,100

UTHAAL-ML-1 Flow Cytometry:^CytoFLEX Flow Cytometer (Beckman Coulter) Cell analysis system used for the qualitative and quantitative measurement of biological and physical properties of cells and other particles. Application flexibility, including an optional 96-well Plate Loader, 13 bandpass filters, and two lasers (Blue and Red). per hour $100 $115 $140 UTHAAL-ML-2 Flow Cytometry:^PhytoCyt flow cytometer (Turner Designs) Portable Cell analysis system used for many applications (e.g., cell counting, viability assays, physiological characterization, etc.). The optical filters and lasers have been optimized to detect the endogenous fluorophores common to phytoplankton, including phycoerythrin, chlorophyll and phycocyanins. One of the detectors has been optimized for green fluorescence detection, which allows labelling with exogenous reagents. per day $200 $230 $280 UTHAAL-ML-3 "" "" per week $1,200 $1,380 $1,680 UTHAAL-ML-4 Microscopy:^IX73 Inverted LED Fluorescence Microscope (Olympus) Fluorescence microscopy is a major tool used to monitor cell physiology and aid in cell identification. The system has imaging capabilities for standard imaging tasks. per hour $50 $58 $70 UTHAAL-ML-5 Microscopy:^BX 42Fl Light Microscopy (Olympus) Suitable for various applications and has imaging capabilities for standard imaging tasks. per hour $30 $35 $42 UTHAAL-ML-6 Microscopy:^ECLIPSE Ts2 Inverted Microscopes (Nikon) Suitable for various applications and has imaging capabilities for standard imaging tasks. per hour $30 $35 $42 UTHAAL-ML-7 Spectrophotometry:^Multiskan™ FC Microplate Photometer (Thermo Scientific) (340–850 nm) Provides fast and accurate measurements for 96-well plates. The wavelength range of 340– 850 nm and is used for a wide variety of applications such as ELISA immunoassays, protein quantification, endotoxin, cytotoxicity and proliferation assays, enzyme assays and growth curves. per hour $50 $58 $70 UTHAAL-ML-8 Spectrophotometry:^Cary 300 UV-Visible Spectrophotometer The wavelength range of 190–900 nm and is used for a wide variety of applications (wavelength accuracy ± 0.2 nm). per hour $50 $58 $70 UTHAAL-ML-9 Fluorescence Spectrophotometer:^Cary Eclipse is a Fluorescence Spectrophotometer Flexible fluorescence spectrometer with multiple data collection modes, including fluorescence, phosphorescence, chemiluminescence, bioluminescence, and time-resolved phosphorescence. Operational emission: 200 - 900 nm Operational excitation: 200 - 900 nm per hour $50 $58 $70 UTHAAL-ML-10 Algal Culturing:^Innova® S44i - Stackable Incubator Shaker^(Three units available) Temperature-controlled (refrigerated) shaker allowing the use of high-volume flasks. Photosynthetic LED light bank provides the ability to evenly light the entire platform across a wide intensity range. per week (per unit) $500 $575 $700 UTHAAL-ML-11 "" "" per month (per unit) $1,500 $1,725 $2,100 UTHAAL-ML-12 Algal Culturing:^Panasonic growth incubators^(Two units available) Temperature-controlled (refrigerated) unit allowing the use of high-volume flasks. Ability to control light intensity and light cycles. per week (per unit) $500 $575 $700 UTHAAL-ML-13 "" "" per month (per unit) $1,500 $1,725 $2,100 UTHAAL-ML-14 Algal Culturing:^Milli-Q® Benchtop Lab Water Purification Systems Water purification system providing ultrapure water for laboratory needs (e.g., buffers and bacterial media preparation). **Complimentary service when using laboratory facilities** **Complimentary service when using laboratory facilities** **Complimentary service when using laboratory facilities** UTHAAL-ML-15 Algal Culturing:^Multi-Cultivator MC 1000-OD Small-scale cultivation device developed for the cultivation of multiple samples. The instrument is primarily intended for the synchronous growth of algae, bacteria or cyanobacteria under defined conditions with a wide range of applications (i.e. toxicological and eco-toxicological testing, optimization of cultivation conditions, phenotypization of various strains). per week $200 $230 $280 UTHAAL-ML-16 "" "" per month $600 $690 $840 UTHAAL-ML-17 Algal Culturing:^Purifier Logic+ Class II, Type A2 Biosafety Cabinets Workstation provides personnel, product, and environmental protection from hazardous particulates such as agents that require Biosafety Level 1 or 2. **Complimentary service when using laboratory facilities** **Complimentary service when using laboratory facilities** **Complimentary service when using laboratory facilities**

^A Ten sample minimum. ^B Samples must be freeze-dried and homogenized. ^C For samples submitted ‘ready-to-analyze’, please contact us for advice regarding sample preparation and submission. ^D Please specify target analytes. ^E For add-on with other analytes only. ^F For samples submitted ‘ready-to-analyze’ (i.e., pre-weighed in tin capsules) ^G A $5 discount will apply to samples received ‘ready-to-analyze’ ^H Per class of co-extractable routine analytes (please see: https://www.trentu.ca/wqc/facilities-services/services) ^I Solid samples include biota, sediments, soils, etc. All prices are subject to change Services subject to availability.

The Bridge to Land Water Sky is Canada's only Indigenous-led Living Lab, inspiring a more resilient agriculture industry and the next wave of farmers. As the only Indigenous-led Living Lab in Canada, the Bridge to Land Water Sky will focus on identifying barriers Indigenous people face when entering the Agricultural Industry and will celebrate Indigenous knowledge as a key factor in building a more innovative and climate-resilient agricultural industry with global impacts. Goals: 1. Improve land management strategies for the mitigation of greenhouse gas emissions and improve carbon sequestration. We will test beneficial practices for revitalization and improvement of our land and soils. 2. Increase food security and sovereignty in Indigenous and non-Indigenous communities to benefit local food production and medicinal and traditional plants for Indigenous communities. 3. Protect biodiversity and water to support healthy ecosystems, people, and environmental co-benefits, including soil health, riparian and wetland buffers, and species at risk (flora, fauna). 4. Create employment and learning opportunities for youth and communities throughout the region, including out-of-classroom learning opportunities, community engagement, practical experiences, mentorships, internships, and relationship-building opportunities. 5. Reimagining Indigenous landowner and producer relationships to strengthen partnerships and mutually beneficial economic and environmental outcomes.

Special Issue of Hydrology and Earth System Sciences (HESS) Understanding and predicting Earth system and hydrological change in cold regions (edited by S. Carey, C. DeBeer, J. Hanesiak, Y. Li, J. Pomeroy, B. Schaefli, M. Weiler, and H. Wheater): https://hess.copernicus.org/articles/special_issue919.html Short description: CCRN observes, diagnoses, and predicts environmental change in the Saskatchewan and Mackenzie River Basins. To integrate existing and new sources of data with improved predictive and observational tools to understand, diagnose and predict interactions amongst the cryospheric, ecological, hydrological, and climatic components of the changing Earth system at multiple scales, with a geographic focus on Western Canada’s rapidly changing cold interior. The cold interior of Western Canada east of the Continental Divide has one of the world's most extreme and variable climates and is experiencing rapid environmental change. In a region which includes a multiplicity of globally-important natural resources and sustains 80% of Canada's agricultural production, changing climate is changing the land, its vegetation and its water. There is an urgent need to understand the nature of these changes, and to develop the improved modelling tools needed to manage uncertain futures. The CCRN brings together the unique expertise of a team of 50 university and government scientists and international collaborators from multiple disciplines to address these challenging and globally-important issues. CCRN integrates existing and new experimental data with modelling and remote sensing products to understand, diagnose and predict changing land, water and climate, and their interactions and feedbacks, for this important region. CCRN uses a network of world class Water, Ecosystem, Cryosphere, and Climate (WECC) observatories to study the detailed connections among changing climate, ecosystems and water in the permafrost regions of the Sub-arctic, the Boreal Forest, the Western Cordillera, and the Prairies. CCRN integrates these and other data to understand the changing regional climate and its effects on large-scale Earth system change and the region's major rivers - the Saskatchewan, Mackenzie and Peace-Athabasca. Current ability to model these effects is limited, yet models are essential to understand and manage change. CCRN works with government, industry, water managers, First Nations communities and other stakeholders to deliver the improved hydrological, ecological and climate modelling tools needed to understand, predict and manage uncertain climate and water futures. CCRN addresses issues of importance not only to Canada, but also the world, and continues to contribute to the work of Canada's Federal, Provincial and Territorial governments, NASA and the Canadian Space Agency, and the World Climate Research Programme.

Core Modelling and Forecasting is a GWF core team which performs world-class, leading-edge water science for cold regions to address the needs of the Canadian economy in adapting to change and managing risks associated with the uncertain water futures and extreme events brought about by climate change. Its research is delivered under eight themes: 1. Spatial Meteorological Forcing Data (https://gwf.usask.ca/core-modelling/research-themes/spatial-meteorological-forcing-data.php); 2. Geospatial Intelligence (https://gwf.usask.ca/core-modelling/research-themes/geospatial-intelligence.php); 3. Current Generation Hydrological Modelling (https://gwf.usask.ca/core-modelling/research-themes/current-generation-modelling.php); 4. Next Generation Hydrological Modelling (https://gwf.usask.ca/core-modelling/research-themes/next-generation-hydrological-modelling.php); 5. Water Resources Management (https://gwf.usask.ca/core-modelling/research-themes/water-resources-management.php); 6. Water Quality Modelling (https://gwf.usask.ca/core-modelling/research-themes/water-quality-modelling.php); 7. Hydrological Forecasting(https://gwf.usask.ca/core-modelling/research-themes/hydrological-forecasting.php); and 8. Hydro-economics Modelling (https://gwf.usask.ca/core-modelling/research-themes/hydro-economics.php).

Since the report of the RNA aptamer for theophylline, theophylline has become a key molecule in chemical biology for designing RNA switches and riboswitches. In addition, theophylline is an important drug for treating airway diseases including asthma. The classic RNA aptamer with excellent selectivity for theophylline has been used to design biosensors, although DNA aptamers are more desirable for stability and cost considerations. In this work, we selected DNA aptamers for theophylline, and all the top sequences shared the same binding motifs. Binding was confirmed using isothermal titration calorimetry and a nuclease digestion assay, showing a dissociation constant (Kd) around 0.5 μM theophylline. The Theo2201 aptamer can be truncated down to 23-mer while still has a Kd of 9.8 μM. The selectivity for theophylline over caffeine is around 250,000-fold based on a strand-displacement assay, which was more than 20-fold higher compared to the classic RNA aptamer. For other tested analogs, the DNA aptamer also showed better selectivity. Using the structure-switching aptamer sensor design method, a detection limit of 17 nM theophylline was achieved in the selection buffer, and a detection limit of 31 nM was obtained in 10% serum.

A DNA Aptamer for Theophylline with Ultrahigh Selectivity Reminiscent of the Classic RNA Aptamer

Crasto, N., Hopkinson, C., Forbes, D. L., Lesack, L., Marsh, P., Spooner, I., & van der Sanden, J. J. (2015). A LiDAR-based decision-tree classification of open water surfaces in an Arctic delta. Remote Sensing of Environment, 164, 90-102. https://doi.org/10.1016/j.rse.2015.04.011

A LiDAR-based decision-tree classification of open water surfaces in an Arctic delta

Seasonal temperature trend and ice phenology in Great Slave lake (GSL), are strongly influenced by warmer inflow from Slave river. The Slave river flows to GSL through Slave river delta (SRD), bringing a rise in temperature that triggers the ice break-up process of the lake. Slave river discharge is subject to multiple stressors including climate warming and upstream water activities, which in turn, directly affects the GSL break-up process. Consequently, monitoring the break-up process at SRD, where the river connects to the lake, serves as an indicator to better understand the cascading effects on GSL ice break-up. This research aims to develop random forest (RF) models to monitor the SRD ice break-up processes, using a combination of satellite images with optical sensors at high spatial resolution, including Landsat-5, Landsat-8, Sentinel-2a, and Sentinel-2b. The RF models were trained using manually selected training pixels to classify ice, open water, and cloud within the SRD. The break-up start period is defined by minimum and maximum thresholds of 60% and 90% on ice fraction, which are a trade-off between maximizing the available images and not including images that are taken after the break-up start. The results show high variability in the rate of break-up within delta using images in recent years with better temporal resolution. Furthermore, a statistically significant trend is observed from 1984 to 2023 using the Mann-Kendall test, with a p-value of 0.05. This study is of great significance to northern and high latitude communities who rely on lake ice for activities such as transportation, and sustenance. Moreover, the break-up of the delta plays a pivotal role in supplying nutrients and sediments, and also in the occurrence of spring flooding. Therefore, the outcomes of this study can be leveraged to shape effective water resource management policies based on the regional characteristics of climate and hydrological patterns.

A Machine Learning Approach to Classify Open Water and Ice Cover on Slave River Delta

Evolutionary coupling is a well investigated phenomenon in software maintenance research and practice. Association rules and two related measures, support and confidence, have been used to identify evolutionary coupling among program entities. However, these measures only emphasize the co-change (i.e., changing together) frequency of entities and cannot determine whether the entities co-evolved by experiencing related changes. Consequently, the approach reports false positives and fails to detect evolutionary coupling among infrequently co-changed entities. We propose a new measure, identifier correspondence (id-correspondence), that quantifies the extent to which changes that occurred to the co-changed entities are related based on identifier similarity. Identifiers are the names given to different program entities such as variables, methods, classes, packages, interfaces, structures, unions etc. We use Dice-Sørensen co-efficient for measuring lexical similarity between the identifiers involved in the changed lines of the co-changed entities. Our investigation on thousands of revisions from nine subject systems covering three programming languages shows that id-correspondence can considerably improve the detection accuracy of evolutionary coupling. It outperforms the existing state-of-the-art evolutionary coupling based techniques with significantly higher recall and F-score in predicting future co-change candidates.

A devastating, flood-producing rainstorm occurred over southern Alberta, Canada, from 19 to 22 June 2013. The long-lived, heavy rainfall event was a result of complex interplays between topographic, synoptic, and convective processes that rendered an accurate simulation of this event a challenging task. In this study, the Weather Research and Forecasting (WRF) Model was used to simulate this event and was validated against several observation datasets. Both the timing and location of the model precipitation agree closely with the observations, indicating that the WRF Model is capable of reproducing this type of severe event. Sensitivity tests with different microphysics schemes were conducted and evaluated using equitable threat and bias frequency scores. The WRF double-moment 6-class microphysics scheme (WDM6) generally performed better when compared with other schemes. The application of a conventional convective/stratiform separation algorithm shows that convective activity was dominant during the early stages, then evolved into predominantly stratiform precipitation later in the event. The HYSPLIT back-trajectory analysis and regional water budget assessments using WRF simulation output suggest that the moisture for the precipitation was mainly from recycling antecedent soil moisture through evaporation and evapotranspiration over the Canadian Prairies and the U.S. Great Plains. This analysis also shows that a small fraction of the moisture can be traced back to the northeastern Pacific, and direct uptake from the Gulf of Mexico was not a significant source in this event.

Obadia, M., DeVries, B., Park, G., Merchant, M. A., and Berg, A. A.: A Spatiotemporal Surface Water Classification in the Mackenzie Delta and Tuktoyaktuk Peninsula, American Geophysical Union Meeting, Virtual, 2021

A Spatiotemporal Surface Water Classification in the Mackenzie Delta and Tuktoyaktuk Peninsula

Climate change is rapidly altering composition, structure, and functioning of the boreal biome, across North America often broadly categorized into ecoregions. The resulting complex changes in different ecoregions present a challenge for efforts to accurately simulate carbon dioxide (CO2) and energy exchanges between boreal forests and the atmosphere with terrestrial ecosystem models (TEMs). Eddy covariance measurements provide valuable information for evaluating the performance of TEMs and guiding their development. Here, we compiled a boreal forest model benchmarking dataset for North America by harmonizing eddy covariance and supporting measurements from eight black spruce (Picea mariana)-dominated, mature forest stands. The eight forest stands, located in six boreal ecoregions of North America, differ in stand characteristics, disturbance history, climate, permafrost conditions and soil properties. By compiling various data streams, the benchmarking dataset comprises data to parameterize, force, and evaluate TEMs. Specifically, it includes half-hourly, gap-filled meteorological forcing data, ancillary data essential for model parameterization, and half-hourly, gap-filled or partitioned component flux data on CO2 (net ecosystem production, gross primary production [GPP], and ecosystem respiration [ER]) and energy (latent [LE] and sensible heat [H]) and their daily aggregates screened based on half-hourly gap-filling quality criteria. We present a case study with the Canadian Land Surface Scheme Including Biogeochemical Cycles (CLASSIC) to: (1) demonstrate the utility of our dataset to benchmark TEMs and (2) provide guidance for model development and refinement. Model skill was evaluated using several statistical metrics and further examined through the flux responses to their environmental controls. Our results suggest that CLASSIC tended to overestimate GPP and ER among all stands. Model performance regarding the energy fluxes (i.e., LE and H) varied greatly among the stands and exhibited a moderate correlation with latitude. We identified strong relationships between simulated fluxes and their environmental controls except for H, thus highlighting current strengths and limitations of CLASSIC

A boreal forest model benchmarking dataset for North America: a case study with the Canadian Land Surface Scheme Including Biogeochemical Cycles (CLASSIC)

The freeze-thaw cycle associated with climatic seasonality is a common phenomenon in cold regions affecting a wide range of subsurface processes. Due to the complex and highly nonlinear nature of the associated hydrologic processes, transient freeze-thaw dynamics are conventionally quantified in a numerical way. Here we present a hybrid analytical-numerical scheme for solving one-dimensional soil (or porous media) temperature profiles when the soil profile is subjected to unidirectional freezing (or thawing) conditions. This scheme divides the partially-frozen soil into multi-layers, each with constant thermal parameters and fixed-temperature boundaries. Temperature profiles within each layer were obtained by solving multiple moving-boundary problems. The proposed hybrid analytical-numerical scheme was tested into a freezing test of silty clay in a permafrost region on the Qinghai-Tibetan Plateau, and its solution was in good agreement with the finite element numerical solution. Results show that the proposed multi-layer method adapted well to the changes in unfrozen water content and thermal properties of soil over a wide range of subzero temperatures. By contrast, the freezing front's migration rate and penetration depth calculated by Neumann's classical solution, which only considers two zones (frozen and unfrozen), was found to be underestimated. As for our proposed multi-layer solution, by dividing the subsurface domain into many layers with smaller proportion ratios (thinner layers close to the freezing front), there was a slower penetration rate of the freezing front resulting in shallower penetration depth. The predicted profiles of temperature, thermal conductivity and diffusivity, heat flux, and dynamics of the freezing front were significantly impacted by the shape of the soil freezing curves and the magnitude of soil grain thermal conductivity, especially for the accuracy of long-term predictions.

Land surface models (LSMs) are used to simulate water and energy fluxes between the land surface and atmosphere. These simulations are useful for water resources management, drought and flood prediction, and numerical climate/weather prediction. However, the usefulness of LSMs are dependent by their ability to reproduce states and fluxes realistically. Accurate measurements of water storage are useful to calibrate and validate LSMs outputs. Geological Weighing Lysimeters (GWLs) are instruments that can provide field-scale estimates of integrated total water storage within a soil profile. We use field estimates of total water storage and subsurface storage to critically evaluate two different land surface models: the Modélisation Environnementale communautaire - Surface Hydrology (MESH) which uses the Canadian Land Surface Scheme (CLASS), and the Structure for Unifying Multiple Modeling Alternatives: (SUMMA). These models have differences in how the processes and properties of the land surface are represented. We attempted to parameterize each model in an equivalent manner, to minimize model differences. Both models were able to reproduce observations of total water storage and subsurface storage reasonably well. However, there were inconsistencies in the simulated timing of snowmelt; depth of soil freezing; total evapotranspiration; partitioning of evaporation between soil evaporation and evaporation of intercepted water; and soil drainage. No one model emerged as better overall, though each model had specific strengths and weaknesses that we describe. Insights from this study can be used to improve model physics and performance.

A decision-tree classification for low-lying complex land cover types within the zone of discontinuous permafrost

The freeze-thaw cycle associated with climatic seasonality is a common phenomenon in cold regions affecting a wide range of subsurface processes. Due to the complex and highly nonlinear nature of the associated hydrologic processes, transient freeze-thaw dynamics are conventionally quantified in a numerical way. Here we present a hybrid analytical-numerical scheme for solving one-dimensional soil (or porous media) temperature profiles when the soil profile is subjected to unidirectional freezing (or thawing) conditions. This scheme divides the partially-frozen soil into multi-layers, each with constant thermal parameters and fixed-temperature boundaries. Temperature profiles within each layer were obtained by solving multiple moving-boundary problems. The proposed hybrid analytical-numerical scheme was tested into a freezing test of silty clay in a permafrost region on the Qinghai-Tibetan Plateau, and its solution was in good agreement with the finite element numerical solution. Results show that the proposed multi-layer method adapted well to the changes in unfrozen water content and thermal properties of soil over a wide range of subzero temperatures. By contrast, the freezing front's migration rate and penetration depth calculated by Neumann's classical solution, which only considers two zones (frozen and unfrozen), was found to be underestimated. As for our proposed multi-layer solution, by dividing the subsurface domain into many layers with smaller proportion ratios (thinner layers close to the freezing front), there was a slower penetration rate of the freezing front resulting in shallower penetration depth. The predicted profiles of temperature, thermal conductivity and diffusivity, heat flux, and dynamics of the freezing front were significantly impacted by the shape of the soil freezing curves and the magnitude of soil grain thermal conductivity, especially for the accuracy of long-term predictions.

A code clone is a pair of code fragments, within or between software systems that are similar. Since code clones often negatively impact the maintainability of a software system, several code clone detection techniques and tools have been proposed and studied over the last decade. However, the clone detection tools are not always perfect and their clone detection reports often contain a number of false positives or irrelevant clones from specific project management or user perspective. To detect all possible similar source code patterns in general, the clone detection tools work on the syntax level while lacking user-specific preferences. This often means the clones must be manually inspected before analysis in order to remove those false positives from consideration. This manual clone validation effort is very time-consuming and often error-prone, in particular for large-scale clone detection. In this paper, we propose a machine learning approach for automating the validation process. First, a training dataset is built by taking code clones from several clone detection tools for different subject systems and then manually validating those clones. Second, several features are extracted from those clones to train the machine learning model by the proposed approach. The trained algorithm is then used to automatically validate clones without human inspection. Thus the proposed approach can be used to remove the false positive clones from the detection results, automatically evaluate the precision of any clone detectors for any given set of datasets, evaluate existing clone benchmark datasets, or even be used to build new clone benchmarks and datasets with minimum effort. In an experiment with clones detected by several clone detectors in several different software systems, we found our approach has an accuracy of up to 87.4% when compared against the manual validation by multiple expert judges. The proposed method also shows better results in several comparative studies with the existing related approaches for clone classification.

Study region The Peace-Athabasca Delta, a Ramsar Wetland of International Importance in northeastern Alberta, is protected within Wood Buffalo National Park and contributes to its UNESCO World Heritage status yet is threatened by climate change and upstream energy projects. Study focus Recent drawdown of the delta’s abundant shallow lakes and rivers has deteriorated vital habitat for wildlife and impaired navigation routes. Here, we report continuous measurements at ~50 lakes during open-water seasons of 2018 and 2019 to improve understanding of hydrological processes causing lake-level variation. New hydrological insights for the region Analyses reveal four patterns of lake-level variation attributable to influential hydrological processes, which provide the basis for a new lake classification scheme: 1) ‘Drawdown’ (≥15 cm decline) by evaporation and/or outflow after ice-jam floods, 2) ‘Stable’ lake levels (<15 cm change) sustained by rainfall, 3) ‘Gradual Rise’ by inundation from the open-drainage network, and 4) ‘Rapid Rise’ by input of river floodwater. River flooding during the open-water season is an under-recognized recharge mechanism yet occurred extensively in the Athabasca sector and appears to be a common occurrence based on the Athabasca River hydrometric record. Lake-level loggers show strong ability to track shifts in hydrological processes, and can be integrated with other methods to decipher their causes and ecological consequences across water-rich landscapes.

A new lake classification scheme for the Peace-Athabasca Delta (Canada) characterizes hydrological processes that cause lake-level variation

This paper presents a novel data fusion technique for improving the snow cover monitoring for a mesoscale Alpine region, in particular in those areas where two information sources disagree. The presented methodological innovation consists in the integration of remote-sensing data products and the numerical simulation results by means of a machine learning classifier (support vector machine), capable to extract information from their quality measures. This differs from the existing approaches where remote sensing is only used for model tuning or data assimilation. The technique has been tested to generate a time series of about 1300 snow maps for the period between October 2012 and July 2016. The results show an average agreement between the fused product and the reference ground data of 96%, compared to 90% of the moderate-resolution imaging spectroradiometer (MODIS) data product and 92% of the numerical model simulation. Moreover, one of the most important results is observed from the analysis of snow cover area (SCA) time series, where the fused product seems to overcome the well-known underestimation of snow in forest of the MODIS product, by accurately reproducing the SCA peaks of winter season.

Mountain water resources are of particular importance for downstream populations but are threatened by decreasing water storage in snowpack and glaciers. Groundwater contribution to mountain streamflow, once assumed to be relatively small, is now understood to represent an important water source to streams. This review presents an overview of research on groundwater in high mountain environments (As classified by Meybeck et al. (2001) as very high, high, and mid-altitude mountains). Coarse geomorphic units, like talus, alluvium, and moraines, are important stores and conduits for high mountain groundwater. Bedrock aquifers contribute to catchment streamflow through shallow, weathered bedrock but also to higher order streams and central valley aquifers through deep fracture flow and mountain-block recharge. Tracer and water balance studies have shown that groundwater contributes substantially to streamflow in many high mountain catchments, particularly during low-flow periods. The percentage of streamflow attributable to groundwater varies greatly through time and between watersheds depending on the geology, topography, climate, and spatial scale. Recharge to high mountain aquifers is spatially variable and comes from a combination of infiltration from rain, snowmelt, and glacier melt, as well as concentrated recharge beneath losing streams, or through fractures and swallow holes. Recent advances suggest that high mountain groundwater may provide some resilience—at least temporarily—to climate-driven glacier and snowpack recession. A paucity of field data and the heterogeneity of alpine landscapes remain important challenges, but new data sources, tracers, and modeling methods continue to expand our understanding of high mountain groundwater flow.

Wastewater surveillance (WWS) is useful to better understand the spreading of coronavirus disease 2019 (COVID-19) in communities, which can help design and implement suitable mitigation measures. The main objective of this study was to develop the Wastewater Viral Load Risk Index (WWVLRI) for three Saskatchewan cities to offer a simple metric to interpret WWS. The index was developed by considering relationships between reproduction number, clinical data, daily per capita concentrations of virus particles in wastewater, and weekly viral load change rate. Trends of daily per capita concentrations of SARS-CoV-2 in wastewater for Saskatoon, Prince Albert, and North Battleford were similar during the pandemic, suggesting that per capita viral load can be useful to quantitatively compare wastewater signals among cities and develop an effective and comprehensible WWVLRI. The effective reproduction number (Rt) and the daily per capita efficiency adjusted viral load thresholds of 85 × 106 and 200 × 106 N2 gene counts (gc)/population day (pd) were determined. These values with rates of change were used to categorize the potential for COVID-19 outbreaks and subsequent declines. The weekly average was considered ‘low risk’ when the per capita viral load was 85 × 106 N2 gc/pd. A ‘medium risk’ occurs when the per capita copies were between 85 × 106 and 200 × 106 N2 gc/pd. with a rate of change <100 %. The start of an outbreak is indicated by a ‘medium-high’ risk classification when the week-over-week rate of change was >100 %, and the absolute magnitude of concentrations of viral particles was >85 × 106 N2 gc/pd. Lastly, a ‘high risk’ occurs when the viral load exceeds 200 × 106 N2 gc/pd. This methodology provides a valuable resource for decision-makers and health authorities, specifically given the limitation of COVID-19 surveillance based on clinical data.

Classification and clustering approaches provide a means to group watersheds according to similar attributes, functions, or behaviours, and can aid in managing natural resources. Although they are widely used, approaches based on hydrological response parameters restrict analyses to regions where well-developed hydrological records exist, and overlook factors contributing to other management concerns, including biogeochemistry and ecology. In the Canadian Prairie, hydrometric gauging is sparse and often seasonal. Moreover, large areas are endorheic and the landscape is highly modified by human activity, complicating classification based solely on hydrological parameters. We compiled climate, geological, topographical, and land-cover data from the Prairie and conducted a classification of watersheds using a hierarchical clustering of principal components. Seven classes were identified based on the clustering of watersheds, including those distinguishing southern Manitoba, the pothole region, river valleys, and grasslands. Important defining variables were climate, elevation, surficial geology, wetland distribution, and land cover. In particular, three classes occur almost exclusively within regions that tend not to contribute to major river systems, and collectively encompass the majority of the study area. The gross difference in key characteristics across the classes suggests that future water management and climate change may carry with them heterogeneous sets of implications for water security across the Prairie. This emphasizes the importance of developing management strategies that target sub-regions expected to behave coherently as current human-induced changes to the landscape will affect how watersheds react to change. The study provides the first classification of watersheds within the Prairie based on climatic and biophysical attributes, with the framework used being applicable to other regions where hydrometric data are sparse. Our findings provide a foundation for addressing questions related to hydrological, biogeochemical, and ecological behaviours at a regional level, enhancing the capacity to address issues of water security.

Watershed classification can identify regions expected to respond similarly to disturbance. Methods should extend beyond hydrology to include other environmental questions, such as ecology and water quality. We developed a classification for the Canadian Prairie and identified seven classes defined by watershed characteristics, including elevation, climate, wetland density, and surficial geology. Results provide a basis for evaluating watershed response to land management and climate condition.

A watershed classification approach that looks beyond hydrology: application to a semi-arid, agricultural region in Canada

Elshamy, M., Abdelhamed, M., Razavi, S., Pietroniro, A., Pomeroy, J., Wheater, H. (May 2021) Advances in Permafrost Modelling: Enhancing MESH/CLASS to represent Permafrost Dynamics. Global Water Futures (GWF) 2021 Fourth Annual Open Science Meeting, virtual event, May 17, 2021 to May 19, 2021

Advances in Permafrost Modelling: Enhancing MESH/CLASS to represent Permafrost Dynamics

Alpine regions receive large volumes of precipitation and are important to local and regional water balances, particularly during baseflow periods of winter cold and summer drought when the larger basin area is frozen and/or water limited. Alpine headwaters in western Canada are expected to warm and receive more precipitation during the coming decades, with implications for groundwater recharge and streamflow generation within these systems and the regional river networks to which they contribute. Throughout the North, thawing peat plateaus and other ice-rich permafrost features are resulting in an increased extent of thermokarst and wetland land cover. This transition places infrastructure and water resources at risk as the structural integrity and reliable flow paths previously maintained by the frozen soils become compromised. Alpine systems are particularly susceptible to hydrological change due to the amplification of climate warming with both latitude and elevation. The inherent spatial heterogeneity of these same systems makes attempts to quantify the impacts of climate change on current and future basin water balance even more challenging, yet few field studies of alpine hydrology have been conducted in northern Canada. Specifically, no hydrological field studies have previously occurred within alpine shrub tundra terrain overlapping the Taiga Cordilleran Ecozone and/or the Mackenzie River basin. The objective of this dissertation is to characterize the spatial and temporal variability in hydrological processes controlling the water balance of an alpine shrub tundra basin. Chapter Two presents five cover classes that are hydrologically distinct based on physiographic, surface, and subsurface characteristics. Glaciofluvial uplands are isolated from the channel network, routing all inputs to aquifer recharge. Peat plateaus have ice-rich permafrost at depth, resulting in limited storage and efficient subsurface runoff to neighbouring fens. Fen and riparian swamp iii cover classes both act as primary contributors to the channel network, although some fen areas may be isolated thermokarst features. These thermokarst features lose water via taliks recharging aquifers and/or evaporative loss from surface ponds. In the context of climate change, permafrost thaw will result in the replacement of peat plateaus with fens, such that both storage capacity and groundwater connections will expand. A conceptual model presents the basin storage compartments and expected flow paths linking the cover classes to each other and the larger area beyond the topographical extent of the study basin. Chapter Three utilizes the land cover classification established in Chapter Two to investigate temporal differences in 2019 open water season basin water balance. During the freshet, a large volume of snowmelt was received, and storage capacity was limited by shallow frost tables and bedfast ice. As a result, runoff generation was highly efficient and streamflow volumes large. The exception to this is the glaciofluvial upland, which channeled all snowmelt to aquifer recharge. As the freshet transitioned to summer, small magnitude rain events began to occur, and evapotranspiration became the primary means of basin water loss. Furthermore, groundwater exchange became more important to the basin water balance, with groundwater discharge from springs in the headwaters sustaining streamflow and channel bed infiltration becoming more prominent as bedfast ice and channel banks thawed. As the summer progressed, cumulative storage, streamflow, and evapotranspiration rates declined as groundwater discharge became the primary input and groundwater recharge the primary output. As climate change continues, a greater proportion of precipitation will be received as rain and the open water season will extend, resulting in a greater proportion of total annual basin outputs occurring via aquifer recharge, although shrubification and permafrost thaw may result in greater influence of evapotranspiration. Chapter Four assesses the basin runoff response following discrete precipitation events and utilizes stable isotope analysis to establish seasonally distinct source water contributions, evaporative influence, and subsurface flow paths during the 2019 open water season. The large volume of snowmelt received during the freshet caused peak streamflow rates, but only 8 % of total freshet discharge was isotopically designated as event water at the main basin outlet. In comparison, the maximum daily and total freshet event water fraction was reduced at the headwater subbasin outlet, where spring sources of groundwater discharge were more influential on streamflow. During the summer months, headwater subbasin streamflow was volumetrically and isotopically unresponsive to rain events and groundwater discharge continued to dominate. At the main outlet, early summer runoff response volumes and event water contributions following precipitation events were greatly reduced, in part due to the smaller magnitude of rain input volumes compared to snowmelt, but also due to the increase in fen storage capacity. By the late summer, the frost table also reached the mineral substrates at depth in the riparian swamp, extending the flow path for rain received by this cover class. As a result, late summer streamflow following rain was composed of even less event water and the hydrograph response was characterized by a lower peak and extended recession limb compared to the early summer event. This dissertation greatly enhances our understanding of the hydrological role alpine tundra plays in sustaining regional river systems via both surface streamflow and aquifer recharge. These findings provide the model structure and parameter values necessary for future hydrological modelling efforts that seek to better represent the contribution of these headwater subbasins to larger regional river systems under current and future climatic conditions.

Object-oriented programming has had a significant impact on software development because it provides programmers with a clear structure of a large system. It encapsulates data and operations into objects, groups objects into classes and dynamically binds operations to program code. With the emergence of multi-core processors, application developers have to explore concurrent programming to take full advantage of multi-core technology. However, when it comes to concurrent programming, object-oriented programming remains elusive as a useful programming tool. Most object-oriented programming languages do have some extensions for concurrency, but concurrency is implemented independently of objects: for example, concurrency in Java is managed separately with the Thread object. We employ a programming model called Lime that combines action systems tightly with object-oriented programming and implements concurrency by extending classes with actions and guarded methods. This provides programmers with a unified and straightforward design view for a concurrent object-oriented program. In this work, using coroutines with guarded methods and actions is proposed as a means of implementing the concurrency extension for objects. Mapping objects to coroutines can result in stack overflow as the number of objects increases. A dynamically segmented stack mechanism, which does not introduce runtime overhead, is implemented to support large-scale concurrency. Since Lime allows guarded methods and actions to "get stuck," a new user-level cooperative scheduler, and a fast coroutine context switch mechanism are implemented to improve the performance. Compared with the traditional segmented stack mechanisms, the new dynamically segmented stack mechanism gets equal performance for more common scenarios. Besides, it outperforms the contemporary stack mechanisms for deep recursion scenarios. Above all, Lime does not only provide the programmers with a unified and straightforward object-oriented programming model for concurrency, but also accomplishes a better performance than concurrent programming languages such as Erlang and Go, in fine-grained, highly concurrent benchmarks.

In recent years, Deep Learning (DL) models have widely been applied to develop safety and security critical systems. The recent evolvement of Deep Neural Networks (DNNs) is the key reason behind the unprecedented achievements in image classification, object detection, medical image analysis, speech recog nition, and autonomous driving. However, DL models often remain a black box for their practitioners due to the lack of interpretability and explainability. DL practitioners generally use standard metrics such as Precision, Recall, and F1 score to evaluate the performance of DL models on the test dataset. However, as high-quality test data is not frequently accessed, the expected level of accuracy of these standard metrics on test datasets cannot justify the trustworthiness of testing adequacy, generality and robustness of DL models. The way we ensure the quality of DL models is still in its infancy; hence, a scalable DL model testing frame work is highly demanded in the context of software testing. The existing techniques for testing traditional software systems could not be directly applicable to DL models because of the fundamental difference in pro gramming paradigm, systems development methodologies, and processes. However, several testing metrics (e.g., Neuron Coverage (NC), Confusion and Bias error metrics, and Multi-granularity metrics) have been proposed leveraging the concept of test coverage in traditional software testing to measure the robustness of DL models and the quality of the test datasets. Although test coverage is highly effective to test traditional software systems, the effectiveness of DL coverage metrics must be evaluated in testing the robustness of DL models and measuring the quality of the test datasets. In addition, the selected testing metrics work on the activated neurons of a DL model. In our study, we consider the neuron count of a DL model differently than the existing studies. For example, according to our calculation the LeNet-5 model has 6508 neurons whereas other studies consider the LeNet-5 model contains 268 neurons only. Therefore, it is also important to in vestigate how neurons’ concept (e.g., the idea of having neurons in the DL model and the way we calculate the number of neurons a DL model does have) impact the testing metrics. In this thesis, we thus conduct an exploratory study for evaluating the effectiveness of the testing metrics to test DL models not only in measuring their robustness but also in assessing the quality of the test datasets. Furthermore, since selected testing metrics work on the activated neurons of a DL model, we also investigate the impact of the neurons’ concepts on the testing metrics. To conduct our experiments, we select popular publicly available datasets (e.g., MNIST, Fashion MNIST, CIFAR-10, ImageNet and so on) and train DL models on them. We also select sate-of-the-art DL models (e.g., VGG-16, VGG-19, ResNet-50, ResNet-101 and so on) trained on the ImageNet dataset. Our experimental results demonstrate that whatever the neuron’s concepts are, NC and Multi-granularity testing metrics are ineffective in evaluating the robustness of DL models and in assessing the quality of the test datasets. In addition, the selection of threshold values has a negligible impact on the NC metric. Increasing the coverage values of the Multi-granularity testing metrics can not separate regular test data from adversarial test data. Our exploratory study also shows that the DL models still make accurate predictions with higher coverage values of Multi-granularity metrics than the false predictions. Therefore, it is not always true that increasing coverage values of the Multi-granularity testing metrics find more defects of DL models. Finally, the Precision and Recall scores show that the Confusion and Bias error metrics are adequate to detect class-level violations of the DL models.

Changes in the water cycle influence the energy balance of the Earth. The water cycle is represented using the water balance equation, in which Evapotranspiration (ET) is a vital parameter. One of the main drivers of the change in ET within a specific area is the change in land cover. This study focuses on estimating ET across the Upper Gundar River Basin located in the state of Tamil Nadu, India. Notable features of this landscape include agriculture throughout the year supported using an extensive network of tanks and borewells, and the presence of Prosopis juliflora, a widely prevalent invasive species known to consume groundwater and moisture. Due to the lack of spatial variability in point ET measurements, ET models using remote sensing imagery as the main forcing data have been widely used to assess the spatial variability and temporal variability based on the principle of surface energy balance. These models are collectively referred to as Surface Energy Balance (SEB) models. The model used in our study is the Surface Energy Balance Algorithm for Land (SEBAL) model to estimate ET for two periods of the year, indicating mid-summer and the end of the northeast monsoon for the years 2006, 2014 and 2021. Since land cover changes drive ET, land cover classification and seasonal change detection are also performed for the same time periods. Imagery from Landsat satellites is used, and one image is chosen to represent the specific season. The major land cover classes chosen in our study are water, pre-growth, agriculture, Prosopis juliflora (prosopis), barren land, and exposed soil. Along with the Landsat imagery, to run SEBAL, Aster DEM is used along with in-situ weather data and GLDAS data. Over 90% levels of overall accuracy were achieved for all year-season combinations for the land cover classification. Using SEBAL, Actual Evapotranspiration (ETa) for all the classes is calculated except the water classes. Due to the lack of in-situ measurements, an intermodal comparison was performed with the EEFlux product available at the same resolution derived using the METRIC algorithm using land cover classes as units of comparison. The comparisons are carried out using correlation coefficient (r), root mean squared error (RMSE), and mean values. Highest mean values were observed for either the agriculture or prosopis class, and the lowest mean value was exhibited by the exposed soil class on all occasions. Within all summers, considering all the years, the average correlation coefficient and RMSE were 0.8, 1.2 mm/day, and for monsoon, the averages were 0.5 and 0.85 mm/day, indicating increased proximity during the monsoon season between SEBAL and EEFlux. Similarly, the range of mean values between classes in summer is 2.12 mm/day, 1.36 mm/day in the monsoon. In terms of the energy fluxes used to determine ETa, a decrease in monsoon is observed for soil heat flux (G), instantaneous net radiant energy (Rn_inst), and net radiation in a day (Rn_24). For sensible heat flux (H), classes with vegetation tend to have lower values in comparison to the classes without vegetation. Finally, average water outflux is calculated encompassing all classes by multiplying the area of a class with mean ETa, and the values observed in summer and monsoon alternatively for the years 2006, 2014, and 2021 in m3/day are 5,142,212, 3,534,906, 2,954,897, 4,046,322, 5,369,191, 4,512,596.

Weather and climate are major factors influencing worldwide wildfire activity. This study assesses surface and atmospheric conditions associated with the 2014 extreme wildfires in the Northwest Territories (NWT) of Canada. Hot and dry conditions led to the NWT experiencing the most severe wildfire season in its recorded history. The season included a record number of cloud-to-ground lightning flashes and set a record for area burned. Lightning was the dominant ignition source and accounted for about 95% of the wildfires. Prolonged periods of smoke led to dramatic reductions in visibility, frequent road closures, and reduced air quality resulting in numerous health alerts. Temporal and spatial patterns of lightning characteristics in 2014, derived from Canadian Lightning Detection Network data, were different from those in other years with, for example, far more positive flashes from 0600 to 1200 utc (midnight to 6:00 am local time). The highest fraction of positive cloud-to-ground flashes (43.1%) occurred in the smoke-dominated North Slave region, which was more than in the Dehcho, South Slave, or Sahtu regions. Mid-tropospheric atmospheric circulation over a large region that included the NWT was classified into the six most common summer patterns. Results showed that ridging and ridge displacements occurred more frequently during 2014 although lightning was associated with all circulation patterns. This study has advanced the understanding of the roles of weather, lightning, and mid-tropospheric circulation patterns associated with extreme wildfires in northwestern Canada.

Yirdaw S.Z., K. R. Snelgrove, F. R. Seglenieks, C. O. Agboma and E.D. Soulis. (2009). An assessment of the WATCLASS hydrological model result of the Mackenzie River basin using the GRACE satellite total water storage measurement. Hydrological Processes pp3391-3400

An assessment of the WATCLASS hydrological model result of the Mackenzie River basin using the GRACE satellite total water storage measurement

Eddy covariance (EC) is one of the most effective ways to directly observe evaporation from a lake surface. However, the deployment of EC systems on lakes is costly and technically challenging which engenders a need for accurate modelling of evaporation from reservoirs for effective management. This study aims to 1) refactor the Canadian Small Lakes Model (CSLM) into modern high-level programming languages in open-source repositories and 2) evaluate evaporation estimates from the CSLM using nine years of EC observations of a pit-lake in Northern Alberta. The CSLM is a 1-D physical lake model simulating a mixing layer and arbitrary thick skin layer which interfaces with the atmosphere and includes a module for ice dynamics. It was developed to interface with the Canadian Global Coupled models as part of the surface classification scheme, and thus utilizes widely accessible forcing data. In this study the CSLM evaporation estimates are also compared to a commonly used bulk transfer method of estimating evaporation. In general, the CSLM had smaller open-water season error (RMSE of 0.70 mm day-1) than the bulk transfer method (RMSE of 0.83 mm day-1). However, if EC data is available, further improvement can be gained by using an Artificial Neural Network to adjust the modelled fluxes (RMSE of 0.51 mm day-1). This final step can be very useful for gap-filling missing data from lake observation networks as there has been recent attention on the limited coverage of direct open water evaporation observations in the literature.

An ever-growing Canadian urban population could be severely impacted by increase in temperature. Canada’s mean temperature is projected to increase by 6-8°C towards the end of the 21st century. The consequence of rising temperatures is an increased likelihood of extreme temperature events like heatwaves and wildfires. The thesis aims to assess changes in extreme temperature in large Canadian urban areas. The research will help in developing mitigation measures like urban planning, which help cope with changing temperature extremes. Predicting urban temperature change will require rigorous assessment of climate models, to account for the uncertainty in projecting temperature in large urban agglomerates. CMIP6 ensemble of models, provide an opportunity for assessment of urban-based projections. The models however, would need to be of fine resolution to fully capture its variability since urban temperature is heavily influenced by local urban features that contribute to Urban heat island (UHI). Historical maximum and minimum temperature trends are analyzed for eighteen urban areas in the Canada with population greater than 250,000 and use twelve CMIP6 models of fine resolution (<1°), and four tier-one emission scenarios to assess maximum, minimum, and mean temperature trends in future. An efficient observation dataset, Serially based station data (SCDNA), was used as a reference observation dataset and a novel bias-correction technique, the Semi-Parametric Quantile Mapping method (SPQM), was used to bias-correct future temperature data. Extreme temperature events were analyzed with the help of eight selected indices of the Expert Team on Climate Change Detection and Indices (ETCCDI), across all the emission scenarios for all the cities in the study. The indices were computed for the entire future time-period (2021-2100) and for three time-slices, T1 (2021-2050), T2 (2040-2070) and T3 (2070-2100) to assess temporal variability. The magnitude, frequency, and duration of the occurrence of extreme events can be analyzed effectively using the ETCCDI indices, classified as absolute, threshold, and duration Indices and percentile indices. The historical temperature trends in Canadian cities were found to be related with urban features like elevation and population-growth but not strongly linked with urban area. Other features of UHI were deemed essential to understand the transitioning of historical and future temperature trends in Canada. Four emission scenarios predict increasing mean temperatures in all Canadian cities, except for the optimistic emission scenario (SSP1-2.6), which shows a marginal decreasing trend in the last quarter of the 21st century. Uneven changes are noted in all the projected indices, for example, in the annual maxima of daily maximum temperature (TXx), i.e., an increase of 0.5 °C and 0.6 °C per decade over the T1 and T2 respectively, and 0.99°C for T3 for the SSP5-8.5. Results show faster rates of warming across Canadian cities especially for the higher emission scenarios (SSP3-7.0 and SSP5-8.5). Spatial trends of extreme temperature indices correlate with temperature trends in individual climate zones in Canada, and the cities associated with a zone, expectedly experience similar trends. Cities in the Prairies and the Great Lakes zones, experience the highest increasing trends over the absolute and threshold indices in the higher emission scenarios, whereas the cities in the Canadian coasts experience higher increasing trends in the percentile indices. Lower emission scenarios also point towards increasing spatial trends in all Canadian cities. The coastal cities also experience the highest trends for the warm-spell duration index (WSDI) and a decreasing trend in the cold-spell duration index (CSDI). Spatial patterns of duration indices in the Canadian coastal cities point towards hotter summers, and milder winters, whereas the cities in the Canadian prairies, the Great Lakes, and Quebec will experience hotter summers with longer durations of extremely hot weather, in addition to persistence of harsher winters. Temperature projections have several applications, for example, in civil engineering applications, where temperature has a great role in the estimation and assessment of concrete and reinforcement deterioration. Another field of research is urban-based mortality studies, a consequence of the increasing frequency and duration of extreme temperature events. Heat-wave analysis, estimated through extreme temperature indices, forms the basis for estimating mortality rates from heat waves and other extreme temperature events. Climate models and CMIP6 models have systematic errors in their development and hence can only predict temperature projections with a limited degree of confidence. An extension of the work in this thesis could be the use of various model performance indicators, that quantitatively assess the performance of temperature projections made by CMIP6 models in Canadian cities. The future temperature projections and estimations of heat waves provide a scientific basis for a better understanding of the temperature patterns and temperature-related extreme events in Canadian cities and thus help facilitate climate change adaptation.

Wastewater treatment plants (WWTPs) are traditionally designed to target the removal of contaminants such as total suspended solids, phosphorous, biological oxygen demand, and ammonia. Recent changes to the Federal Wastewater Systems Effluent Regulations (WSER) in Canada require all WWTPs to be operating with secondary treatment or equivalent by 2021. Upgrades being implemented at WWTPs across the country will improve the quality of final effluent discharged into the receiving waters. However, over the past several years, contaminants of emerging concern such as pharmaceuticals, personal care products, and endocrine disrupting compounds have become widely prevalent in wastewater. These compounds are not monitored or targets for removal in Canada causing them to be routinely discharged into surface waters. The Grand River watershed is the largest watershed in southern Ontario and receives effluent discharge from 30 WWTPs. Several studies have been conducted in the Grand River to assess the impacts of effluent discharge on fish found in the river. The two largest WWTPs are the Kitchener and Waterloo WWTPs, both of which having recently undergone upgrades to improve nitrification processes and improve the overall effluent quality. Studies linked effluent from the plant’s pre-upgrade, to several adverse impacts on fish, such as severe cases of intersex and altered hormone production. Upgrades at the Kitchener WWTP were shown to reduce these impacts on fish. Effluent from both Waterloo and Kitchener have been collected and analyzed for pharmaceuticals and estrogens since before the upgrades providing the unique opportunity to evaluate the change in effluent quality and composition over time. In addition to the Kitchener and Waterloo WWTPs, nine secondary WWTPs across southern Ontario were studied to compare the composition of influent and effluent as well as evaluate the apparent removal of various pharmaceuticals and estrogens. Despite all the plants being classified as having secondary level treatment there was a considerable amount of variability in their ability to treat the incoming influent. Pharmaceuticals of interest were ibuprofen, naproxen, carbamazepine, and venlafaxine because of their different behaviour during treatment. Ibuprofen and naproxen were significantly reduced at all plants, with an increased reduction at plants achieving better nitrification. Carbamazepine and venlafaxine are recalcitrant and remained untreated. Of the estrogens measured, estriol was significantly reduced across all plants while 17α-ethinylestradiol had no difference post treatment. Estrone and 17β-estradiol were both reduced to varying degrees and were more influenced by external factors such as treatment type and biotransformation. Although there was compound specific variability, the total estrogenicity was significantly reduced post treatment at all plants except those with poor nitrification. Through the analysis of the pharmaceuticals and estrogens as well as nutrient data, nitrification was related to the apparent removal of these non-target compounds (although a direct relationship cannot be established). This correlates with the findings at the Kitchener and Waterloo WWTPs. With the introduction of nitrification at both plants there was a decrease in ammonia concentrations, improved treatment of ibuprofen, naproxen, estrone, and estradiol. There was also a decrease in the total estrogenicity of the effluent discharged from the plants. While venlafaxine, carbamazepine and ethinylestradiol concentrations remained unchanged post upgrades. Understanding the composition and concentration of contaminants in influent and effluent can provide insight into treatment processes influencing the removal and biotransformation of these compounds. This information is important when deciding on the regulation of these contaminants in effluent discharge. Chemical analysis of these compounds is also critical in developing relationships between contaminant exposure to impacts found in the Grand River. This data can aid in validating predictive models linking contaminants to specific biological endpoints.

Mutton, D., Arain, M.A, Davison B., Princz, D. (2022) Analysis of carbon and water cycle variability in Canadian watersheds using coupled MESH-CLASSIC model. Assessing the impact of climate change on the McKenzie Creek in the Great Lakes region. Global Water Futures Annual Science Meeting, University of Saskatoon, Saskatoon, Saskatchewan, Canada (held virtually), May 16-18, 2022.

Analysis of carbon and water cycle variability in Canadian watersheds using coupled MESH-CLASSIC model. Assessing the impact of climate change on the McKenzie Creek in the Great Lakes region

Recent advances in the Mod´elisation Environmentale Communautaire Surface and Hydrology system (MESH) allows for vector-based routing to better represent the reality of the catchment structure and water processes within the catchment. MESH has also been coupled with the Canadian Land Surface Scheme including Biogeochemical Cycles (CLASSIC) allows for the simulation of carbon, water, and energy cycles at a catchment scale. In this study, the MESH-CLASSIC model was tested in three catchments across Canada, the Groundhog River catchment (a Boreal Forest catchment in northern Ontario), the Big Creek catchment (a managed agricultural catchment in southern Ontario), and the White Gull Creek catchment (a boreal forest and wetland in northern Saskatchewan). The vector-based MESH-CLASSIC model simulations were performed for historic and future climate change scenarios, RCP 4.5 and 8.5. Model biases in simulating hydrological processes such as stream flow, evapotranspiration, snow mass, and soil moisture will be evaluated, and model performance will be evaluated with underperforming areas identified with the intent to improve these processes and the model’s capability. Early research suggests that the model overestimates streamflow and snow mass estimates, as well as predicts major events will start earlier than they do.

Research uses coupled MESH-CLASSIC model to predict future carbon, water, and energy cycle in three watersheds across Canada

Analysis of coupled MESH-CLASSIC model performance in Canadian watersheds

Biomaterials used in biomedical implants, diagnostic devices, and in-situ sensors, all face the issue of biofouling. Surface modification of biomaterial surfaces with antifouling polymers can prevent non-specific adsorption of proteins and other bio-foulants onto these surfaces. Although there are many antifouling polymers to chose from, getting the polymers onto different materials is challenging as the surface modification process is dependent on the substrate’s surface chemistry. This limits the kinds of materials that are able to be modified, especially in devices made with several materials that must be modified as a single unit. Therefore, the goal of this research is to develop an effective antifouling surface modification that is compatible with different types and classes of biomaterials. A three-step modification approach was taken to form dense antifouling polymer brushes. The surfaces were first activated using oxygen plasma to increase the density of surface hydroxyl groups. Next, a silane coupling agent with an Atom Transfer Radical Polymerization (ATRP) initiator was attached to the activated surfaces. Finally, an antifouling zwitterionic monomer was polymerized on the surface using an aqueous controlled living radical polymerization technique, Surface Initiated - Activators Regenerated by Electron Transfer – Atom Transfer Radical Polymerization (SI-ARGET-ATRP). Two zwitterionic antifouling polymers, poly(carboxybetaine methacrylate) (pCBMA), and poly(sulfobetaine methacrylate) (pSBMA) were investigated. Clinically- and environmentally-relevant materials were studied and include poly(dimethylsiloxane) (PDMS), poly(ether ether ketone) (PEEK), titanium, silicon, and 3D printed stainless steel. Water contact angle (WCA) analysis showed that surfaces modified with zwitterionic polymers became more hydrophilic. WCA analysis may not be suitable for evaluating non-modified 3D printed surfaces due to their poor surface finish, and this material requires further surface topography characterization. Atomic force microscopy (AFM) and ellipsometry showed that the zwitterionic polymer layers did not necessarily have to be thick to produce their hydrophilic effect. AFM also revealed that each step of the surface modification process produced different roughness effects on all of the different surfaces. The zwitterionic layer with the smoother surface tended to better resist bovine serum albumin (BSA) adsorption. Radiolabelled BSA experiments showed reduced fouling on all 2D samples but to different degrees. The pCBMA modification was not successful in preventing BSA fouling on 3D printed 316L stainless steel. Full or partial BSA fouling may be due to the hydrolytic instability of the silane coupling agent, used to form covalent bonds between the antifouling polymers and the different surfaces, although further investigation is required to validate this hypothesis. Improving the long-term stability of silanes or research with other multi-surface compatible coupling agents that have better long-term stability in aqueous solutions should be pursued.

Ice jam formation is a key concern for many rivers in cold regions. Ice jams can lead to destructive floods during the ice season and create major disturbances in the aquatic environment. In the past years, numerous approaches, including numerical, empirical and data-driven models, have been applied to understand and potentially help reduce the detrimental impacts of ice-jam formation and flooding on riverine communities. Although these approaches are capable of solving many ice-related problems, they still have some limitations. Recent advancements in machine learning offers many methodological opportunities to deal with a plethora of data. This study explores the use of modelling output (e.g. hydrological and hydraulic model results and global circulation model (GCM) output) in various machine learning algorithms to predict river-ice hydraulic processes, such as ice-jam formation and mid-winter breakup along the Saint John River, which is a transboundary and transborder river in North America. A hydrodynamic model (RIVICE) was applied to simulate hundreds of river ice scenarios and pre-breakup hydraulic conditions of the river. The simulated hydraulic conditions of the river were then used in machine learning classification algorithms to forecast the location of ice-jam formation during spring breakup. Since recent changes in climatic conditions make the river more vulnerable to ice-jam formation, such as mid-winter breakup, this study examined the impacts of future climate on mid-winter breakup. A hydrological model (MESH) was applied to derive the streamflow conditions of the river under future climatic conditions. The temperature and precipitation data were derived from GCM output for similar future climatic conditions. The data were then applied in a machine-learning classification algorithm to predict the probability of mid-winter breakup along the river. This research will help understand the hydro-climatic impacts on the ice-jam processes and offer new knowledge to manage these complex river ice processes.

Synthetic aperture radar (SAR) is a widely used tool for Earth observation activities. It is particularly effective during times of persistent cloud cover, low light conditions, or where in situ measurements are challenging. The intensity measured by a polarimetric SAR has proven effective for characterizing Arctic tundra landscapes due to the unique backscattering signatures associated with different cover types. However, recently, there has been increased interest in exploiting novel interferometric SAR (InSAR) techniques that rely on both the amplitude and absolute phase of a pair of acquisitions to produce coherence measurements, although the simultaneous use of both intensity and interferometric coherence in Arctic tundra image classification has not been widely tested. In this study, a time series of dual-polarimetric (VV, VH) Sentinel-1 SAR/InSAR data collected over one growing season, in addition to a digital elevation model (DEM), was used to characterize an Arctic tundra study site spanning a hydrologically dynamic coastal delta, open tundra, and high topographic relief from mountainous terrain. SAR intensity and coherence patterns based on repeat-pass interferometry were analyzed in terms of ecological structure (i.e., graminoid, or woody) and hydrology (i.e., wet, or dry) using machine learning methods. Six hydro-ecological cover types were delineated using time-series statistical descriptors (i.e., mean, standard deviation, etc.) as model inputs. Model evaluations indicated SAR intensity to have better predictive power than coherence, especially for wet landcover classes due to temporal decorrelation. However, accuracies improved when both intensity and coherence were used, highlighting the complementarity of these two measures. Combining time-series SAR/InSAR data with terrain derivatives resulted in the highest per-class F1 score values, ranging from 0.682 to 0.955. The developed methodology is independent of atmospheric conditions (i.e., cloud cover or sunlight) as it does not rely on optical information, and thus can be regularly updated over forthcoming seasons or annually to support ecosystem monitoring.

Causes of software architectural change are clas- sified as perfective, preventive, corrective, and adaptive. Change classification is used to promote common approaches for address- ing similar changes, produce appropriate design documentation for a release, construct a developer’s profile, form a balanced team, support code review, etc. However, automated architectural change classification techniques are in their infancy, perhaps due to the lack of a benchmark dataset and the need for extensive human involvement. To address these shortcomings, we present a benchmark dataset and a text classifier for determining the architectural change rationale from commit descriptions. First, we explored source code properties for change classification independent of project activity descriptions and found poor outcomes. Next, through extensive analysis, we identified the challenges of classifying architectural change from text and pro- posed a new classifier that uses concept tokens derived from the concept analysis of change samples. We also studied the sensitivity of change classification of various types of tokens present in commit messages. The experimental outcomes employing 10- fold and cross-project validation techniques with five popular open-source systems show that the F1 score of our proposed classifier is around 70%. The precision and recall are mostly consistent among all categories of change and more promising than competing methods for text classification.

In Canada's Arctic tundra region, permafrost is continuous, and the landscape is rich in patterned features. Polygonal terrain, which includes both high- and low-centered features and their wet trenches below, is considered to be high-latitude wetlands in the continuous permafrost region. These prominent hydrological features retain and transport water within widespread ice-wedge networks and govern many ecosystem dynamics. Due to the meter-scale spatial gradients of these processes, mapping of polygonal wetland networks necessitates high-resolution imagery. To date, most studies have used optical imagery for this task; however, these sensors are affected by cloud cover and polar darkness, limiting image availability and repeatability. Thus, our overall objective was to evaluate high-resolution hybrid compact polarimetric (HCP) imagery from the recently launched Radarsat Constellation Mission (RCM), in fusion with ArcticDEM topographic data, for Arctic landscape mapping with a focus on polygonal wetlands. RCM's 5 m Stripmap beam mode, which has yet to be studied for such a task, represents an innovative HCP synthetic aperture radar (SAR) data source since it allows for polarimetric decomposition methods, despite being a dual-pol system. Within this study, we present a seven-input channel Convolutional Neural Network (CNN) model for the classification of ice-wedge dominated landscapes. A range of model hyperparameters as well as the effect of SAR speckle filtering on classification accuracy, have been examined. The optimized CNN achieved a high classification accuracy (0.931 mean Intersection Over Union; mIOU) for three semantic classes representative of the study area, namely polygonal wetlands, open water, and uplands. These results were superior in comparison to a benchmark machine learning (ML) Random Forest (RF) algorithm, thus demonstrating the proposed CNN's potential for regional-scale permafrost feature mapping. Notably, the optimal CNN architecture used unfiltered SAR data as input, underscoring the importance of spatial resolution when classifying polygonal terrain with deep learning (DL). These findings have important implications regarding the design, tuning, and sensitivity of CNN algorithms, and on the efficacy of HCP SAR, for mapping Arctic regions.

The seasonal temperature trends and ice phenology in the Great Slave Lake (GSL) are significantly influenced by inflow from the Slave River. The river undergoes a sequence of mechanical break-ups all the way to the GSL, initiating the GSL break-up process. Additionally, upstream water management practices impact the discharge of the Slave River and, consequently, the ice break-up of the GSL. Therefore, monitoring the break-up process at the Slave River Delta (SRD), where the river meets the lake, is crucial for understanding the cascading effects of upstream activities on GSL ice break-up. This research aimed to use Random Forest (RF) models to monitor the ice break-up processes at the SRD using a combination of satellite images with relatively high spatial resolution, including Landsat-5, Landsat-8, Sentinel-2a, and Sentinel-2b. The RF models were trained using selected training pixels to classify ice, open water, and cloud. The onset of break-up was determined by data-driven thresholds on the ice fraction in images with less than 20% cloud coverage. Analysis of break-up timing from 1984 to 2023 revealed a significant earlier trend using the Mann–Kendall test with a p-value of 0.05. Furthermore, break-up data in recent years show a high degree of variability in the break-up rate using images in recent years with better temporal resolution.

Significant challenges from changes in climate and land use face sustainable water use in the Canadian Prairies ecozone. The region has experienced significant warming since the mid-20th century, and continued warming of an additional 2 ∘C by 2050 is expected. This paper aims to enhance understanding of climate controls on Prairie basin hydrology through numerical model experiments. It approaches this by developing a basin-classification-based virtual modelling framework for a portion of the Prairie region and applying the modelling framework to investigate the hydrological sensitivity of one Prairie basin class (High Elevation Grasslands) to changes in climate. High Elevation Grasslands dominate much of central and southern Alberta and parts of south-western Saskatchewan, with outliers in eastern Saskatchewan and western Manitoba. The experiments revealed that High Elevation Grassland snowpacks are highly sensitive to changes in climate but that this varies geographically. Spring maximum snow water equivalent in grasslands decreases 8 % ∘C−1 of warming. Climate scenario simulations indicated that a 2 ∘C increase in temperature requires at least an increase of 20 % in mean annual precipitation for there to be enough additional snowfall to compensate for enhanced melt losses. The sensitivity in runoff is less linear and varies substantially across the study domain: simulations using 6 ∘C of warming, and a 30 % increase in mean annual precipitation yields simulated decreases in annual runoff of 40 % in climates of the western Prairie but 55 % increases in climates of eastern portions. These results can be used to identify those areas of the region that are most sensitive to climate change and highlight focus areas for monitoring and adaptation. The results also demonstrate how a basin classification-based virtual modelling framework can be applied to evaluate regional-scale impacts of climate change with relatively high spatial resolution in a robust, effective and efficient manner.

Assessing hydrological sensitivity of grassland basins in the Canadian Prairies to climate using a basin classification-based virtual modelling approach

Wetland drainage has been pervasive in the North American Prairie Pothole Region. There is strong evidence that this drainage increases the hydrological connectivity of previously isolated wetlands and, in turn, runoff response to snowmelt and rainfall. It can be hard to disentangle the role of climate from the influence of wetland drainage in observed records. In this study, a basin-classification-based virtual modelling approach is described that can isolate these effects on runoff regimes. The basin class which was examined, entitled Pothole Till, extends throughout much of Canada's portion of the Prairie Pothole Region. Three knowledge gaps were addressed. First, it was determined that the spatial pattern in which wetlands are drained has little influence on how much the runoff regime was altered. Second, no threshold could be identified below which wetland drainage has no effect on the runoff regime, with drainage thresholds as low as 10 % in the area being evaluated. Third, wetter regions were less sensitive to drainage as they tend to be better hydrologically connected, even in the absence of drainage. Low flows were the least affected by drainage. Conversely, during extremely wet years, runoff depths could double as the result of complete wetland removal. Simulated median annual runoff depths were the most responsive, potentially tripling under typical conditions with high degrees of wetland drainage. As storage capacity is removed from the landscape through wetland drainage, the size of the storage deficit of median years begins to decrease and to converge on those of the extreme wet years. Model simulations of flood frequency suggest that, because of these changes in antecedent conditions, precipitation that once could generate a median event with wetland drainage can generate what would have been a maximum event without wetland drainage. The advantage of the basin-classification-based virtual modelling approach employed here is that it simulated a long period that included a wide variety of precipitation and antecedent storage conditions across a diversity of wetland complexes. This has allowed seemingly disparate results of past research to be put into context and finds that conflicting results are often only because of differences in spatial scale and temporal scope of investigation. A conceptual framework is provided that shows, in general, how annual runoff in different climatic and drainage situations will likely respond to wetland drainage in the Prairie Pothole Region.

Spence, C., He, Z., Shook, K. R., Pomeroy, J. W., Whitfield, C. J., and Wolfe, J. D. (2022). Assessing runoff sensitivity of North American Prairie Pothole Region basins to wetland drainage using a basin classification-based virtual modelling approach, Hydrol. Earth Syst. Sci., 26, 5555-5575, https://doi.org/10.5194/hess-26-5555-2022

Assessing runoff sensitivity of North American Prairie Pothole Region basins to wetland drainage using a basin classification-based virtual modelling approach

Climate change presents unique challenges for grape growers across the world. In Ontario, three distinct viticultural regions are experiencing climatic shifts towards warmer growing seasons. According to historical records collected from Environment and Climate Change Canada, Lake Erie North Shore has transitioned from an intermediate-to-warm growing season classification, the Niagara Peninsula from the lower to upper limits of the intermediate zone, and Prince Edward County from cool to intermediate, when analyzing their average growing season temperatures. Terroir is directly related to vine water status, an indicator of grapevine stress. Biophysical responses controlled by air temperature and precipitation include fluctuations in vapour pressure deficits, evapotranspiration, and water-use-efficiency rates, as well as soil water content levels. By conducting an extensive literature review, the development of a conceptual model addresses how variations in climatic controls, under the scope of climate change, may influence grapevine water status, biophysical responses, and associated production outcomes for Ontario vineyards. Cool-to-intermediate air temperatures, when paired with increased precipitation will lead to no or low vine stress, increasing photosynthesis and transpiration rates, and variable plant water-use-efficiency levels, producing higher yields and lower quality grapes, if no management strategies are applied. Oppositely, higher air temperature as a product of climate change, when paired with variable precipitation may produce mild-to-severe stress, reducing yield, and increasing grape quality. With the appropriate management strategies, both traditional and new, growers may be able to accommodate for the influence of climate change on their vineyards.

Lakes are regarded as sentinels of change, where shifts in environmental conditions significantly affect lake phenology. A significant consequence of the change is the perceived increase in the frequency, magnitude, and severity of algal blooms in lakes globally. Algal blooms/increased productivity in lakes pose significant ecological, economic and health risks, impacting fisheries, tourism, and freshwater access. The impacts of external nutrient loading from anthropogenic sources are well documented; however, blooms have been observed to occur in even remote lakes. Climate change is a hypothesized driver of these recent algal bloom trends, such as increasing global air temperatures, water temperatures, lake ice loss, precipitation intensity, and drought. Past research on the impact of climatic drivers on algal biomass dynamics has often been limited to lab, mesocosm, or short termed observations, due to limited in situ data. New remote sensing data products make use of historic multispectral satellite image archives to provide greater spatial and temporal coverage of algal biomass concentrations, allowing for longer time series observational studies to be conducted over large areas. Using data provided by the European Space Agency (ESA) Climate Change Initiative (CCI) Lakes project (product version 2.0.0), daily chlorophyll-a (chl-a; proxy of algal biomass), Lake Surface Water Temperature (LSWT) and Lake Ice Cover (LIC) from 2002 to 2020 were derived from five North American Great Lakes: Great Bear Lake (GBL), Great Slave Lake (GSL), Lake Athabasca (LA), Lake Winnipeg (LW), and Lake Erie (LE). Additional atmospheric and lake physical variables were provided by the European Centre for Medium-Range Weather Forecasts (ECMWF) ERA5-Land data as part of the ERA5 climate reanalysis product including: 2-m air temperature (T2m), Total Precipitation (PPT), Surface Net Solar Radiation (SNSR), Surface Runoff (SR) and Subsurface Runoff (SSR), Wind Speed (WS) and Lake Mix-Layer Depth (LMLD). Such data products allow for comprehensive time series analysis on the interaction effects of atmospheric and lake physical parameters on algal biomass dynamics. Winter temperatures exhibit the highest rate of change relative to other seasons, where LIC loss is important for Northern hemisphere lakes; however, its effect on algal biomass dynamics is relatively unknown. To investigate how LIC duration alters algal biomass in North American Great Lakes, annual and seasonal algal biomass, LSWT and LIC parameters were calculated for the five study lakes using ESA CCI Lakes data. Algal biomasses (β = 0.01 – 0.75 μg L-1 yr-1) and LSWT (β = 0.03 – 0.14 K yr-1) were found to increase, with a general decrease in LIC (β = -0.88 – -1.08 Days yr-1) from 2002 to 2020. Vector autoregressions (VARs) showed that in Northern Lakes (NL; GBL, GSL and LA), LSWT and LIC parameters provide greater explanatory power for annual/seasonal chl-a concentrations (median adj. r2 = 0.75) compared to Southern Lakes (median adj. r2 = 0.46). Additionally, LIC parameters were found to provide higher explanatory power for NLs during the spring season compared to LSWT. However, higher explanatory power does not indicate predictive capacity, where machine learning methods may provide stronger predictive models. To determine if LIC may act as a predictor of algal biomass parameters, multiple linear regression (MLR) and artificial neural networks (ANN) were constructed using per-pixel observations of annual/seasonal algal biomass, LSWT, and LIC parameters. Irrespective of season, LSWT only models returned lower prediction error (median NRMSE = 0.82) compared to LIC only models (median NRMSE = 0.93). However, models consisting of both LIC and LSWT returned the lowest predictive error (median NRMSE = 0.75). While LIC did not act as a strong predictor of algal biomass, a random forest (RF) classifier was used to determine whether LIC could classify the presence of lake-specific anomalies in chl-a concentrations. The RF model found that LIC parameters (ice on/off) had the highest mean accuracy decrease on average for NLs during the spring season. LIC timings are changing, where it was found to have greater importance on springtime abnormal algal biomass growth in NLs. While LIC was important at this time compared to LSWT, the impact of other important atmospheric and lake physical variables on algal biomass dynamics are not well understood, particularly at a smaller temporal scale (i.e., daily). To assess the potential interaction effects between algal biomass, atmospheric, and lake physical parameters, a network analysis was conducted using a High Order Dynamic Gaussian Bayesian Network (HO-DGBN) for the original time series, the stationary, non-stationary, and residual signals at varying temporal ranges (Δ: daily, three days, weekly, biweekly, and monthly averages). It was found that LSWT, T2m and SNSR were the most important parameters on average, where LSWT exhibited the highest importance on the daily scale compared to the monthly. Additionally, LMLD returned increased importance at longer temporal frequencies, while SSR returned increased importance at shorter temporal frequencies. Temperature interactions were mixed, typically returning both positive and negative interactions, while SNSR typically exhibited a positive interaction with chl-a, while LMLD exhibited a frequent negative interaction. PPT and WS were found to be the least important parameters in all study lakes. This thesis provides some of the first analytical uses of the ESA CCI Lakes product; a product that undergoes regular updates (every two years or so) as new satellite and in situ data become available, and algorithms for the retrieval of chl-a, LSWT and LIC are being improved. As such, improvements are expected in future releases of the product, limiting the accuracy of some findings in the thesis. Of the data presented, there is evidence that LIC is a significant contributor to spring algal biomass dynamics for NLs; however, Southern Lakes (SL; LW and LE) exhibit more complex interactions, likely due to anthropogenic impacts. This thesis identifies the complexity of LSWT interactions with algal biomass and identifies LMLD as a predominantly negative effect in the development of algal biomass. Algal biomasses are increasing, where increases in LSWT yield higher algal biomass peaks (at varying times throughout the year) within the study lakes. Future climate scenarios may provide conditions favorable for algal biomass growth, where Northern landscapes are at the greatest risk.

Global gridded precipitation products have proven essential for many applications ranging from hydrological modeling and climate model validation to natural hazard risk assessment. They provide a global picture of how precipitation varies across time and space, specifically in regions where ground-based observations are scarce. While the application of global precipitation products has become widespread, there is limited knowledge on how well these products represent the magnitude and frequency of extreme precipitation—the key features in triggering flood hazards. Here, five global precipitation datasets (MSWEP, CFSR, CPC, PERSIANN-CDR, and WFDEI) are compared to each other and to surface observations. The spatial variability of relatively high precipitation events (tail heaviness) and the resulting discrepancy among datasets in the predicted precipitation return levels were evaluated for the time period 1979–2017. The analysis shows that 1) these products do not provide a consistent representation of the behavior of extremes as quantified by the tail heaviness, 2) there is strong spatial variability in the tail index, 3) the spatial patterns of the tail heaviness generally match the Köppen–Geiger climate classification, and 4) the predicted return levels for 100 and 1000 years differ significantly among the gridded products. More generally, our findings reveal shortcomings of global precipitation products in representing extremes and highlight that there is no single global product that performs best for all regions and climates.

The topic of satellite remote sensing of lake ice has gained considerable attention in recent years. Optical satellite data from the Moderate Resolution Imaging Spectroradiometer (MODIS) allow for the monitoring of lake ice cover (an Essential Climate Variable or ECV), and dates associated with ice phenology (freeze-up, break-up, and ice cover duration) over large areas in an era where ground-based observational networks have nearly vanished in many northern countries. Ice phenology dates as well as dates of maximum and minimum ice cover extent (for lakes that do not form a complete ice cover in winter or do not totally lose their ice cover in summer) are useful for assessing long-term trends and variability in climate, particularly due to their sensitivity to changes in near-surface air temperature. Existing knowledge-driven (threshold-based) retrieval algorithms for lake ice cover mapping that use top-of-atmosphere (TOA) reflectance products do not perform well under lower solar illumination conditions (i.e. large solar zenith angles), resulting in low TOA reflectance. This research assessed the capability of four machine learning classifiers (i.e. multinomial logistic regression, MLR; support vector machine, SVM; random forest, RF; gradient boosting trees, GBT) for mapping lake ice cover, water and cloud cover during both break-up and freeze-up periods using the MODIS/Terra L1B TOA (MOD02) product. The classifiers were trained and validated using samples collected from 17 large lakes across the Northern Hemisphere (Europe and North America); lakes that represent different characteristics with regards to area, latitude, freezing frequency, and ice duration. Following an accuracy assessment using random k-fold cross-validation (k = 100), all machine learning classifiers using a 7-band combination (visible, near-infrared and shortwave-infrared) were found to be able to produce overall classification accuracies above 94%. Both RF and GBT provided overall and class-specific accuracies above 98% and a more visually accurate depiction of lake ice, water and cloud cover. The two tree-based classifiers offered the most robust spatial transferability over the 17 lakes and performed consistently well across ice seasons. However, only RF was relatively insensitive to the choice of the hyperparameters compared to the other three classifiers. The results demonstrate the potential of RF for mapping lake ice cover globally from MODIS TOA reflectance data.

Assessment of machine learning classifiers for global lake ice cover mapping from MODIS TOA reflectance data

Yirdaw, S.Z., Snelgrove, K.R., Seglenieks, F.R., Agboma, C.O. and Soulis, E.D. (2009). Assessment of the WATCLASS Hydrological Model Result of the Mackenzie River Basin Using the GRACE Satellite Total Water Storage Measurement. Hydrologic Processes, https://doi.org/10.1002/hyp.7450.

Assessment of the WATCLASS Hydrological Model Result of the Mackenzie River Basin Using the GRACE Satellite Total Water Storage Measurement

Yirdaw, S.Z., K.R. Snelgrove, F.R. Seglenieks, C.O. Agboma, and E.D. Soulis, 2009: Assessment of the WATCLASS hydrological model result of the MacKenzie River Basin using the result of the GRACE satellite total water storage measurement. Hydrol. Processes, 23, 3391-3400.

Assessment of the WATCLASS hydrological model result of the MacKenzie River Basin using the result of the GRACE satellite total water storage measurement

When a programmer makes changes to a target program entity (files, classes, methods), it is important to identify which other entities might also get impacted. These entities constitute the impact set for the target entity. Association rules have been widely used for discovering the impact sets. However, such rules only depend on the previous co-change history of the program entities ignoring the fact that similar entities might often need to be updated together consistently even if they did not co-change before. Considering this fact, we investigate whether cloning relationships among program entities can be associated with association rules to help us better identify the impact sets. In our research, we particularly investigate whether the impact set detection capability of a clone detector can be utilized to enhance the capability of the state-of-the-art association rule mining technique, Tarmaq, in discovering impact sets. We use the well known clone detector called NiCad in our investigation and consider both regular and micro-clones. Our evolutionary analysis on thousands of commit operations of eight diverse subject systems reveals that consideration of code clones can enhance the impact set detection accuracy of Tarmaq with a significantly higher precision and recall. Micro-clones of 3LOC and 4LOC and regular code clones of 5LOC to 20LOC contribute the most towards enhancing the detection accuracy.

Simulation-optimization techniques in support of groundwater management are computationally expensive. To tackle such computational burden, a variety of surrogate modeling-frameworks have been proposed, where a cheaper-to-run model referred to as a surrogate is used in lieu of a computationally intensive model. These frameworks are generally based on what referred herein to as ‘global surrogate modelling’ where a single surrogate approximates the underlying response surface of a model. Such classic frameworks, however, are sub-optimal when the response surface is complex and/or high-dimensional. This paper proposes a novel ‘local surrogate modelling’ framework that simultaneously builds and evolves multiple local surrogates, guided by an automatic clustering method. Unlike traditional clustering methods that select the number of clusters a priori, the proposed automatic clustering method concurrently determines the optimum number of clusters and the clustering scheme itself. To serve as the surrogate, Artificial Neural Networks (ANNs) are used. The proposed framework is applied to solve a computationally intensive groundwater remediation optimization problem. This study shows that the proposed automatic clustering-based local surrogate modeling is effective and reliable while reducing at least 60 percent of the computational burden.

Release notes are admitted as an essential technical document in software maintenance. They summarize the main changes, e.g. bug fixes and new features, that have happened in the software since the previous release. Manually producing release notes is a time-consuming and challenging task. For that reason, sometimes developers neglect to write release notes. For example, we collect data from GitHub with over 1900 releases, and among them, 37% of the release notes are empty. To mitigate this problem, we propose an automatic release notes generation approach by applying the text summarization techniques, i.e. TextRank. To improve the keyword extraction method of traditional TextRank, we integrate the GloVe word embedding technique with TextRank. After generating release notes automatically, we apply machine learning algorithms to classify the release note contents (or sentences). We classify the contents into six categories, e.g. bug fixes and performance improvements, to represent the release notes better for users. We use the evaluation metric, e.g. ROUGE, to evaluate the automatically generated release notes. We also compare the performance of our technique with two popular extractive algorithms, e.g. Luhn’s and latent semantic analysis (LSA). Our evaluation results show that the improved TextRank method outperforms the two algorithms.

Automatically generating release notes with content classification models

Beaver (Castor canadensis and Castor fiber) are regarded widely as ecosystem engineers and the dams they create are well-known for their ability to drastically alter the hydrology of rivers. As a result, beaver are increasingly being included in green infrastructure practices to combat the effects of climate change and enhance ecosystem resilience. Both drought and flood mitigation capabilities have been observed in watersheds with beaver dam structures; however, how dams possess contrasting mitigation abilities is not fully understood since most studies neglect to acknowledge variation in beaver dam structures. In this study, an extensive cross-site survey of the physical and hydrologic properties of beaver dams was conducted in the Canadian Rocky Mountains in Alberta. This research aimed to improve the understanding of the hydrology of beaver dams by categorizing dams using their intrinsic properties and landscape settings to identify fundamental patterns that may be applicable across landscape types. The dam flow type classification from Woo and Waddington (1990) was evaluated in this new context and adapted to include two new flow types. The survey of intrinsic beaver dam properties revealed significant differences in dam structure across different sites. Physical differences in dam structure altered the dynamics and variance of pond storage and certain dam attributes related to the landscape setting. For instance, dam material influenced dam height and water source influenced dam length. However, a closer analysis of large rain events showed that the physical structure of dams alters seasonal dynamics of pond storage but not the response to rain events. Overall, this research shows that beaver dams can be both structurally and hydrologically very different from each other. Establishing broadly applicable classifications is vital to understanding the ecosystem resilience and mitigation services beaver dams provide.

In Arctic-boreal regions, climate change induced permafrost thaw causes complex interactions between topography, microbiology, hydrology and vegetation distribution that impact the emission rate of radiatively important gasses like methane (CH4). Vegetation distribution, for example, can be used as a proxy for permafrost thaw and the greenhouse gas emissions associated with each stage. To scale emissions up from the site level, more accurate remotely sensed data is needed to produce the necessary vegetation cover maps. Analyses of data collected from unpiloted aerial systems (UAS) have proven effective at classifying vegetation distribution into land cover classes that often include multiple species. However, focusing classification at the level of species functional group (SFG) could further improve scaling efforts by representing vegetation distribution, and thus permafrost thaw impacts and gas emissions, more accurately. High resolution, UAS-derived, hyperspectral imagery was collected in tandem with vegetation surveys, porewater CH4 concentration and isotopes as well as chamber flux measurements to predict six SFG at two boreal peatland sites: Lutose, Alberta (59.5°N 117.2°W) and Smith Creek, NWT (63.2°N 123.3°W). Using a Random Forest Model, we classified the imagery into predetermined land cover classes representative of a permafrost thaw gradient (Permafrost Peat Plateau, Plateau Edge, Young Bog, Intermediate Bog, Mature/Old Bog, and Fen). Each class was spectrally distinct from one another. We then used an Artificial Neural Network to predict the distribution of six SFGs, chosen based on their association with higher or lower CH4 emissions (Carex/Sedge, Scheuchzeria spp., Sphagnum spp., Black Spruce, Lichens, and Cloudberry). Five spectral indices, indicative of structural or chemical properties of vegetation, were used to predict each SFG. We determined which two indices best predicted the distribution of each functional group using a jackknife validation approach. We found we could confidently predict five of the six SFGs, and could produce distribution maps of each SFG in both sites. These data could then be used to parameterize ecosystem models used to predict and then scale CH4 emissions at the site and regional scale

Burke, S.A., Palace, M.W., Sullivan, F., et al. (2022) Beyond what the eye can see: using high-resolution UAS-derived hyperspectral imagery to classify boreal peatland vegetation. American Geophysical Meeting, Chicago, United States of America, (December 12-16). https://agu.confex.com/agu/fm22/meetingapp.cgi/Paper/1148519

Beyond what the eye can see: using high-resolution UAS-derived hyperspectral imagery to classify boreal peatland vegetation

Flow management has the potential to significantly affect ecosystem condition. Shallow lakes in arid regions are especially susceptible to flow management changes, which can have important implications for the formation of cyanobacterial blooms. Here, we reveal water quality shifts associated with changing source water inflow management. Using in situ monitoring data, we studied a seven-year time span during which inflows to a shallow, eutrophic drinking water reservoir transitioned from primarily natural landscape runoff (2014–2015) to managed flows from a larger upstream reservoir (Lake Diefenbaker; 2016–2020) and identified significant changes in cyanobacteria (as phycocyanin) using generalized additive models to classify cyanobacterial bloom formation. We then connected changes in water source with shifts in chemistry and the occurrence of cyanobacterial blooms using principal components analysis. Phycocyanin was greater in years with managed reservoir inflow from a mesotrophic upstream reservoir (2016–2020), but dissolved organic matter (DOM) and specific conductivity, important determinants of drinking water quality, were greatest in years when landscape runoff dominated lake water source (2014–2015). Most notably, despite changing rapidly, it took multiple years for lake water to return to a consistent and reduced level of DOM after managed inflows from the upstream reservoir were resumed, an observation that underscores how resilience may be hindered by weak resistance to change and slow recovery. Environmental flows for water quality are rarely defined, yet we show that trade-offs exist between poor water quality via elevated conductivity and DOM and higher bloom risk, depending on water source. Our work highlights the importance of source water quality, not just quantity, to water security, and our findings have important implications for water managers who must protect ecosystem services while adapting to projected hydroclimatic change.

The Global Water Futures program (GWF) was granted $77.8 million by the Canada First Research Excellence Fund to conduct research on the forecasting and management of water futures in Canada as part of an effort to combat projected risks associated with global climate change. The production of scientific knowledge is a clear objective of the GWF program, and the evaluation and enumeration of research outcomes is a key metric. The goal of this work is to create a comprehensive bibliographic analysis of research outputs across the full extent of the GWF program including metadata such as the title, author, publication date and geographic locations of the works. Processes to incorporate quality control, classification and validation were documented to ensure these outputs are effectively managed, monitored and evaluated. Links to the resources are also made available to ensure they are easily accessible to a wide range of audiences. Enhanced accessibility is key in sharing critical climate change research and expanding international understanding of climate-water issues. The review and evaluation of existing procedures for research output reporting provided insight to propose improved processes to increase efficiency and accuracy. By establishing a consistent organizational tool for all forms of research outputs, opportunities are created for future widespread information sharing and global collaboration. Analysis of annual reports generated by 65 projects across for main partner universities (University of Saskatchewan, University of Waterloo, McMaster University, and Wilfrid Laurier University) listed 4,708 total outputs between 2017 and 2022. Conference presentations, refereed publications, and data publications constitute most of this total, accounting for 48%, 26%, and 10% of overall outputs, respectively. Other products, including non-refereed articles, model code, and book chapters, each make up less than 3% of total GWF reported outputs. By looking at the distribution of each output type annually, we were able to examine the impact of global circumstances such as COVID-19 on total output production over a significant period. Understanding the effects of large-scale events on project development and management are critical for future adaptation in water research that will enhance research opportunities and their subsequent data findings.

The design and maintenance of APIs (Application Programming Interfaces) are complex tasks due to the constantly changing requirements of their users. Despite the efforts of their designers, APIs may suffer from a number of issues (such as incomplete or erroneous documentation, poor performance, and backward incompatibility). To maintain a healthy client base, API designers must learn these issues to fix them. Question answering sites, such as Stack Overflow (SO), have become a popular place for discussing API issues. These posts about API issues are invaluable to API designers, not only because they can help to learn more about the problem but also because they can facilitate learning the requirements of API users. However, the unstructured nature of posts and the abundance of non-issue posts make the task of detecting SO posts concerning API issues difficult and challenging. In this paper, we first develop a supervised learning approach using a Conditional Random Field (CRF), a statistical modeling method, to identify API issue-related sentences. We use the above information together with different features collected from posts, the experience of users, readability metrics and centrality measures of collaboration network to build a technique, called CAPS, that can classify SO posts concerning API issues. In total, we consider 34 features along eight different dimensions. Evaluation of CAPS using carefully curated SO posts on three popular API types reveals that the technique outperforms all three baseline approaches we consider in this study. We then conduct studies to find important features and also evaluate the performance of the CRF-based technique for classifying issue sentences. Comparison with two other baseline approaches shows that the technique has high potential. We also test the generalizability of CAPS results, evaluate the effectiveness of different classifiers, and identify the impact of different feature sets.

CAPS: a supervised technique for classifying Stack Overflow posts concerning API issues

Permafrost thaw/degradation in the Northern Hemisphere due to global warming is projected to accelerate in coming decades. Assessment of this trend requires improved understanding of the evolution and dynamics of permafrost areas. Land surface models (LSMs) are well-suited for this due to their physical basis and large-scale applicability. However, LSM application is challenging because (a) LSMs demand extensive and accurate meteorological forcing data, which are not readily available for historic conditions and only available with significant biases for future climate, (b) LSMs possess a large number of model parameters, and (c) observations of thermal/hydraulic regimes to constrain those parameters are severely limited. This study addresses these challenges by applying the MESH-CLASS modeling framework (Modélisation Environmenntale communautaire—Surface et Hydrology embedding the Canadian Land Surface Scheme) to three regions within the Mackenzie River Basin, Canada, under various meteorological forcing data sets, using the variogram analysis of response surfaces framework for sensitivity analysis and threshold-based identifiability analysis. The study shows that the modeler may face complex trade-offs when choosing a forcing data set; for current and future scenarios, forcing data require multi-variate bias correction, and some data sets enable the representation of some aspects of permafrost dynamics, but are inadequate for others. The results identify the most influential model parameters and show that permafrost simulation is most sensitive to parameters controlling surface insulation and runoff generation. But the identifiability analysis reveals that many of the most influential parameters are unidentifiable. These conclusions can inform future efforts for data collection and model parameterization.

Permafrost plays an important role in the hydrology of arctic/subarctic regions. However, permafrost thaw/degradation has been observed over recent decades in the Northern Hemisphere and is projected to accelerate. Hence, understanding the evolution of permafrost areas is urgently needed. Land surface models (LSMs) are well-suited for predicting permafrost dynamics due to their physical basis and large-scale applicability. However, LSM application is challenging because of the large number of model parameters and the complex memory of state variables. Significant interactions among the underlying processes and the paucity of observations of thermal/hydraulic regimes add further difficulty. This study addresses the challenges of LSM application by evaluating the uncertainty due to meteorological forcing, assessing the sensitivity of simulated permafrost dynamics to LSM parameters, and highlighting issues of parameter identifiability. Modelling experiments are implemented using the MESH-CLASS framework. The VARS sensitivity analysis and traditional threshold-based identifiability analysis are used to assess various aspects of permafrost dynamics for three regions within the Mackenzie River Basin. The study shows that the modeller may face significant trade-offs when choosing a forcing dataset as some datasets enable the representation of some aspects of permafrost dynamics, while being inadequate for others. The results also emphasize the high sensitivity of various aspects of permafrost simulation to parameters controlling surface insulation and soil texture; a detailed list of influential parameters is presented. Identifiability analysis reveals that many of the most influential parameters for permafrost simulation are unidentifiable. These conclusions will hopefully inform future efforts in data collection and model parametrization.

Human-induced disturbances of terrestrial and aquatic ecosystems continue at alarming rates. With the advent of both raw sensor and analysis-ready datasets, the need to monitor ecosystem disturbances is now more imperative than ever; yet the task is becoming increasingly complex with increasing sources and varieties of earth observation data. In this research, computer vision methods and tools are interrogated to understand their capability for comparing spatial patterns. A critical survey of literature provides evidence that computer vision methods are relatively robust to scale and highlights issues involved in parameterization of computer vision models for characterizing significant pattern information in a geographic context. Utilizing two widely used pattern indices to compare spatial patterns in simulated and real-world datasets revealed their potential to detect subtle changes in spatial patterns which would not otherwise be feasible using traditional pixel-level techniques. A texture-based CNN model was developed to extract spatially relevant information for landscape similarity comparison; the CNN feature maps proved to be effective in distinguishing agriculture landscapes from other landscape types (e.g., forest and mountainous landscapes). For real-world human disturbance monitoring, a U-Net CNN was developed and compared with a random forest model. Both modeling frameworks exhibit promising potential to map placer mining disturbance; however, random forests proved simple to train and deploy for placer mapping, while the U-Net may be used to augment RF as it is capable of reducing misclassification errors and will benefit from increasing availability of detailed training data.

Projections of change in high-flow extremes with global warming vary widely among, and within, large midlatitude river basins. The spatial variability of these changes is attributable to multiple causes. One possible and little-studied cause of changes in high-flow extremes is a change in the synchrony of mainstem and tributary streamflow during high-flow extremes at the mainstem-tributary confluence. We examined reconstructed and simulated naturalized daily streamflow at confluences on the Columbia River in western North America, quantifying changes in synchrony in future streamflow projections and estimating the impact of these changes on high-flow extremes. In the Columbia River basin, projected flow regimes across colder tributaries initially diverge with warming as they respond to climate change at different rates, leading to a general decrease in synchrony, and lower high-flow extremes, relative to a scenario with no changes in synchrony. Where future warming is sufficiently large to cause most subbasins upstream from a confluence to transition toward a rain-dominated, warm regime, the decreasing trend in synchrony reverses itself. At one confluence with a major tributary (the Willamette River), where the mainstem and tributary flow regimes are initially very different, warming increases synchrony and, therefore, high-flow magnitudes. These results may be generalizable to the class of large rivers with large contributions to flood risk from the snow (i.e., cold) regime, but that also receive considerable discharge from tributaries that drain warmer basins.

The microbiome has been described as an additional host “organ” with well-established beneficial roles. In addition to aiding in digestion of food and uptake of nutrients, microbiota in guts of vertebrates are responsible for regulating several beneficial functions, including stimulating immune responses and maintaining homeostasis. However, effects of exposures to chemicals on both structure and function of the gut microbiome of fishes are understudied. The overall purpose of research reported in this thesis was to characterize the effects of polycyclic aromatic hydrocarbons (PAHs) on the gut microbiomes of freshwater fishes, using both laboratory- and field-based assessments. PAHs have a number of well-characterized deleterious impacts in fish and are known modulators of the aryl hydrocarbon receptor, a receptor that has a bidirectional relationship with the microbiome. As such, this chemical class was selected to investigate targeted effects on the microbiome. The objectives of this thesis were to: (1) Determine if aqueous exposures to BaP modulate the gut microbiome in fathead minnows (Pimephales promelas) and to discern whether these community shifts were sex-dependent; (2) Assess whether a dietary exposure to BaP also affects genomic and active microbiomes in juvenile fathead minnows; and (3) Evaluate the gut microbiomes in native fishes following exposure to an oil spill of heavy crude on the North Saskatchewan River. To accomplish this, adult male and female fathead minnows were aqueously exposed to a short-term environmentally-relevant low concentrations of benzo[a]pyrene (BaP) and composition of the gut microbiome were assessed with 16S rRNA metagenetics. Following this, juvenile fathead minnows were exposed via the diet to environmentally-relevant higher concentrations of BaP for two weeks, and composition of both the genomic and active gut microbiome were assessed with DNA- and RNA-based 16S rRNA metagenetics, respectively. Lastly, fishes from the North Saskatchewan River were collected a year after an oil spill, and differences in the microbiome based on fish species and measured PAH muscle concentrations were assessed with 16S rRNA metagenetics. Studies presented in this thesis revealed a clear influence of PAHs on the microbiome, even with relatively small exposures, while exposures to higher concentrations of PAHs resulted in microbial communities of the gut that were distinctly altered, with lost community structure as determined by co-occurrence networks. Notably, microbiomes in guts of male and female fish were affected differently by exposure to BaP, suggesting sex specific responses to the chemical. Moreover, certain bacterial taxa were correlated with exposure to BaP, many of which were associated with hydrocarbon degradation and disease. Predicted functional analyses revealed several pathways that were correlated with BaP exposure, including increases in aromatic degradation pathways. Many of the conclusions were similar when analyzing the genomic and active microbiomes, but the active and DNA-normalized (the RNA/DNA ratio of bacterial abundances) active microbiomes provided greater resolution of the effects of BaP exposure. Finally, with the field study, it was determined that among goldeye (Hiodon alosoides), walleye (Sander vitreus), northern pike (Esox lucius), and shorthead redhorse (Moxostoma macrolepidotum), host species drive the assemblages of gut microbiomes. Additionally, several taxa associated with hydrocarbon degradation, inflammation, and disease were correlated with PAH concentrations in muscle tissue across the field-collected fishes, and walleye community composition was correlated with concentrations of PAHs in muscle tissue. Across all studies, PAH exposure can be a driver of bacterial community composition, and Desulfovibrionaceae, Shewanellaceae and Chitinophagaceae were positively correlated with PAH exposure in both laboratory-exposed fathead minnows and wild-caught fishes. Overall, this thesis provides novel data and new understanding into the effects of PAHs on the gut microbiomes of freshwater fish that can ultimately be used to better understand the connection between a toxicant and an adverse outcome via the microbiome.

The microbiome has been described as an additional host “organ” with well-established beneficial roles. In addition to aiding in digestion of food and uptake of nutrients, microbiota in guts of vertebrates are responsible for regulating several beneficial functions, including stimulating immune responses and maintaining homeostasis. However, effects of exposures to chemicals on both structure and function of the gut microbiome of fishes are understudied. The overall purpose of research reported in this thesis was to characterize the effects of polycyclic aromatic hydrocarbons (PAHs) on the gut microbiomes of freshwater fishes, using both laboratory- and field-based assessments. PAHs have a number of well-characterized deleterious impacts in fish and are known modulators of the aryl hydrocarbon receptor, a receptor that has a bidirectional relationship with the microbiome. As such, this chemical class was selected to investigate targeted effects on the microbiome. The objectives of this thesis were to: (1) Determine if aqueous exposures to BaP modulate the gut microbiome in fathead minnows (Pimephales promelas) and to discern whether these community shifts were sex-dependent; (2) Assess whether a dietary exposure to BaP also affects genomic and active microbiomes in juvenile fathead minnows; and (3) Evaluate the gut microbiomes in native fishes following exposure to an oil spill of heavy crude on the North Saskatchewan River. To accomplish this, adult male and female fathead minnows were aqueously exposed to a short-term environmentally-relevant low concentrations of benzo[a]pyrene (BaP) and composition of the gut microbiome were assessed with 16S rRNA metagenetics. Following this, juvenile fathead minnows were exposed via the diet to environmentally-relevant higher concentrations of BaP for two weeks, and composition of both the genomic and active gut microbiome were assessed with DNA- and RNA-based 16S rRNA metagenetics, respectively. Lastly, fishes from the North Saskatchewan River were collected a year after an oil spill, and differences in the microbiome based on fish species and measured PAH muscle concentrations were assessed with 16S rRNA metagenetics. Studies presented in this thesis revealed a clear influence of PAHs on the microbiome, even with relatively small exposures, while exposures to higher concentrations of PAHs resulted in microbial communities of the gut that were distinctly altered, with lost community structure as determined by co-occurrence networks. Notably, microbiomes in guts of male and female fish were affected differently by exposure to BaP, suggesting sex specific responses to the chemical. Moreover, certain bacterial taxa were correlated with exposure to BaP, many of which were associated with hydrocarbon degradation and disease. Predicted functional analyses revealed several pathways that were correlated with BaP exposure, including increases in aromatic degradation pathways. Many of the conclusions were similar when analyzing the genomic and active microbiomes, but the active and DNA-normalized (the RNA/DNA ratio of bacterial abundances) active microbiomes provided greater resolution of the effects of BaP exposure. Finally, with the field study, it was determined that among goldeye (Hiodon alosoides), walleye (Sander vitreus), northern pike (Esox lucius), and shorthead redhorse (Moxostoma macrolepidotum), host species drive the assemblages of gut microbiomes. Additionally, several taxa associated with hydrocarbon degradation, inflammation, and disease were correlated with PAH concentrations in muscle tissue across the field-collected fishes, and walleye community composition was correlated with concentrations of PAHs in muscle tissue. Across all studies, PAH exposure can be a driver of bacterial community composition, and Desulfovibrionaceae, Shewanellaceae and Chitinophagaceae were positively correlated with PAH exposure in both laboratory-exposed fathead minnows and wild-caught fishes. Overall, this thesis provides novel data and new understanding into the effects of PAHs on the gut microbiomes of freshwater fish that can ultimately be used to better understand the connection between a toxicant and an adverse outcome via the microbiome.

Characterizing alpine basin land cover classes and their influence on water storage and routing in the Mackenzie Mountains, Kaska & Shúhtaot'ine shared territory

Vegetation growth and productivity in Canada's boreal are governed by a characteristically short growing season, which is largely driven by the Freeze/Thaw(F/T) cycles that constrain the supply of water and nutrients through seasonally frozen soils. Since much of the vegetation in the Canadian boreal consists of evergreen species which do not experience large seasonal cycles in photosynthetic biomass, monitoring this growing season through the use of visible and near-infrared wavelengths via spectral indices such as the Normalized Difference Vegetation Index (NDVI) has proven difficult. To adequately capture growing season constraints in these northern environments, microwave remote sensing offers potential. L-band passive microwave observations are sensitive to near-surface soil moisture conditions and can monitor F/T states effectively due to the high contrast in permittivity between frozen and thawed soils. We characterize F/T information using products from both the Soil Moisture Ocean Salinity (SMOS) and Soil Moisture Active Passive (SMAP) missions. Daily F/T retrievals are classified into distinct growing season phases based on a 7-day moving window approach and are used to generate a suite of 12 temporal metrics over the 2017–2019 period including growing season length, timing of the fall freeze transition, ephemeral F/T events, and the total number of state transitions. Several key metrics are also generated from in-situ soil temperature datasets for comparison with the SMOS and SMAP datasets. Uncorrelated F/T metrics were then leveraged to delineate unique regions of F/T-derived growing season characteristics using a K-means clustering approach. Regions derived from SMOS and SMAP F/T retrievals were assessed for their ability to capture unique spatial constraints on vegetation productivity with reference to modelled Gross Primary Productivity (GPP) obtained from SMAP and a MODIS/Fluxnet synergy product. Our results indicate that both SMOS and SMAP-derived F/T metrics correspond with unique spatial patterns in vegetation productivity, illustrating the F/T cycle constraints on the seasonal availability of soil moisture, nutrients and suitable soil temperatures required for vegetation productivity across the Canadian boreal. In addition, the relationship between the SMAP and SMOS F/T-derived growing season length metrics and reference GPP yielded rates of change at 5.30 and 5.64 gC m?2 yr?1 per 1-day increase in growing season length. These estimated rates of change are similar to those identified by studies using complex process-based ecosystem models and in-situ eddy covariance data from flux towers. These similarities highlight the potential of this simple and robust remotely sensed approach for capturing climatic drivers of land cover and vegetation productivity not currently represented in common Canadian ecological regions.

Qu, Y., Scott, K.A. (2024) Classification of Ice and Water In a Regulated River Using Convolutional Neural Networks, Geoinformatics 2024

Classification of Ice and Water In a Regulated River Using Convolutional Neural Networks

Tang, W., and Carey, S.K. 2022.?Classifying Annual Daily Hydrographs in Western North America using t-Distributed Stochastic Neighbor Embedding (t-SNE). Hydrological Processes. https://doi.org/10.1002/hyp.14473.

Classifying Annual Daily Hydrographs in Western North America using t-Distributed Stochastic Neighbor Embedding (t-SNE

Classifying terrain in a muskeg-wetland regime for application to a GRU type distributed hydrologic model

A code clone is a pair of similar code fragments, within or between software systems. To detect each possible clone pair from a software system while handling the complex code structures, the clone detection tools undergo a lot of generalization of the original source codes. The generalization often results in returning code fragments that are only coincidentally similar and not considered clones by users, and hence requires manual validation of the reported possible clones by users which is often both time-consuming and challenging. In this paper, we propose a machine learning based tool 'CloneCognition' (Open Source Codes: https://github.com/pseudoPixels/CloneCognition ; Video Demonstration: https://www.youtube.com/watch?v=KYQjmdr8rsw) to automate the laborious manual validation process. The tool runs on top of any code clone detection tools to facilitate the clone validation process. The tool shows promising clone classification performance with an accuracy of up to 87.4%. The tool also exhibits significant improvement in the results when compared with state-of-the-art techniques for code clone validation.

Bellisario, L., L.D. Boudreau, D.L. Verseghy, W.R. Rouse, and P.D. Blanken, (1999). Comparing the performance of the Canadian Land Surface Scheme (CLASS) for two subarctic terrain types. Atmos-Ocean, 38, 181-204.

Comparing the performance of the Canadian Land Surface Scheme (CLASS) for two subarctic terrain types

Computationally efficient calibration of WATCLASS Hydrologic models using surrogate optimization

Shantz, S., Strickert, G. (2023) Confluences of Water, Art and Science. Invited Talk to Class at Western University Feb 14th, 1:30 - 3:00 PM

Invited Talk to Class at Western University Feb 14th, 1:30 - 3:00 PM

Coupled MESH-CLASSIC model for catchment-scale water and carbon cycle studies

The North American beaver (Castor canadensis) is regarded widely as an ecosystem engineer and the dams they create are well known for their ability to drastically alter the hydrology of river basins. As a result, beavers are increasingly being included in green infrastructure practices to combat the effects of climate change and enhance ecosystem resilience. Both drought and flood mitigation capabilities have been observed in watersheds with beaver dam structures; however, how dams can possess such contrasting mitigation abilities is not fully understood as most studies neglect to acknowledge the incredible variation in beaver dam structures. In this study, an extensive cross-site survey of dam properties and water flow through dams in the Canadian Rocky Mountains in Alberta was conducted. The dam flow type classification from Woo and Waddington (1990) was evaluated in this new context and adapted to include two new flow types not found in the original study. The survey revealed significant differences in dam structure across the different sites. Physical differences in dam structure altered the dynamics and variance in the ponds’ storage and certain dam attributes related to landscape setting. However, a closer analysis of large rain events surprisingly showed little influence of dam physical structure on storage and recession limbs. These results reveal that variation in dam structure alters the temporal dynamics of pond storage but also emphasizes that some metrics, including response to rain events, may remain constant regardless of dam structure. This combination of variability and stability may be the secret to the contrasting mitigation abilities possessed by beaver dams. Further research may be able to use these results to better predict streamflow routing through beaver ponds and improve the prediction and modelling of beaver dam effects.

Various functional DNA molecules have been used for the detection of environmental contaminants in water, but their practical applications have been limited. To address this gap, this review highlights the efforts to develop field-deployable water quality biosensors. The biosensor devices include microfluidic, lateral flow and paper-based devices, and other novel ideas such as the conversion of glucometers for the detection of environmental analytes. In addition, we also review DNA-functionalized hydrogels and their use in diffusive gradients in thin films (DGT) devices. We classify the sensors into one-step and two-step assays and discuss their practical implications. While the review is focused on works reported in the last five years, some classic early works are cited as well. Overall, most of the existing work only tested spiked water samples. Future work needs to shift to real environmental samples and the comparison of DNA-based sensors to standard analytical methods.

The Köppen-Geiger (KG) climate classification has been widely used to determine the climate at global and regional scales using precipitation and temperature data. KG maps are typically developed using a single product; however, uncertainties in KG climate types resulting from different precipitation and temperature datasets have not been explored in detail. Here, we assess seven global datasets to show uncertainties in KG classification from 1980 to 2017. Using a pairwise comparison at global and zonal scales, we quantify the similarity among the seven KG maps. Gauge- and reanalysis-based KG maps have a notable difference. Spatially, the highest and lowest similarity is observed for the North and South Temperate zones, respectively. Notably, 17% of grids among the seven maps show variations even in the major KG climate types, while 35% of grids are described by more than one KG climate subtype. Strong uncertainty is observed in south Asia, central and south Africa, western America, and northeastern Australia. We created two KG master maps (0.5° resolution) by merging the climate maps directly and by combining the precipitation and temperature data from the seven datasets. These master maps are more robust than the individual ones showing coherent spatial patterns. This study reveals the large uncertainty in climate classification and offers two robust KG maps that may help to better evaluate historical climate and quantify future climate shifts.

Detailed investigation of discrepancies in Köppen-Geiger climate classification using seven global gridded products

If two or more program entities (such as files, classes, methods) co-change (i.e., change together) frequently during software evolution, then it is likely that these two entities are coupled (i.e., the entities are related). Such a coupling is termed as evolutionary coupling in the literature. The concept of traditional evolutionary coupling restricts us to assume coupling among only those entities that changed together in the past. The entities that did not co-change in the past might also have coupling. However, such couplings can not be retrieved using the current concept of detecting evolutionary coupling in the literature. In this paper, we investigate whether we can detect such couplings by applying transitive rules on the evolutionary couplings detected using the traditional mechanism. We call these couplings that we detect using our proposed mechanism as transitive evolutionary couplings. According to our research on thousands of revisions of four subject systems, transitive evolutionary couplings combined with the traditional ones provide us with 13.96% higher recall and 5.56% higher precision in detecting future co-change candidates when compared with a state-of-the-art technique.

Snow interception in cold regions needleleaf forest canopies is a crucial process that controls local snow accumulation and redistribution over >20% of the Earth's land surface. Various ground-based methods exist to measure intercepted snow load, however all are based on single-tree measurements and are difficult to implement. No research has focussed on detecting large areal intercepted snow loads and no studies have assessed the use of satellite observations. In this study, four remote sensing indices (NDSI, NDVI, albedo, and land surface temperature (LST)) were retrieved from Landsat images to study their sensitivity to canopy intercepted snow and the possibility of using them to detect the presence of intercepted snow. The results indicate that presence of intercepted snow on canopy increased NDSI and albedo, but decreased NDVI. Intercepted snow presence also decreased the areal variability of NDSI and NDVI while increasing that of albedo. For these three indices, the differences between snow-free and snowcovered canopies were correlated to topography and forest canopy cover. Of these indices, NDSI changed the greatest. Intercepted snow noticeably decreased the LST difference between forest and open areas in springtime while the influence in wintertime was relatively smaller. An intercepted snow detection approach that uses both NDSI and NDVI to classify pixels into either snowcovered canopy or other (snow-free canopy and non-forest areas) is proposed here. A case study applying this approach compared remote sensing detection to simulations by the snow interception and sublimation model implemented in the Cold Regions Hydrological Modelling platform (CRHM). This used local meteorological observations from the pine, spruce and fir forest covered Marmot Creek Research Basin in the Canadian Rockies. The remote sensing detection of intercepted snow agreed well with CRHM simulations for continuous forests (83%) and less well for sparse forests (72%) and clearings with small trees (70%). Therefore, the approach is suitable for intercepted snow detection over continuous evergreen canopies. This technique provides a new capability for large-scale snow interception model validation and data assimilation to cold regions hydrological forecasting models.

Vernal pools are small, temporary, forested wetlands of ecological importance with a high sensitivity to changing climate and land-use patterns. These ecosystems are under considerable development pressure in southeastern Georgian Bay, where mapping techniques are required to inform wise land-use decisions. Our mapping approach combines common machine learning techniques [random forest, support vector machines (SVMs)] with object-based image analysis. Using multispectral image segmentation on high-resolution orthoimagery, we first created objects and assigned classes based on field collected data. We then supplied machine learning algorithms with data from freely available sources (Ontario orthoimagery and Sentinel 2) and tested accuracy on a reserved dataset. We achieved producer’s accuracies of 85 and 79% and user’s accuracies of 78 and 84% for random forest and SVMs models, respectively. Difficulty differentiating between small, dark shadows and small, obscured pools accounted for many of the omission and commission errors. Our automated approach of vernal pool classification provides a relatively accurate, consistent, and fast mapping strategy compared to manual photointerpretation. Our models can be applied on a regional basis to help verify the locations of pools in an area of Ontario that is in critical need of more detailed ecological information.

There are many real-world hydrological problems on the Canadian Prairies for which existing tools are poorly suited, due to the region’s complex cold-region hydrology and its equally complex hydrography, which is dominated by depressions, poorly defined and results in dynamic drainage basin contributing areas to streamflow. The complexities of the problems are compounded by changes in hydrology due to climate change, and by changes in depressional storage capacities due to wetland drainage. Although some hydrological models (CRHM, MESH) are able to simulate Prairie hydrology, including the effects of changes in climate, they do not have the ability to simulate detailed local-scale hydraulics. Conversely, hydraulic models which can simulate these small-scale features do not have the ability to simulate prairie hydrological processes. The Prairie Hydrology Design and Analysis Product (PHyDAP) deploys the research results of the GWF Prairie Water Project to produce a tool useful for solving local-scale water problems in the prairies. A classification of prairie basin types (Wolfe et al. 2019), was used to inform, design and parameterize a collection of Cold Region Hydrological Modelling platform (CRHM) models of “virtual” prairie basins. Each virtual basin model simulates the hydrological functioning of one of seven basin classes. PHyDAP consists of decades-long outputs from the virtual basin models, run for approximately 4000 basins, each approximately 100 km², across the Canadian Prairies. Each basin’s model is forced with local gridded meteorological data, using both historical values and downscaled simulations of future climates, for long time periods. The output variables are time series of rainfall, evaporation from ponded water and the runoff from uplands. We are working with partner organizations who will use the PHyDAP values as inputs to hydraulic models, such as SWMM, to simulate the complex local hydraulic conditions including the local depressions. This will facilitate calculation of the effects of climate change and wetland drainage/restoration on local return period flows and flooding, which can inform iterative design of infrastructure such as culverts. Examples of the data created for PHyDAP and its potential uses for solving small scale hydrological problems on the Prairies are shown.

Preferential flowpaths transport phosphorus (P) to agricultural tile drains. However, if and to what extent this may vary with soil texture, moisture conditions, and P placement is poorly understood. This study investigated (a) interactions between soil texture, antecedent moisture conditions, and the relative contributions of matrix and preferential flow and (b) associated P distributions through the soil profile when fertilizers were applied to the surface or subsurface. Brilliant blue dye was used to stain subsurface flowpaths in clay and silt loam plots during simulated rainfall events under wet and dry conditions. Fertilizer P was applied to the surface or via subsurface placement to plots of different soil texture and moisture condition. Photographs of dye stains were analysed to classify the flow patterns as matrix dominated or macropore dominated, and soils within plots were analysed for their water-extractable P (WEP) content. Preferential flow occurred under all soil texture and moisture conditions. Dye penetrated deeper into clay soils via macropores and had lower interaction with the soil matrix, compared with silt loam soil. Moisture conditions influenced preferential flowpaths in clay, with dry clay having deeper infiltration (92 ± 7.6 cm) and less dye–matrix interaction than wet clay (77 ± 4.7 cm). Depth of staining did not differ between wet (56 ± 7.2 cm) and dry (50 ± 6.6 cm) silt loam, nor did dominant flowpaths. WEP distribution in the top 10 cm of the soil profile differed with fertilizer placement, but no differences in soil WEP were observed at depth. These results demonstrate that large rainfall events following drought conditions in clay soil may be prone to rapid P transport to tile drains due to increased preferential flow, whereas flow in silt loams is less affected by antecedent moisture. Subsurface placement of fertilizer may minimize the risk of subsurface P transport, particularily in clay.

In addition to aiding in digestion of food and uptake of nutrients, microbiota in guts of vertebrates are responsible for regulating several beneficial functions, including development of an organism and maintaining homeostasis. However, little is known about effects of exposures to chemicals on structure and function of gut microbiota of fishes. To assess effects of exposure to polycyclic aromatic hydrocarbons (PAHs) on gut microbiota, male and female fathead minnows (Pimephales promelas) were exposed to environmentally-relevant concentrations of the legacy PAH benzo[a]pyrene (BaP) in water. Measured concentrations of BaP ranged from 2.3 × 10−3 to 1.3 μg L−1. The community of microbiota in the gut were assessed by use of 16S rRNA metagenetics. Exposure to environmentally-relevant aqueous concentrations of BaP did not alter expression levels of mRNA for cyp1a1, a “classic” biomarker of exposure to BaP, but resulted in shifts in relative compositions of gut microbiota in females rather than males. Results presented here illustrate that in addition to effects on more well-studied molecular endpoints, relative compositions of the microbiota in guts of fish can also quickly respond to exposure to chemicals, which can provide additional mechanisms for adverse effects on individuals.

Hobbi, S., Papalexiou, S.M., Rupa, C., Tang, G., Clark, M.P., Markonis, I., 2021. Discrepancies in Köppen-Geiger Climate Classification Using Ten Global Gridded Products. AGU Fall Meeting 2021, AGU.?

Discrepancies in Köppen-Geiger Climate Classification Using Ten Global Gridded Products

Holmes Jason , Discrete-time Markov chain modelling of the Ontario air quality health index & image classification of Diatom genera using decision tree-based classifiers, 2021

Discrete-time Markov chain modelling of the Ontario air quality health index & image classification of Diatom genera using decision tree-based classifiers

Differential impacts of policies or changes in environmental conditions on people is a growing area of interest to decision-makers, yet remains an often neglected area of study for the environmental valuation literature. Using data from a large national survey of over 24,000 people conducted in Canada, this paper implements a latent class Kuhn-Tucker recreation demand model to assess differences in preferences and values for nature-based activities. Preferences are disaggregated by self-reported Indigeneity, immigration status, and gender. We find that Indigenous people receive 63% greater benefits from participating in nature-based activities compared to non-Indigenous people living in Canada. Immigrants have the lowest participation in, and benefits associated with, nature-based activities. Similarly, women receive 21% lesser benefits associated with nature-based activities when compared to men. These results demonstrate that Indigenous peoples may be more vulnerable to adverse impacts on nature-based activities such as land-use changes, climate change, and government policies. The study also highlights the importance of disaggregated data and incorporating aspects of identity in the ecosystem service literature towards more equitable decision-making and reconciliation.

Distinguishing alpine valley landcover classes based on surface and subsurface hydrology

Phytoplankton monitoring is essential for better understanding and mitigation of phytoplankton bloom formation. We present a microfluidic cytometer with two imaging modalities for onsite detection and identification of phytoplankton: a lensless imaging mode for morphological features, and a fluorescence imaging mode for autofluorescence signal of phytoplankton. Both imaging modes are integrated in a microfluidic device with a field of view (FoV) of 3.7 mm × 2.4 mm and a depth of field (DoF) of 0.8 mm. The particles in the water flow channel can be detected and classified with automated image processing algorithms and machine learning models using their morphology and fluorescence features. The performance of the device was demonstrated by measuring Chlamydomonas, Euglena, and non-fluorescent beads in both separate and mixed flow samples. The recall rates for Chlamydomonas and Euglena ware 93.6% and 94.4%. The dual-modality imaging approach enabled observing both morphology and fluorescence features with a large DoF and FoV which contribute to high-throughput analysis. Moreover, this imaging flow cytometer platform is portable, low-cost, and shows potential in the onsite phytoplankton monitoring.

Tile-drainage area has expanded across the Canadian Lake Erie watershed in recent decades, and effects on phosphorus (P) loading are unclear. Eleven years (2010 to 2021) of daily P, total suspended solids (TSS), discharge, and climatological data were aggregated from three Canadian tributaries that form a gradient of tiled areas: East Sydenham River (ESR, 60% tile), Thames River (TR, 48% tile), and Grand River (GR, 23% tile). Instead of using traditional seasons (winter, spring, summer, fall), we classified seasons by air temperature to highlight hydrological periods of importance for P loss through tile drains. Seasons included frozen (<?3.2 °C), thawing (?3.2 – 6.7 °C), bare (6.7 – 15.9 °C), and growing (>15.9 °C). Nonparametric comparisons revealed that during every season, the ESR and TR had significantly higher soluble reactive P (SRP) and total P (TP) concentrations than the GR. For %SRP, the ESR was significantly higher than the other rivers during every season, while for TSS, the GR was significantly higher than the other rivers during every season. Only during the thawing season were positive relationships observed in every river between year-over-year tile-drainage proportion and associated P loadings and concentrations. The ESR was the only river to yield significant relationships between tile drainage and P in all seasons except the frozen season. Our findings suggest that increases in tile-drainage area can lead to increases in SRP loading to Lake Erie from Canadian tributaries, especially during the thawing season. However, effects of tile drainage are moderated by differences in soil texture, land-use-land-cover, climate, and point sources.

Large-scale pre-trained transformer models such as BERT have become ubiquitous in Natural Language Processing (NLP) research and applications. They bring significant improvements to both academia benchmarking tasks and industry applications: the average score on the General Language Understanding Evaluation benchmark (GLUE) has increased from 74 to 90+; commercial search engines such as Google and Microsoft Bing are also applying BERT-like models to search. Despite their exciting power, these increasingly large transformer-based models are notorious for having billions of parameters and being slow in both training and inference, making deployment difficult when inference time and resources are limited. Therefore, model efficiency has become a more important and urgent problem in the transformer era. In this thesis, we propose and innovate methods for efficient NLP models. We choose to specifically focus on inference efficiency: pre-trained models are almost always publicly available, and fine-tuning is performed on relatively small datasets without strict time constraints; inference, by contrast, needs to be performed repetitively and typically in a real-time setting. First, we propose the early exiting idea for transformers. Considering that the transformer model has multiple layers with identical structures, we try to reduce the number of layers used for inference by dynamic early exiting. During inference, if an intermediate transformer layer predicts an output of high confidence, we directly exit from this layer and use the current output as the final one. We apply the early exiting idea on sequence classification tasks and show that it is able to greatly improve inference efficiency. We then explore a few extensions to the early exiting idea: (1) early exiting for low-resource datasets - in this case, the straightforward fine-tuning methods fail to train the model to its full potential and we propose a method to better balance all layers of the model; (2) early exiting for regression datasets - in this case, the output is no longer a distribution where we can directly estimate confidence, and we design a learning-to-exit module to explicitly learn confidence estimation; (3) early exiting for document reranking - in this case, the two classes that the model tries to distinguish are highly asymmetric and we design an asymmetric early exiting method to better handle this task. We also extend early exiting to another direction - selective prediction. In this setting, if we have low confidence in the final prediction, we abstain from making predictions at all. We propose better ways for confidence estimation and also discuss a few applications for selective prediction. Finally, we discuss the combination of multiple efficiency methods, including early exiting itself and other popular methods such as distillation, pruning, quantization, etc. We propose a conceptual framework to treat each efficiency method as an operator. We conduct experiments to show interesting properties of these operators when they combine, which provide useful guidelines for designing and evaluating the application of combining multiple efficiency methods. The thesis presents a series of modeling and experimental contributions for efficient transformer models. We not only largely reduce the inference time for many NLP and IR applications, but also provide insights to understand the efficiency problem from a novel perspective.

Climate change, water stress, and rapid population growth have increased the need to manage water resources through innovative sustainable technologies. Decentralized systems such as membrane treatment trains have become increasingly important to provide high volumes of potable water to millions of people. Pressure-driven membrane systems have dominated separation processes due to their low cost, small footprint, ease of operation, and high permeate quality. Conventionally, pressure-driven membranes are classified into microfiltration (MF), ultrafiltration (UF), nanofiltration (NF), and reverse osmosis (RO). MF and UF membranes operate under low pressure (< 7 bar, <~100 psi). They can separate a variety of large particles such as bacteria, natural organic matter, suspended solids, and colloids. In contrast, NF and RO membranes are more energy-intense due to operating at high pressures (7 – 80 bar, ~100 – 1200 psi) and can remove small molecules such as ions, pharmaceuticals, and heavy metals. Fouling is a primary challenge with membranes that compromises the membrane performance, increases energy consumption, and reduces the membrane lifetime. Many strategies are used to address fouling, such as pre-treatment (pH adjustment, screening, coagulation), membrane modification (chemical and morphological properties), and membrane cleaning (physical, chemical). However, such strategies increase operational expenditures, produce waste products that can impact the environment, and negatively impact membrane lifetimes. Recently, electrically conductive membranes (ECMs) have been introduced to address the challenges with traditional membranes. They contain conductive surfaces that offer self-cleaning and antifouling properties across the surface in response to electrical potential externally applied to them. ECMs are advantageous as compared to traditional membranes because (a) they are more effective in treating foulants as they target foulants at the membrane/solvent interface, (b) they are more economical and environmentally friendly as they reduce the need for chemical consumption, and (c) they can be responsive to fouling conditions as their antifouling mechanisms can be easily manipulated by changing the applied current type (positive, negative, direct current, alternating current) to match the foulant. ECMs have been formed from all categories of membranes (MF, UF, NF, MD, FO, and RO) with a range of applications. Despite the remarkable progress in demonstrating their excellent antifouling performance, there are many hurdles to overcome before they can be commercialized. Two of these are (a) a fundamental understanding of their underlying mechanisms, (b) surface materials that can withstand extreme chemical and electrical conditions. In this work, we have comprehensively discussed antifouling mechanisms with respect to surface polarization and elaborated on the impact of electrically-induced mechanisms on four major fouling categories. i.e., biofouling, organic fouling, mineral scaling, and oil wetting. In addition, we characterized surface properties of a common electrically conductive composite membrane composed of carbon nanotubes (CNTs) and polyvinyl alcohol (PVA). We then investigated the impact of cross-linkers in CNT/PVA network on transmembrane flux, electrical conductivity, hydrophilicity, and surface roughness. In addition, we proposed standard, practical, and straightforward methodologies to detect and quantify the electrochemical, physical, and mechanical stability of ECMs, using chronoamperometry and cyclic voltammetry, an evaluation of polymer leaching from membranes, and micro mechanical scratch testing, respectively. Our methods can readily be extended to all membrane-based separation processes and different membrane materials (carbonaceous materials, ceramics, metal-based, and polymers). To demonstrate the antifouling properties of ECMs, we challenged ECMs with mixed-bacterial cultures in a flow-through system. Although ECMs showed high rejection, comparable flux, and excellent self-cleaning performance under application of electrical potential, understanding the relationship between applied electrical currents and antifouling mechanisms demands a well-controlled investigation. To this end, we quantified the impact of electrochemically-induced acidic conditions, alkaline conditions, and H2O2 concentration on model bacteria, Escherichia Coli. We first quantified the electrochemical potential of CNT-based ECMs in generating stressors such as protons, hydroxyl ions, and H2O2, under a range of applied electrical currents (± 0-150 mA). Next, these individual stressors with identical magnitude were imposed on E. Coli cells and biofilms in batch and flow-through systems, respectively. This thesis guides researchers to understand the underlying antifouling mechanisms associated with ECMs, how to match the mechanisms to the application of ECMs, and offers benchmarks for making practical ECMs.

Aim High-throughput pipelines supported by eDNA metabarcoding have been applied in various freshwater ecosystems. Both eDNA in ethanol (EtOH) samples (ES-eDNA) and in water samples (WS-eDNA) can provide comprehensive classification lists with good taxonomic resolution and coverage for determining freshwater biodiversity and biomonitoring. But, the advantages of ES-eDNA metabarcoding over WS-eDNA metabarcoding remain unclear for routine assessments of diversity of benthic macroinvertebrates in streams. Location Qiantang River Basin, China. Methods Here, we compared ES-eDNA and WS-eDNA metabarcoding to evaluate the performance of two eDNA workflows in determining biodiversity and recovery of damaged macroinvertebrate communities. All eDNA samples from the environment and bulk specimen of macroinvertebrates were processed into available molecular operational taxonomic units (MOTUs) and identified to the level of genus. Results WS-eDNA detected more exact sequence variants (ESVs) (formerly referred to as operational taxonomic units; OTUs), than did ES-eDNA (2,866 vs. 2,406), but fewer macroinvertebrate ESVs (381 vs. 481). Among sampling sites, the two eDNA workflows exhibited relatively large dissimilarity on inferred community composition (p < .001). Furthermore, ES-eDNA metabarcoding exhibited more consistent with morphological identification approaches than did WS-eDNA metabarcoding (24.24% vs. 17.63%, p = .002), especially for species identified by traditional morphology (morphotaxa). Main conclusions Based on the attributes of ES-eDNA and WS-eDNA, it is suggested that ES-eDNA metabarcoding performs better than does WS-eDNA metabarcoding in detecting local biodiversity and was consistent with morphological results, while WS-eDNA was more suitable for exploring biodiversity patterns on a broad scale, as it is the easiest and most convenient way to collect samples. Results of this study suggest ES-eDNA metabarcoding could be an option in building molecular measurement biomonitoring programme based on EtOH sample used for preserving biological samples.

Accurate estimation of precipitation fields remains a grand challenge in cold regions hydrology due to the sparseness of precipitation gauges and lack of quantitative precipitation estimation from ground based weather radars. It is even more challenging in cold regions mountains due to blockage of ground based weather radars, and the complexity of modelling precipitation. Satellite remote sensing provides an alternative to precipitation monitoring in complex terrain when retrieval algorithms are able to represent a range of hydrometeor types. The Global Precipitation Measurement (GPM) satellite constellation has been successfully monitoring liquid precipitation since early 2014 through the Integrated Multi-SatellitE Retrievals for GPM (IMERG) algorithm. IMERG uses a constellation of satellites with passive microwave and infrared sensors to estimate precipitation, which is intercalibrated using GPM’s core platform Dual-frequency Precipitation Radar (DPR). The DPR sensor has been working efficiently to retrieve liquid precipitation because reflectivity-rainfall relationships are well established due to a more uniform Particle Size Distribution (PSD); however, DPR’s snowfall retrieval has been suboptimal because of the diversity of PSDs, especially in complex terrain. For example, the same radar reflectivity can represent differing precipitation rates for different PSDs. Therefore, proper PSD-specific reflectivity-snowfall (Z-S) relationships need to be established to improve the accuracy of DPR’s snowfall retrievals and, consequently, IMERG precipitation estimates. This study aims to develop PSD-specific Z-S relationships to improve satellite snowfall estimates in the complex terrain of the Fortress Mountain Snow Laboratory, Canadian Rockies, Alberta. Observations from instruments located between 2100 and 2310 m above sea level included a Parsivel-2 optical disdrometer to determine hydrometeor phase and PSD at ~ 3 m above the surface; a Micro Rain Radar-2 (MRR-2) to establish Z-S relationships for each classified PSD at the near-surface and at different levels of the atmosphere; and a network of Alter-shielded weighing precipitation gauges, including one next to the Parsivel-2 and MRR-2, to quantify the precipitation rate used on the Z-S relationship equations. The study period comprises matching Parsivel-2 and MRR-2 observed events during the Storms Across the Continental Divide Experiment (SPADE), April-June 2019, and between January 2020 and March 2022. Preliminary results during SPADE events indicate that snowfall rate percentage differences between PSD-specific and general Z-S relationships range from -51% to 18%. This study’s findings explore the biases of satellite precipitation in cold mountain river basins. Improved precipitation satellite estimates can supplement precipitation gauges and modelling to yield an improved characterization of snowfall in mountains.

Rising human populations and ever-increasing demand for potable water result in increased municipal wastewater production. The collection, treatment, and management of municipal wastewaters include energy-intensive processes leading to the generation and emission of greenhouse, potentially toxic, and odorous gases. The main goal of this thesis was to advance knowledge of greenhouse gas (including carbon dioxide, CO2; methane, CH4; and nitrous oxide, N2O) and smelly compound (including ammonia, NH3; and hydrogen sulphide, H2S) emissions from typical municipal wastewater treatment plants (MWTPs) to accurately describe their emission rate estimates (EREs) using operating parameters. This research included laboratory and field assessments of greenhouse gas (GHG) and odour emissions in conjunction with monitored operating parameters. Laboratory-scale reactors simulating open-to-air treatment processes including primary and secondary clarifiers and anaerobic, anoxic, and aerobic reactors, were used to monitor gas EREs using wastewater samples taken from the analogous MWTP processes in winter and summer seasons. The Saskatoon Wastewater Treatment plan (SWTP) is a state-of-the-art biological nutrient removal (BNR) type MWTP and a Class IV treatment facility in Canada which was selected as a case study given its highly variable seasonal temperatures from −40 °C to 30 °C and its geographic location near the University of Saskatchewan. The experimental results were then used to develop a variety of novel machine learning models describing gas EREs with further optimization of operating parameters using genetic algorithm (GA). Studied machine learning models were artificial data generation algorithms (including generative adversarial network, GAN) and data-driven models (including artificial neural network, ANN; adaptive network-based fuzzy inference systems, ANFIS; and linear/non-linear regression models). To my knowledge, this is the first application of GAN used for MWTP modelling purposes. Results indicated that anaerobic digestion EREs averagely reached 4,443 kg CH4/d, 9,145 kg CO2/d, and 59.7 kg H2S/d. In contrast, GHG and odour ERE variabilities given ambient temperature changes were more noticeable for open-to-air treatment processes such that the winter EREs were 45,129 kg CO2/d, 21.9 kg CH4/d, 3.20 kg N2O/d, and insignificant for H2S and NH3. The higher temperature for the summer samples resulted in increased EREs for CH4, N2O, and H2S EREs of 33.0 kg CH4/d, 3.87 kg N2O/d, and 2.29 kg H2S/d, respectively, and still insignificant NH3 emissions. However, the CO2 EREs were reduced to 37,794 kg CO2/d, and interestingly, NH3 emissions were still negligible. Overall, the aerobic reactor was the dominant source of GHG emissions for both seasons, and changes in the aerobic reactor aeration rates (in reactor) and BNR treatment configurations (from site) further impacted the EREs. The integration of field monitoring data with data-driven models showed that the ANN, ANFIS, and regression models provided reasonable EREs using: (1) volatile fatty acids, total/fixed/volatile solids, pH, and inflow rate for anaerobic digestion biogas generations; and (2) hydraulic retention time, temperature, total organic carbon, dissolved oxygen, phosphate, and nitrogen concentrations for aerobic GHG emissions. However, when both model accuracy and uncertainty were considered there appears to be a compromise between these parameters with no model having simultaneously both high accuracy and low uncertainty. Additionally, and interestingly, virtual data augmentation using GAN was found to be a valuable resource in supplementation of limited data for improved modelling outcomes. GA was also coupled with the data-driven models to determine optimal operating parameters resulting in either GHG emission maximization given biogas could be beneficial for energy generation or GHG emission minimization given the aerobic reactor is an open-to-air process that can impact nearby residential neighbourhood air quality. The current study provides a hybrid methodology of mathematical modelling and experiments that can be used to accurately estimate and optimize the GHG and odour EREs from other MWTPs in Canada and worldwide.

Phosphorus (P) loss in agricultural discharge has typically been associated with surface runoff; however, tile drains have been identified as a key P pathway due to preferential transport. Identifying when and where these pathways are active may establish high-risk periods and regions that are vulnerable for P loss. A synthesis of high-frequency, runoff data from eight cropped fields across the Great Lakes region of North America over a 3-yr period showed that both surface and tile flow occurred year-round, although tile flow occurred more frequently. The relative timing of surface and tile flow activation was classified into four response types to infer runoff-generation processes. Response types were found to vary with season and soil texture. In most events across all sites, tile responses preceded surface flow, whereas the occurrence of surface flow prior to tile flow was uncommon. The simultaneous activation of pathways, indicating rapid connectivity through the vadose zone, was seldom observed at the loam sites but occurred at clay sites during spring and summer. Surface flow at the loam sites was often generated as saturation-excess, a phenomenon rarely observed on the clay sites. Contrary to expectations, significant differences in P loads in tiles were not apparent under the different response types. This may be due to the frequency of the water quality sampling or may indicate that factors other than surface-tile hydrologic connectivity drive tile P concentrations. This work provides new insight into spatial and temporal differences in runoff mechanisms in tile-drained landscapes.

Daniel Wu (2019). Evaluating impacts of thinning in a temperate pine forest using CLASS-CTEM Model (B.Sc. thesis). McMaster University, Hamilton, Canada. Thesis

Evaluating impacts of thinning in a temperate pine forest using CLASS-CTEM Model (B.Sc. thesis

High-latitude ecosystems have experienced substantial warming over the past 40 years, which is expected to continue into the foreseeable future. Consequently, an increase in vegetation growth has occurred throughout the circumpolar North as documented through remote sensing and plot-level studies. A major component of this change is shrub expansion (shrubbing) in arctic and subarctic ecotones. However, these changes are highly variable depending on plant species, topographic position, hydrology, soils and other ecosystem properties. Changes in shrub and other vegetation properties are critical to document due to their first-order control on water, energy and carbon balances. This study uses a combination of multi-temporal LiDAR (Light Detection and Ranging) and field surveys to measure temporal changes in shrub vegetation cover over the Wolf Creek Research Basin (WCRB), a 180 km2 long-term watershed research facility located ~15 km south of Whitehorse, Yukon Territory. This work focuses on the smaller Granger Basin, a 7.6 km2 subarctic headwater catchment that straddles WCRB’s subalpine and alpine tundra ecozones with a wide range of elevation, landscape topography, and vegetation. Airborne LiDAR surveys of WCRB were conducted in August 2007 and 2018, providing an ideal opportunity to explore vegetation changes between survey years. Vegetation surveys were conducted throughout Granger Basin in summer 2019 to evaluate shrub properties for comparisons to the LiDAR. Machine learning classification algorithms were used to predict shrub presence/absence in 2018 based on rasterized LiDAR metrics with up to 97% overall independent accuracy compared to field validation points, with the best-performing model applied to the 2007 LiDAR to create binary shrub cover layers to compare between survey years. Results show a 63.3% total increase in detectable shrub cover > 0.45 m in height throughout Granger Basin between 2007 and 2018, with an average yearly expansion of 5.8%. These changes in detectable shrub cover were compared across terrain derivatives created using the LiDAR to quantify the influence of topography on shrub expansion. The terrain comparison results show that shrubs in the study area are located in and are preferentially expanding into lower and flatter areas near stream networks, at lower slope positions and with a higher potential for topographic wetness. The greatest differences in terrain derivative value distributions across the shrub and non-shrub change categories were found in terms of stream distance, elevation, and relative slope position. This expansion of shrubs into higher-resource areas is consistent with previous studies and is supported by established physical processes. As vegetation responses to warming have far-reaching influences on surface energy exchange, nutrient cycling, and the overall water balance, this increase in detectable shrub cover has a wide range of impacts on the future of northern watersheds. Overall, the findings from this research reinforce the documented increase in pan-Arctic shrub vegetation in recent years, quantify the variation in shrub expansion over terrain derivatives at the landscape scale, and demonstrate the feasibility of using LiDAR to compare changes in shrub properties over time.

As wetlands around the world are being lost, policies are implemented to help protect further destruction and loss of valuable services that wetlands provide. In Alberta, wetland policy has been put in place with the goals of protecting the most valuable wetlands and replacing necessary loss of wetlands to maintain functional value. To help the policy meet its objectives, the Alberta Wetland Rapid Evaluation Tool-Actual (ABWRET-A) was developed and implemented in Alberta’s settled area in 2015 as a standardized way to give a value score via functional assessment to any wetland in the province, with the hopes that the most valuable wetlands will be conserved. These assessment tools are in constant need of review and improvement to make sure they are helping meet policy goals. I assess biases made in the selection for ABWRET-A calibration wetlands and determine how these biases affect ABWRET-A scoring to determine if subsequent scores provided by this tool are over or under estimating wetland value. I also assess the wetlands that underwent ABWRET-A evaluation and were drained or filled in under a permit in the 1.5 yr after ABWRET-A implementation in Alberta’s settled region to determine whether they mirror the calibration wetlands. I found that the calibration dataset comprised larger, more permanently ponded wetlands distributed closer to roads than the general wetland population. I also found that the calibration dataset included fewer bogs and more fens. I found that larger wetlands and wetlands classified as fens received higher ABWRET-A scores, whereas wetlands close to roads received lower scores. Consequently, I surmise that the scores being given out since ABWRET-A’s implementation are likely underestimates. This is corroborated by a lower distribution of scores in the wetlands permitted for drainage than policy recommends. The wetlands being targeted for permitted loss were also smaller, more road-proximate, and concentrated around major cities, implying permanent regional loss of those wetlands and their functions. Based on these findings, I make suggestions for improving ABWRET-A, including adding calibration sites to better capture the natural variability of wetlands in the area to improve ABWRET-A’s accuracy in estimating relative wetland value.

Supercells are the most violent thunderstorm, which can produce strong tornadoes, large hail, damaging winds, and flash flooding. From June to August, the Canadian Prairies (i.e., a major agricultural region of Canada) are greatly affected by various supercells, including low-precipitation (LP), high-precipitation (HP), and classic supercells. Despite the enormous socioeconomic impact of supercell thunderstorms over the Canadian Prairies, it remains poorly understood which environmental conditions are favorable for the development and maintenance of supercells. Therefore, this work aims to study the dynamical and thermodynamical characteristics of the environment during different supercells by examining severe weather parameters derived from soundings. Specifically, this study analyzed 23 LP, 15 HP, and 16 classic-type supercells that occurred over the Canadian Prairies and compared their dynamical and thermodynamical characteristics with their US Great Plains counterparts. Preliminary results show that LP supercell develops predominantly over Alberta, Classic Supercell forms over Saskatchewan, and HP supercell develops over Manitoba, respectively. The dynamical and thermodynamical atmospheric variables of the prairie LP supercell are significantly lower than those of the U.S. Great Plains counterpart. HP supercells over the Canadian Prairies usually develop in an environment with relatively lower convective available potential energy than their U.S. Great Plains counterparts. The development of the classic supercell is controlled by an elevated mixed layer that is associated with the Rocky Mountains. Our result will help operational weather forecasters prepare for improved severe weather forecasting, saving lives and properties.

BACOLI is a Fortran software package for solving one-dimensional parabolic partial differential equations (PDEs) with separated boundary conditions by B-spline adaptive collocation methods. A distinguishing feature of BACOLI is its ability to estimate and control error and correspondingly adapt meshes in both space and time. Many models of scientific interest, however, can be formulated as multiscale parabolic PDE systems, that is, models that couple a system of parabolic PDEs describing dynamics on a global scale with a system of ordinary differential equations describing dynamics on a local scale. This article describes the Fortran software eBACOLI, the extension of BACOLI to solve such multiscale models. The performance of the extended software is demonstrated to be statistically equivalent to the original for purely parabolic PDE systems. Results from eBACOLI are given for various multiscale models from the extended problem class considered.

Testing software is considered to be one of the most crucial phases in software development life cycle. Software bug fixing requires a significant amount of time and effort. A rich body of recent research explored ways to predict bugs in software artifacts using machine learning based techniques. For a reliable and trustworthy prediction, it is crucial to also consider the explainability aspects of such machine learning models. In this paper, we show how the feature transformation techniques can significantly improve the prediction accuracy and build confidence in building bug prediction models. We propose a novel approach for improved bug prediction that first extracts the features, then finds a weighted transformation of these features using a genetic algorithm that best separates bugs from non-bugs when plotted in a low-dimensional space, and finally, trains the machine learning model using the transformed dataset. In our experiment with real-life bug datasets, the random forest and k-nearest neighbor classifier models that leveraged feature transformation showed 4.25% improvement in recall values on an average of over 8 software systems when compared to the models built on original data.

Testing and debugging software to fix bugs is considered one of the most important stages of the software life cycle. Many studies have investigated ways to predict bugs in software artifacts using machine learning techniques. It is important to consider the explanatory aspects of such models for reliable prediction. In this paper, we show how feature transformation can significantly improve prediction accuracy and provide insight into the inner workings of bug prediction models. We propose a new approach for bug prediction that first extracts the features, then finds a weighted transformation of these features using a genetic algorithm that best separates bugs from non-bugs when plotted in a low-dimensional space, and finally, trains predictive models using the transformed dataset. In our experiment using the proposed feature transformation, the traditional machine learning and deep learning classifiers achieved an average improvement of 4.25% and 9.6% in recall values for bug classification over 8 software systems compared to the models built on original data. We also examined the generalizability of our concept for multiclass classification tasks such as commit classification in software systems and found modest improvements in F1-scores (sometimes up to 3%) for traditional machine learning models and 4% with deep learning models.

Sakib Mostafa, Shamse Tasnim Cynthia, Banani Roy, Debajyoti Mondal (2024) Feature transformation for improved software bug detection and commit classification, Journal of Systems and Software, Volume 219, 2025, 112205, ISSN 0164-1212

Feature transformation for improved software bug detection and commit classification

Cowichan Lake lamprey (Entosphenus macrostomus) is a threatened species resident to Mesachie Lake, Cowichan Lake, and adjoining Bear Lake and their major tributaries in British Columbia. Decreases in trapping success have created concerns that the population is declining. Some potential threats include water use, climate change, and management actions. Owing to the absence of long-term data on population trends, little information is available to estimate habitat quality and factors that influence it. We sought to fill this gap by examining associations between habitat area and variables representing suspected key drivers of habitat availability. Critical habitat areas were imaged using an unmanned aerial vehicle over a period of three years at three sites at Cowichan Lake and a subsequent habitat area was classified. Meteorological and anthropogenic controls on habitat area were investigated through automatic relevance detection regression models. The major driver of habitat area during the critical spawning period was water level during the storage season, which also depends on the meteorological variables and anthropogenic control. It is recommended that regulation of the weir should aim to ensure that the water level remains above the 1 m mark, which roughly equates to the 67% coverage of water on the habitat area used for spawning.

In this paper, we develop a radar snow water equivalent (SWE) retrieval algorithm based on a parameterized forward model of bicontinuous dense media radiative transfer (Bic-DMRT). The algorithm is based on retrieving the absorption loss of the snowpack which is directly proportional to the SWE. In the algorithm, Bic-DMRT is first applied to generate a lookup table (LUT) of snowpack backscattering at X- and Ku-band. Regression training is applied to the LUT to transform the dual-frequency backscatter into functions of two parameters: the scattering albedo at X-band and SWE. The background scattering is subtracted from the SnowSAR data to give the volume scattering of snow. Classification of SnowSAR data is applied to provide a priori information. Based on the obtained volume scattering and the priori information, a cost function is established to find SWE. Performance of the retrieval algorithm was tested using three sets of airborne SnowSAR data acquired over mixed areas in Finland and open tundra landscape in Canada. It is shown that the retrieval algorithm has a root-mean-square error below 30 mm of SWE and a correlation coefficient above 0.64.

Flow management has the potential to significantly affect ecosystem condition. Shallow lakes in arid regions are especially susceptible to flow management changes, which can have important implications for the formation of cyanobacterial blooms. Here, we reveal water quality shifts associated with changing source water inflow management. Using in situ monitoring data, we studied a seven-year time span during which inflows to a shallow, eutrophic drinking water reservoir transitioned from primarily natural landscape runoff (2014–2015) to managed flows from a larger upstream reservoir (Lake Diefenbaker; 2016–2020) and identified significant changes in cyanobacteria (as phycocyanin) using generalized additive models to classify cyanobacterial bloom formation. We then connected changes in water source with shifts in chemistry and the occurrence of cyanobacterial blooms using principal components analysis. Phycocyanin was greater in years with managed reservoir inflow from a mesotrophic upstream reservoir (2016–2020), but dissolved organic matter (DOM) and specific conductivity, important determinants of drinking water quality, were greatest in years when landscape runoff dominated lake water source (2014–2015). Most notably, despite changing rapidly, it took multiple years for lake water to return to a consistent and reduced level of DOM after managed inflows from the upstream reservoir were resumed, an observation that underscores how resilience may be hindered by weak resistance to change and slow recovery. Environmental flows for water quality are rarely defined, yet we show that trade-offs exist between poor water quality via elevated conductivity and DOM and higher bloom risk, depending on water source. Our work highlights the importance of source water quality, not just quantity, to water security, and our findings have important implications for water managers who must protect ecosystem services while adapting to projected hydroclimatic change.

Smart packaging of fresh produce is an emerging technology toward reduction of waste and preservation of consumer health and safety. Smart packaging systems also help to prolong the shelf life of perishable foods during transport and mass storage, which are difficult to regulate otherwise. The use of these ever-progressing technologies in the packaging of fruits has the potential to result in many positive consequences, including improved fruit quality, reduced waste, and associated improved public health. In this review, we examine the role of smart packaging in fruit packaging, current-state-of-the-art, challenges, and prospects. First, we discuss the motivation behind fruit quality monitoring and maintenance, followed by the background on the development process of fruits, factors used in determining fruit quality, and the classification of smart packaging technologies. Then, we discuss conventional freshness sensors for packaged fruits including direct and indirect freshness indicators. After that, we provide examples of possible smart packaging systems and sensors that can be used in monitoring fruits quality, followed by several strategies to mitigate premature fruit decay, and active packaging technologies. Finally, we discuss the prospects of smart packaging application for fruit quality monitoring along with the associated challenges and prospects.

Fluvial systems in southern Ontario are regularly affected by widespread early-spring flood events primarily caused by rain-on-snow events. Recent studies have shown an increase in winter floods in this region due to increasing winter temperature and precipitation. Streamflow simulations are associated with uncertainties mainly due to the different scenarios of greenhouse gas emissions, global climate models (GCMs) or the choice of the hydrological model. The internal variability of climate, defined as the chaotic variability of atmospheric circulation due to natural internal processes within the climate system, is also a source of uncertainties to consider. Uncertainties of internal variability can be assessed using hydrological models fed by downscaled data of a global climate model large ensemble (GCM-LE), but GCM outputs have too coarse of a scale to be used in hydrological modeling. The Canadian Regional Climate Model Large Ensemble (CRCM5-LE), a 50-member ensemble downscaled from the Canadian Earth System Model version 2 Large Ensemble (CanESM2-LE), was developed to simulate local climate variability over northeastern North America under different future climate scenarios. In this study, CRCM5-LE temperature and precipitation projections under an RCP8.5 scenario were used as input in the Precipitation Runoff Modeling System (PRMS) to simulate streamflow at a near-future horizon (2026–2055) for four watersheds in southern Ontario. To investigate the role of the internal variability of climate in the modulation of streamflow, the 50 members were first grouped in classes of similar projected change in January–February streamflow and temperature and precipitation between 1961–1990 and 2026–2055. Then, the regional change in geopotential height (Z500) from CanESM2-LE was calculated for each class. Model simulations showed an average January–February increase in streamflow of 18 % (±8.7) in Big Creek, 30.5 % (±10.8) in Grand River, 29.8 % (±10.4) in Thames River and 31.2 % (±13.3) in Credit River. A total of 14 % of all ensemble members projected positive Z500 anomalies in North America's eastern coast enhancing rain, snowmelt and streamflow volume in January–February. For these members the increase of streamflow is expected to be as high as 31.6 % (±8.1) in Big Creek, 48.3 % (±11.1) in Grand River, 47 % (±9.6) in Thames River and 53.7 % (±15) in Credit River. Conversely, 14 % of the ensemble projected negative Z500 anomalies in North America's eastern coast and were associated with a much lower increase in streamflow: 8.3 % (±7.8) in Big Creek, 18.8 % (±5.8) in Grand River, 17.8 % (±6.4) in Thames River and 18.6 % (±6.5) in Credit River. These results provide important information to researchers, managers, policymakers and society about the expected ranges of increase in winter streamflow in a highly populated region of Canada, and they will help to explain how the internal variability of climate is expected to modulate the future streamflow in this region.

Time series of wetland methane fluxes measured by eddy covariance require gap-filling to estimate daily, seasonal, and annual emissions. Gap-filling methane fluxes is challenging because of high variability and complex responses to multiple drivers. To date, there is no widely established gap-filling standard for wetland methane fluxes, with regards both to the best model algorithms and predictors. This study synthesizes results of different gap-filling methods systematically applied at 17 wetland sites spanning boreal to tropical regions and including all major wetland classes and two rice paddies. Procedures are proposed for: 1) creating realistic artificial gap scenarios, 2) training and evaluating gap-filling models without overstating performance, and 3) predicting half-hourly methane fluxes and annual emissions with realistic uncertainty estimates. Performance is compared between a conventional method (marginal distribution sampling) and four machine learning algorithms. The conventional method achieved similar median performance as the machine learning models but was worse than the best machine learning models and relatively insensitive to predictor choices. Of the machine learning models, decision tree algorithms performed the best in cross-validation experiments, even with a baseline predictor set, and artificial neural networks showed comparable performance when using all predictors. Soil temperature was frequently the most important predictor whilst water table depth was important at sites with substantial water table fluctuations, highlighting the value of data on wetland soil conditions. Raw gap-filling uncertainties from the machine learning models were underestimated and we propose a method to calibrate uncertainties to observations. The python code for model development, evaluation, and uncertainty estimation is publicly available. This study outlines a modular and robust machine learning workflow and makes recommendations for, and evaluates an improved baseline of, methane gap-filling models that can be implemented in multi-site syntheses or standardized products from regional and global flux networks (e.g., FLUXNET).

Aquatic environments with high levels of dissolved ferrous iron and low levels of sulfate serve as an important systems for exploring biogeochemical processes relevant to the early Earth. Boreal Shield lakes, which number in the tens of millions globally, commonly develop seasonally anoxic waters that become iron rich and sulfate poor, yet the iron–sulfur microbiology of these systems has been poorly examined. Here we use genome-resolved metagenomics and enrichment cultivation to explore the metabolic diversity and ecology of anoxygenic photosynthesis and iron/sulfur cycling in the anoxic water columns of three Boreal Shield lakes. We recovered four high-completeness and low-contamination draft genome bins assigned to the class Chlorobia (formerly phylum Chlorobi) from environmental metagenome data and enriched two novel sulfide-oxidizing species, also from the Chlorobia. The sequenced genomes of both enriched species, including the novel “Candidatus Chlorobium canadense”, encoded the cyc2 gene that is associated with photoferrotrophy among cultured Chlorobia members, along with genes for phototrophic sulfide oxidation. One environmental genome bin also encoded cyc2. Despite the presence of cyc2 in the corresponding draft genome, we were unable to induce photoferrotrophy in “Ca. Chlorobium canadense”. Genomic potential for phototrophic sulfide oxidation was more commonly detected than cyc2 among environmental genome bins of Chlorobia, and metagenome and cultivation data suggested the potential for cryptic sulfur cycling to fuel sulfide-based growth. Overall, our results provide an important basis for further probing the functional role of cyc2 and indicate that anoxygenic photoautotrophs in Boreal Shield lakes could have underexplored photophysiology pertinent to understanding Earth’s early microbial communities.

Changes in the frequency and intensity of extreme precipitation resulting from climate change are responsible for natural disasters such as severe floods and have been a major study focus during the last decades. Previous studies have mainly focused on the trends of annual maxima precipitation at global and regional scales. However, little is known about how extreme precipitation trends change among different climate types. This study offers a global analysis of extreme precipitation changes in terms of climate type by using over 8500 gauge-based records. We focus on the period 1964 to 2013 when global warming was accelerating. A climate type is assigned to each station based on the Köppen Geiger (KG) climate classification, resulting in 30 KG climate subtypes. Mann-Kendall test and Sen’s slope estimator are applied to each time series, measuring the magnitude and significance of trends. The heaviness of the tail for each station is assessed based on the shape parameter of the Generalized Extreme Value distribution. Our results indicate a decreasing trend for the majority of stations associated with some of the arid, temperate, and continental subtypes (i.e., hot semi-arid (BSh); hot-summer temperate (Csa); warm-summer temperate (Csb); and warm, dry-summer continental (Dsb)). An increasing trend is observed for the stations associated with the remaining KG subtypes, especially stations associated with dry-summer subarctic (Dsc) and monsoon-influenced extremely cold subarctic (Dwd). A significant increasing trend is estimated for 9.7% of stations located in the eastern USA, Asia, and northern Europe. However, only 2% of stations, mainly in eastern Australia and the central USA have a significant decreasing trend. The heaviness of the tail is the largest in the Polar major climate type (E), followed by Tropical (A), Dry (B), Continental (D), and Temperate (C). For the climate subtypes, large heavy-tailed extremes are observed in extremely cold subarctic (Dfd), polar tundra (ET), and tropical monsoon (Am), while only light-tailed extremes were observed in subpolar oceanic (Cfc). This study reveals the relationship of extreme precipitation characteristics (e.g., tail heaviness and trend) with the climate types at the global scale.

Global analysis of extreme precipitation changes in the Koumlppen-Geiger climate classification

Climate change is contributing to extreme climate events such as prolonged heat waves, hurricanes, and flooding. Climate classification schemes have become critical tools in investigating these events. One of the most widely used schemes is the Köppen-Geiger (KG) classification, which groups the world’s climate types using multiple variables based on precipitation and temperature data. Studies that apply the KG classification have a variety of purposes, including to present the geographical distribution of climate types, to measure shifts among climate types, to study changes in extreme events at regional scales, and to present future projections of climate types. However, several aspects of KG classification have not been thoroughly investigated in the literature: First, few studies have explored the differences among climate types at the global scale derived from multiple sources of precipitation and temperature data; second, little research has looked at changes in extreme precipitation in the KG climate classification at a global scale. This research work points out discrepancies in global climate types by analysing climate maps derived from different globally gridded datasets of precipitation and temperature from 1980 to 2017. Similarity and uncertainty among KG maps at the global and zonal scales are presented. By reducing uncertainty in maps, the research presents robust representations of KG climate types in a new map. This map was applied to assign the climate types of daily station rainfall records (1964 to 2013) to measure changes in extreme precipitation in the KG climate classification. For stations associated with different KG climate types, an analysis was carried out on the annual maxima precipitation time series to measure the trend and heaviness of the tail using the Mann-Kendall test and extreme value theory, respectively. Results from this thesis are as follows: (1) there was large uncertainty in the KG climate classification in the Middle East, northern Russia, eastern, and central Africa; (2) the highest and lowest similarity among the KG maps was observed in the North and South Temperate zones; (3) WFDEI is likely the most reliable dataset to determine KG climate types; (4) of all station records, those associated with Af, Am, Aw, and Cwa climate type showed larger variation in the magnitude of extreme precipitation trends; (5) a significant increasing trend was found in 9.7% of stations in the eastern USA, Asia, and northern Europe, while a significant decreasing trend was observed in only 2% of stations in eastern Australia and central USA; (6) a decreasing extreme precipitation trend was seen only over the majority of stations associated with BSh, Csa, Csb, and Dsb, whereas an increasing trend was observed in the remaining climate types; and (7) large heavy-tailed extremes were observed in Dfd, ET, and Am, while only light-tailed extremes were observed in Cfc. These results will be useful for scientists studying KG climate classification and the relationship between extreme precipitation changes and climate types.

Hobbi, Salma (2021) Global characteristics of extreme precipitation and variation of climate types from Köppen-Geiger classification using different datasets, USASK Harvest - Theses and Dissertations, https://hdl.handle.net/10388/13492

Global characteristics of extreme precipitation and variation of climate types from Köppen-Geiger classification using different datasets

A graphical analysis method is applied over the province of Alberta, Canada using publicly available water level data from standard monitoring wells to evaluate and categorize aquifer recovery. Agriculture in the province relies heavily upon surface water for irrigation, which is increasingly unreliable due to climate change and increasing climate variability. Due to an expected future reliance on groundwater, it is necessary to better understand groundwater flow and aquifer characteristics across Alberta to prevent over-allocation of groundwater resources. Water level data from provincial monitoring well hydrographs are examined and graphically analyzed to broadly characterize recovery in agriculturally significant regions of the province of Alberta, Canada. Through this analysis, the presence of a recharge boundary within a recovery curve can be ascertained. Of the 292 monitoring wells originally screened, recovery curve analysis is conducted on 49 monitoring wells. Using graphical analysis of recovery curves within monitoring well hydrographs, the presence or absence of recovery or aquifer replenishment in an area immediately surrounding monitoring well screens is determined. 785 recovery curves from the 49 monitoring wells are subsequently categorized as either “enhanced recovery”, “normal recovery”, or “inconclusive”, with continuing discussion and analysis focusing on results from 36 wells located in three significant aquifers within the province. These aquifers include the Paskapoo aquifer, aquifers within the irrigation districts of southern Alberta, and surficial aquifers within agriculturally rich regions of the province. Results demonstrate the presence of a potential recharge signal deviating from standard Theis recovery curve in 97.22% of the 36 monitoring wells studied. In individual wells, recovery curve classifications vary over time, with some recovery curves being classified as “normal recovery”, and some being classified as “enhanced recovery”, showing signs of a possible recharge boundary. This classification depends on the characterization of late-time recovery curve behavior, as pumping signals transition to regional aquifer signals over time. Analyzed hydrographs show the influence and effects of changes in groundwater pumping on surrounding water levels, including through change in water policy. This method provides information about the presence or absence of recharge over a large area, in contrast to traditional methods of determining recharge which cover smaller areas in comparison. However, a comprehensive database of monitoring well data are required to facilitate analysis, as 48.76% of recovery curves analyzed were classified as “inconclusive”. It is recommended that results from this method are paired with data such as climate indices or agricultural usage, to help determine possible correlations between results and climatic, geographic, or agricultural factors.

New members of a software team can struggle to locate user requirements if proper software engineering principles are not practiced. Reading through code, finding relevant methods, classes and files take a significant portion of software development time. Many times developers have to fix issues in code written by others. Having a good tool support for this code browsing activity can reduce human effort and increase overall developers' productivity. To help program comprehension activities, building an abstract code summary of a software system from the call graph is an active research area. A call graph is a visual representation of caller-callee relationships between different methods of a software project. Call graphs can be difficult to comprehend for a larger code-base. The motivation is to extract the essence from the call graph by finding execution scenarios from a call graph and then cluster them together by concentrating the information in the code-base. Later, different techniques are applied to label nodes in the abstract code summary tree. In this thesis, we focus on static call graphs for creating an abstract code summary tree as it clusters all possible program scenarios and groups similar scenarios together. Previous work on static call graph clusters execution paths and uses only one information retrieval technique without any feedback from developers. First, to advance existing work, we introduced new information retrieval techniques alongside human-involved evaluation. We found that developers prefer node labels generated by terms in method names with TFIDF (term frequency-inverse document frequency). Second, from our observation, we introduced two new types of information (text description using comments and execution patterns) for abstraction nodes to provide better overview. Finally, we introduced an interactive software tool which can be used to browse the code-base in a guided way by targeting specific units of the source code. In the user study, we found developers can use our tool to overview a project alongside finding help for doing particular jobs such as locating relevant files and understanding relevant domain knowledge.

Permafrost thaw has been observed in recent decades in the Northern Hemisphere and is expected to accelerate with continued global warming. Predicting the future of permafrost requires proper representation of the interrelated surface/subsurface thermal and hydrologic regimes. Land surface models (LSMs) are well suited for such predictions, as they couple heat and water interactions across soil-vegetation-atmosphere interfaces and can be applied over large scales. LSMs, however, are challenged by the long-term thermal and hydraulic memories of permafrost and the paucity of historical records to represent permafrost dynamics under transient climate conditions. In this study, we aim to understand better how LSMs function under different spin-up states, which facilitates addressing the challenge of model initialization by characterizing the impact of initial climate conditions and initial soil frozen and liquid water contents on the simulation length required to reach equilibrium. Further, we quantify how the uncertainty in model initialization propagates to simulated permafrost dynamics. Modelling experiments are conducted with the Modélisation Environmentale Communautaire—Surface and Hydrology (MESH) framework and its embedded Canadian land surface scheme (CLASS). The study area is in the Liard River basin in the Northwest Territories of Canada with sporadic and discontinuous regions. Results show that uncertainty in model initialization controls various attributes of simulated permafrost, especially the active layer thickness, which could change by 0.5–1.5 m depending on the initial condition chosen. The least number of spin-up cycles is achieved with near field capacity condition, but the number of cycles varies depending on the spin-up year climate. We advise an extended spin-up of 200–1000 cycles to ensure proper model initialization under different climatic conditions and initial soil moisture contents.

Evolutionary coupling is a well investigated phenomenon in software maintenance research and practice. Association rules and two related measures, support and confidence, have been used to identify evolutionary coupling among program entities. However, these measures only emphasize the co-change (i.e., changing together) frequency of entities and cannot determine whether the entities co-evolved by experiencing related changes. Consequently, the approach reports false positives and fails to detect evolutionary coupling among infrequently co-changed entities. We propose a new measure, identifier correspondence (id-correspondence), that quantifies the extent to which changes that occurred to the co-changed entities are related based on identifier similarity. Identifiers are the names given to different program entities such as variables, methods, classes, packages, interfaces, structures, unions etc. We use Dice-Sørensen co-efficient for measuring lexical similarity between the identifiers involved in the changed lines of the co-changed entities. Our investigation on thousands of revisions from nine subject systems covering three programming languages shows that id-correspondence can considerably improve the detection accuracy of evolutionary coupling. It outperforms the existing state-of-the-art evolutionary coupling based techniques with significantly higher recall and F-score in predicting future co-change candidates.

Ice-wedge polygon classification with deep learning: evaluation of hybrid compact polarimetric RADARSAT constellation mission and arcticDEM

Phragmites australis (Cav.) Trin. ex Steudel subspecies australis is one of the worst plant invaders in wetlands of North America. Remote sensing is the most cost-effective method to track its spread given its widespread distribution and rapid colonization rate. We hypothesize that the morphological and/or physiological features associated with different phenological states of Phragmites can influence their reflectance signal and thus affect mapping accuracies. We tested this hypothesis by comparing classification accuracies of cloud-free images acquired by Landsat 7, Landsat 8, and Sentinel 2 at roughly monthly intervals over a calendar year for two wetlands in southern Ontario. We used the Support Vector Machines classification and employed field observations and image acquired from unmanned aerial vehicle (8 cm) to perform accuracy assessments. The highest Phragmites producer’s, user’s, and overall accuracy (96.00, 91.11, and 88.56% respectively) were provided by images acquired in late summer and fall period. During this period, green, Near Infrared, and Short-Wave Infrared bands generated more unique reflectance signals for Phragmites. Both Normalized Difference Vegetation Index and Normalized Difference Water Index showed significant difference between Phragmites and the most confused classes (cattail; Typha latifolia L., and meadow marsh) during the late summer and fall period. Since meadow marsh separated out best from Phragmites and cattail in the February image, we used it to mask the meadow marsh in the July image to reduce confusion. The unique reflectance signal of Phragmites in late summer and fall is likely due to prolonged greenness of Phragmites when compared to other wetland vegetation, large, distinct inflorescence, and the water content of Phragmites during this period.

Streamflow in Western North America (WNA) has been experiencing pronounced changes in terms of volume and timing over the past century, primarily driven by natural climate variability and human-induced climate changes. This thesis advances on previous work by revealing the most recent streamflow changes in WNA using a comprehensive suite of classical hydrometric methods along with novel Deep Learning (DL) based approaches for change detection and classifica- tion. More than 500 natural streams were included in the analysis across western Canada and the United States. Trend analyses based on the Mann-Kendall test were conducted on a wide selection of classic hydrometric indicators to represent varying aspects of streamflow over 43 years from 1979 to 2021. A general geograph- ical divide at approximately 46◦N degrees latitude indicates that total streamflow is increasing to the north while declining to the south. Declining late summer flows (July–September) were also widespread across the WNA domain, coinciding with an overall reduction in precipitation. Some changing patterns are regional specific, including: 1) increased winter low flows at high latitudes; 2) earlier spring freshet in Rocky Mountains; 3) increased autumns flows in coastal Pacific North- west; and 4) dramatic drying in southwestern United States. In addition to classic hydrometrics, trend analysis was performed on Latent Features (LFs), which were extracted by Variation AutoEncoder (VAE) from raw streamflow data and are considered “machine-learned hydrometrics”. Some LFs with direct hydrological implications were closely associated with the classical hydrometric indicators such as flow quantity, seasonal distribution, timing and magnitude of freshet, and snow- to-rain transition. The changing patterns of streamflows revealed by LFs show direct agreement with the hydrometric trends. By reconstructing hydrographs from select LFs, VAE also provides a mechanism to project changes in streamflow patterns in the future. Furthermore, a parametric t-SNE method based on DL technology was developed to visualize similarity among a large number of hydro- graphs on a 2-D map. This novel method allowed fast grouping of hydrologically similar rivers based on their flow regime type and provides new opportunities for streamflow classification and regionalization.

Umair, M., Melton, J.R., Roy, A., et al. (2023) Implementation of plant hydraulics in the Canadian land surface scheme including biogeochemical cycles (CLASSIC).

Implementation of plant hydraulics in the Canadian land surface scheme including biogeochemical cycles (CLASSIC)

Implications during transitional periods of improvements to the snow processes in the land surface scheme – hydrological model WATCLASS

The collection efficiency of a typical precipitation gauge-shield configuration decreases with increasing wind speed, with a high scatter for a given wind speed. The high scatter in the collection efficiency for a given wind speed arises in part from the variability in the characteristics of falling snow and atmospheric turbulence. This study uses weighing gauge data collected at the Marshall Field Site near Boulder, Colorado, during the WMO Solid Precipitation Intercomparison Experiment (SPICE). Particle diameter and fall speed data from a laser disdrometer were used to show that the scatter in the collection efficiency can be reduced by considering the fall speed of solid precipitation particles. The collection efficiency was divided into two classes depending on the measured mean-event particle fall speed during precipitation events. Slower-falling particles were associated with a lower collection efficiency. A new transfer function (i.e., the relationship between collection efficiency and other meteorological variables, such as wind speed or air temperature) that includes the fall speed of the hydrometeors was developed. The root-mean-square error of the adjusted precipitation with the new transfer function with respect to a weighing gauge placed in a double fence intercomparison reference was lower than using previously developed transfer functions that only consider wind speed and air temperature. This shows that the measured fall speed of solid precipitation with a laser disdrometer accounts for a large amount of the observed scatter in weighing gauge collection efficiency.

Conventional assessment and evaluation of sediment quality are based on laboratory-based ecotoxicological and chemical measurements with lack of concern for ecological relevance. Microbiotas in sediment are responsive to pollutants and can be used as alternative ecological indicators of sediment pollutants; however, the linkage between the microbial ecology and ecotoxicological endpoints in response to sediment contamination has been poorly evaluated. Here, in situ microbiotas from the Three Gorges Reservoir (TGR) area of the Yangtze River were characterized by DNA metabarcoding approaches, and then, changes of in situ microbiotas were compared with the ecotoxicological endpoint, aryl hydrocarbon receptor (AhR) mediated activity, and level of polycyclic aromatic hydrocarbons (PAHs) in sediments. PAHs and organic pollutant mixtures mediating AhR activity had different effects on the structures of microbiotas. Specifically, Shannon indices of protistan communities were negatively correlated with the levels of AhR mediated activity and PAHs. The sediment AhR activity was positively correlated with the relative abundance of prokaryotic Acetobacteraceae, but had a negative correlation with protistan Oxytrichidae. Furthermore, a quantitative classification model was built to predict the level of AhR activity based on the relative abundances of Acetobacteraceae and Oxytrichidae. These results suggested that in situ Protista communities could provide a useful tool for monitoring and assessing ecological stressors. The observed responses of microbial community provided supplementary evidence to support that the AhR-active pollutants, such as PAHs, were the primary stressors of the aquatic community in TGR area.

With the increasing availability of SAR imagery in recent years, more research is being conducted using deep learning (DL) for the classification of ice and open water; however, ice and open water classification using conventional DL methods such as convolutional neural networks (CNNs) is not yet accurate enough to replace manual analysis for operational ice chart mapping. Understanding the uncertainties associated with CNN model predictions can help to quantify errors and, therefore, guide efforts on potential enhancements using more–advanced DL models and/or synergistic approaches. This paper evaluates an approach for estimating the aleatoric uncertainty [a measure used to identify the noise inherent in data] of CNN probabilities to map ice and open water with a custom loss function applied to RADARSAT–2 HH and HV observations. The images were acquired during the 2014 ice season of Lake Erie and Lake Ontario, two of the five Laurentian Great Lakes of North America. Operational image analysis charts from the Canadian Ice Service (CIS), which are based on visual interpretation of SAR imagery, are used to provide training and testing labels for the CNN model and to evaluate the accuracy of the model predictions. Bathymetry, as a variable that has an impact on the ice regime of lakes, was also incorporated during model training in supplementary experiments. Adding aleatoric loss and bathymetry information improved the accuracy of mapping water and ice. Results are evaluated quantitatively (accuracy metrics) and qualitatively (visual comparisons). Ice and open water scores were improved in some sections of the lakes by using aleatoric loss and including bathymetry. In Lake Erie, the ice score was improved by ∼2 on average in the shallow near–shore zone as a result of better mapping of dark ice (low backscatter) in the western basin. As for Lake Ontario, the open water score was improved by ∼6 on average in the deepest profundal off–shore zone.

Abstract: Herbicides and safeners have been formulated together to help protect crop plants from the injurious effects of herbicides while maintaining the ability of the herbicides to selectively remove targeted weeds. These groups of compounds became important as agriculture increased to sustain the world's ever-increasing population. Selected emerging safeners (SESs), Mefenpyr di-ethyl (MEF) and Cyprosulfamide (CPS) and their co-herbicides (Fenoxaprop-p-ethyl (FEN) and Isoxaflutole (ISO)) used during pre- and post-emergence of cereals and grains were used for this study. The mobility of safeners varied in the environment in relation to their chemical properties, such as octanol-water solubility partition coefficient. Herbicides and safeners have been found in the aquatic environment because they could dissipate from the point of application through leaching, surface runoff, and volatilization. The sorption of safeners such as CPS to soil was found to be governed by soil pH. Hence, their ability to leach varies for soils depending on their acidity/alkalinity. Safeners are classified as inert for regulatory purposes because they mostly act by upregulating detoxifying enzymes. Despite their presence in surface water and their mobility, there is limited or no data regarding their toxicity to non-target organisms in aquatic environments. Thus, the study focused on evaluating and assessing the toxicity of the SESs on non-target organisms such as Daphnia magna-a cladoceran invertebrate, and Zebrafish (D. rerio)- an aquatic vertebrate animal model. Specifically, the primary objectives of this study were to 1) evaluate and further expand our knowledge on the toxicity of selected safeners for which little toxicological knowledge regarding non-target species like daphnids and fish exist; 2) to evaluate these chemicals singly and in a mixture in an in vivo approach that helps to assess acute, sublethal and chronic effects on the selected model aquatic animals; 3) use simulation models of molecular docking to understand further the interaction of SESs and their herbicides with respect to their binding affinity and predict toxicity on receptors such as growth receptor (4XNN)and hatching enzymes (ZHE1). To address the information gap regarding toxicity of selected safeners, in vivo studies and biochemical assays were done to understand possible reactions in the system of the animals after exposure to serial concentrations of SESs in acute and chronic water-borne exposures. Results from the study showed sublethal and lethal effects on both organisms exposed to MEF and FEN singly and mixed at environmentally relevant concentrations. Lower concentrations of these chemicals caused deformities and inhibited hatching rate, while such effects were not observed for CPS and ISO at the same concentrations. MEF was classified as moderate to high-risk level according to ADMET software. On the contrary, CPS had low to moderate risk levels for both D. magna and D. rerio. A biphasic plot indicating a hormetic reaction was obtained for reproduction in D. magna, suggesting that MEF might be an endocrine-disrupting chemical and induce stress. Lower concentrations (< 3 mg/L) of MEF caused deformations and inhibited the hatching rate in D. rerio embryos. Mixture studies showed infra-additive reactions on endpoints, including survival and hatching rate. Biochemical activities showed a downregulation in GST in some of the chemicals and an inhibition in SOD activities, the organism is mounting an antioxidant response to maintain homeostasis. Molecular docking scores suggested that MEF, CPS, ISO and FEN all binds to the studied receptors (growth (4XNN) and hatching (ZHE1)). Therefore, MEF and CPS have potentials to affect the activities of the selected receptors. Likewise, toxicity estimation software (TEST) predicted MEF and FEN to be developmental toxicants, and that FEN was predicted to have mutagenic potential using consensus method to calculate the end point. While MEF was more potent singly, its toxicity was reduced when in a mixture with another toxic compound (FEN), and this is beneficial to the organism. In the same vein, CPS, which has low potency, also suppressed the toxicity of ISO. The key findings from this study were that even if SESs contributed to mitigating the effects of herbicides on the animal models, there are sublethal effects associated with it, and the adverse outcome is still a cause for concern. Also, safeners categorized as inert should be re-evaluated to account for all their toxic potential.

During the past few decades, contamination of sediments by persistent toxic substances (PTSs) has been observed in estuarine and coastal areas on the west coast of South Korea. The contaminants are suspected to cause toxicities in aquatic biota, but little is known about their ecological effects, particularly on benthic microbial communities. In this study, an eDNA-based assessment was applied along with classic assessments of exposure, such as chemistry and in vitro bioassays, to evaluate condition of benthic bacterial communities subjected to PTSs. Two strategies were adopted for the study. One was to conduct a comprehensive assessment in space (by comparing seawater and freshwater sites at five coastal regions) and in time (by following change over a 5-y period). Although we found that bacterial composition varied among and within years, some phyla, such as Proteobacteria (28.7%), Actinobacteria (13.1%), Firmicutes (12.7%), and Chloroflexi (12.5%) were consistently dominated across the study regions. Certain bacterial groups, such as Firmicutes and Verrucomicrobia have been linked to contamination at some sites in the study area and at specific points in time. Bacterial communities were not significantly correlated with salinity or AhR- and ER-mediated potencies, whereas concentrations of PAHs, APs, and certain metals (Cd and Hg) exhibited significant associations to the structure of bacterial communities at the phylum level. In fact, the relative abundance of microbes in the phylum Planctomycetes was significantly and negatively correlated with concentrations of PAHs and metals. Thus, the relative abundance of Planctomycetes could be used as an indicator of sedimentary contamination by PAHs and/or metals. Based on our correlation analyses, Cd and ER-mediated potencies were associated more with bacterial abundances at the class taxonomic level than were other PTSs and metals. Overall, the eDNA-based assessment was useful by augmenting more traditional measures of exposure and responses in a sediment triad approach and has potential as a more rapid screening tool.

Talus and moraine serve as vital groundwater reservoirs, with their associated capability to modulate the baseflow of major rivers that originate in headwater environments. Recent field-based studies conducted in the Canadian Rocky Mountains have identified a nonlinear storage-discharge relationship expressed by these surficial alpine aquifers and the importance of their spatial positioning and extent in headwater environments. However, few studies have tried to upscale our current small-scale understanding of these surficial units, to better understand how their storage-discharge dynamics influence basin-scale (i.e. 10^3-10^4 km^2) hydrology. This study aimed to develop a means to integrate several representative features associated with alpine aquifers into a basin-scale hydrological model to potentially improve its capability to estimate and predict the baseflow of mountain rivers. Specifically, this study developed a simple object-oriented image classification workflow to map the spatial extent and distribution of talus and moraine, among other alpine landcover, in addition to validating the capability of a previously discerned simple exponential function to emulate the aforementioned groundwater storage-discharge relationship expressed by alpine aquifers. The resulting object-oriented workflow did well to capture the spatial variability of the aquifers present in the 2228 km^2 Upper Bow River Basin in Alberta Canada, yielding an overall accuracy rating of 90%, while providing an efficient means to extract the aquifer’s spatial characteristics. The exponential function was then tested in a small first-order watershed in the Canadian Rocky Mountains and simulated watershed’s groundwater storage-discharge dynamics to a similar accuracy compared to a distributed physically-based groundwater flow model implemented in the same area. This suggests that the function potentially has the capacity to be integrated as the new baseflow function in a basin-scale hydrologic model and likely improve its capability to accurately estimate and predict the baseflow of mountain rivers. The underlying framework of the Modélisation Environnementale communautaire (MEC) - Surface Hydrology (MESH) model, a basin-scale hydrological model widely implemented in alpine regions in Canada, is presented to demonstrate how these representative features associated with alpine aquifers could be integrated into such a model.

Integration of surrogate optimization and pca for calibration of hydrologic models, a WATCLASS case study

Groundwater vulnerability assessments, often presented in the form of a thematic map, provide a measure of the relative susceptibility of a groundwater system to contamination introduced at or near the ground surface. However, most groundwater vulnerability assessments rely on deterministic, point estimates based on averaged input parameters, and result in a single output value without any indication of the uncertainty or variation around this value. To facilitate the most effective application and interpretation of groundwater vulnerability assessments, a method for incorporating the uncertainty associated with the natural variation of input parameters into groundwater vulnerability assessments was developed and demonstrated in south-central Saskatchewan. A comprehensive literature review and synthesis, including a review of the conceptual basis of intrinsic groundwater vulnerability assessment methods, a critical evaluation of common and representative methods, and a review of the current research in the field illustrated opportunities for extending the application of these methods as decision support tools. A modified, depth-defined Aquifer Vulnerability Index (AVI) method was developed based on statistically derived, depth-defined hydraulic conductivity distributions generated from hydraulic conductivity data for the Pleistocene-aged glacial till aquitards of the Interior Plains region of Saskatchewan. This modified AVI method was used to produce three sets of vulnerability indices based on the range of probable hydraulic conductivity values, allowing for the pseudo-quantitative assessment of the uncertainty associated with the variability of the input parameter. A final vulnerability map was produced showing the mean (expected) AVI value with an overlay indicating areas of elevated uncertainty. Comparisons of the modified AVI method with a classic AVI assessment revealed the impact of geological controls over groundwater vulnerability assessment results. The methods developed for incorporating and presenting uncertainty in groundwater vulnerability assessments are not limited to local applications of the AVI method, but can be applied to any deterministic vulnerability assessment method where a statistical characterization of the input parameters is possible. Furthermore, valuable and accessible reference information presented here in the form of a hydraulic conductivity database, summary tables, and conceptual models will aid in the effective selection, application, and interpretation of intrinsic groundwater vulnerability assessments.

Groundwater vulnerability assessments, often presented in the form of a thematic map, provide a measure of the relative susceptibility of a groundwater system to contamination introduced at or near the ground surface. However, most groundwater vulnerability assessments rely on deterministic, point estimates based on averaged input parameters, and result in a single output value without any indication of the uncertainty or variation around this value. To facilitate the most effective application and interpretation of groundwater vulnerability assessments, a method for incorporating the uncertainty associated with the natural variation of input parameters into groundwater vulnerability assessments was developed and demonstrated in south-central Saskatchewan. A comprehensive literature review and synthesis, including a review of the conceptual basis of intrinsic groundwater vulnerability assessment methods, a critical evaluation of common and representative methods, and a review of the current research in the field illustrated opportunities for extending the application of these methods as decision support tools. A modified, depth-defined Aquifer Vulnerability Index (AVI) method was developed based on statistically derived, depth-defined hydraulic conductivity distributions generated from hydraulic conductivity data for the Pleistocene-aged glacial till aquitards of the Interior Plains region of Saskatchewan. This modified AVI method was used to produce three sets of vulnerability indices based on the range of probable hydraulic conductivity values, allowing for the pseudo-quantitative assessment of the uncertainty associated with the variability of the input parameter. A final vulnerability map was produced showing the mean (expected) AVI value with an overlay indicating areas of elevated uncertainty. Comparisons of the modified AVI method with a classic AVI assessment revealed the impact of geological controls over groundwater vulnerability assessment results. The methods developed for incorporating and presenting uncertainty in groundwater vulnerability assessments are not limited to local applications of the AVI method, but can be applied to any deterministic vulnerability assessment method where a statistical characterization of the input parameters is possible. Furthermore, valuable and accessible reference information presented here in the form of a hydraulic conductivity database, summary tables, and conceptual models will aid in the effective selection, application, and interpretation of intrinsic groundwater vulnerability assessments.

Despite the high historical losses attributed to flood events, Canadian flood mitigation efforts have been hindered by a dearth of current, accessible flood extent/risk models and maps. Such resources often entail large datasets and high computational requirements. This study presents a novel, computationally efficient flood inundation modeling framework (“InundatEd”) using the height above nearest drainage (HAND)-based solution for Manning's equation, implemented in a big-data discrete global grid system (DGGS)-based architecture with a web-GIS (Geographic Information Systems) platform. Specifically, this study aimed to develop, present, and validate InundatEd through binary classification comparisons to recently observed flood events. The framework is divided into multiple swappable modules including GIS pre-processing; regional regression; inundation models; and web-GIS visualization. Extent testing and processing speed results indicate the value of a DGGS-based architecture alongside a simple conceptual inundation model and a dynamic user interface.

While designing a software system, minimizing coupling among program entities (such as files, classes, methods) is always desirable. If a software entity is coupled with many other entities, this might be an indication of poor software design because changing that entity will likely have ripple change effects on the other coupled entities. Evolutionary coupling, also known as change coupling, is a well investigated way of identifying coupling among program entities. Existing studies have investigated whether file level or class level evolutionary couplings are related with software bug-proneness. While these existing studies have mixed findings regarding the relationship between bug-proneness and evolutionary coupling, none of these studies investigated whether method level (i.e., function level for procedural languages) evolutionary coupling is correlated with bug-proneness. Investigation considering a finer granularity (i.e., such as method level granularity) can help us pinpoint which methods in the files or classes are actually responsible for coupling as well as bug-proneness. In this research, we investigate method level evolutionary coupling through mining association rules and analyze whether this coupling is correlated with software bug-proneness. According to our investigation on thousands of commit operations from the evolutionary history of six open-source subject systems, method level evolutionary coupling generally has a good positive correlation with software bug-proneness. Our regression analysis indicates that evolutionary coupling and bug-proneness mostly have a linear relationship between them. We also observe that methods that experience bug-fixes during evolution generally have a significantly higher number of evolutionary coupling links than the methods that do not experience bug-fixes. We realize that minimizing method level evolutionary coupling links can help us minimize bugs in software systems. Our prototype tool is capable of identifying highly coupled methods along with their coupling links so that programmers can find possibilities of minimizing those links for reducing bug-proneness of software systems.

Assessment of remote sensing derived freeze/thaw products from L-band radiometry requires ground validation. There is growing interest in utilizing soil moisture networks to meet this validation requirement, although it remains unclear whether the current configuration of these networks is appropriate. To address this issue, a small-scale L-band radiometry study was conducted from November 2014 to April 2015 to capture F/T events. Soil moisture probes measuring soil temperature and real dielectric permittivity were installed within a surface-based L-band radiometer footprint vertically at the surface and horizontally at 2.5, 5 and 10 cm depths. A binary freeze/thaw product was derived using radiometer brightness temperatures and compared to the binary F/T classification using soil temperature, real dielectric permittivity and air temperature measurements. The results of the study found that shallow probe depths (vertical and 2.5 cm) resulted in an improvement in the radiometer F/T product accuracy over the standard 5 cm instrument depth currently utilized at soil moisture networks. Air temperature based approaches for validation result in lower accuracy for F/T events with no snow or wet snow, but performed similarly to soil measurements of temperature and permittivity during dry snow F/T events.

The large stocks of soil organic carbon (SOC) in soils and deposits of the northern permafrost region are sensitive to global warming and permafrost thawing. The potential release of this carbon (C) as greenhouse gases to the atmosphere does not only depend on the total quantity of soil organic matter (SOM) affected by warming and thawing, but it also depends on its lability (i.e., the rate at which it will decay). In this study we develop a simple and robust classification scheme of SOM lability for the main types of soils and deposits in the northern permafrost region. The classification is based on widely available soil geochemical parameters and landscape unit classes, which makes it useful for upscaling to the entire northern permafrost region. We have analyzed the relationship between C content and C-CO2 production rates of soil samples in two different types of laboratory incubation experiments. In one experiment, ca. 240 soil samples from four study areas were incubated using the same protocol (at 5 ∘C, aerobically) over a period of 1 year. Here we present C release rates measured on day 343 of incubation. These long-term results are compared to those obtained from short-term incubations of ca. 1000 samples (at 12 ∘C, aerobically) from an additional three study areas. In these experiments, C-CO2 production rates were measured over the first 4 d of incubation. We have focused our analyses on the relationship between C-CO2 production per gram dry weight per day (µgC-CO2 gdw−1 d−1) and C content (%C of dry weight) in the samples, but we show that relationships are consistent when using C ∕ N ratios or different production units such as µgC per gram soil C per day (µgC-CO2 gC−1 d−1) or per cm3 of soil per day (µgC-CO2 cm−3 d−1). C content of the samples is positively correlated to C-CO2 production rates but explains less than 50 % of the observed variability when the full datasets are considered. A partitioning of the data into landscape units greatly reduces variance and provides consistent results between incubation experiments. These results indicate that relative SOM lability decreases in the order of Late Holocene eolian deposits to alluvial deposits and mineral soils (including peaty wetlands) to Pleistocene yedoma deposits to C-enriched pockets in cryoturbated soils to peat deposits. Thus, three of the most important SOC storage classes in the northern permafrost region (yedoma, cryoturbated soils and peatlands) show low relative SOM lability. Previous research has suggested that SOM in these pools is relatively undecomposed, and the reasons for the observed low rates of decomposition in our experiments need urgent attention if we want to better constrain the magnitude of the thawing permafrost carbon feedback on global warming.

Lability classification of soil organic matter in the northern permafrost region

Harmful Algal Bloom (HABs) reports have increased globally, where climate change is considered a primary driver. While the role of temperature and precipitation on bloom formation is well understood on HAB formation, lake ice remains understudied. Reduced ice duration periods may alter algal growth, extent, duration and timing due to earlier light penetration, shifts in mixing, and changes to thermal regimes. Northern lakes are at an elevated risk due to the greater rate of air temperature change at high latitudes. To determine the importance of lake ice in the prediction of HABs, an observational time series (2002-2019) was analyzed utilizing new remote sensing data products provided by the ESA CCI Lakes+ project to determine the potential link between lake ice (lake ice on-off DOY and lake ice duration), lake surface water temperature (LSWT; mean/max, peak LSWT DOY, LSWT anomaly days) and algal biomass parameters (mean and max chlorophyll-a (chl-a, a proxy of algal biomass), high chl-a extent, duration and peak DOY) for five North American Great Lakes (Lake Erie, Lake Winnipeg, Lake Athabasca, Great Slave Lake, and Great Bear Lake). A mean reduction in lake ice duration of ~0.47-0.57 days per year for northern lakes, with an increase of ~1.05-1.12 µg L-1/yr in mean chl-a concentrations was found. Multiple Linear Regression (MLR) tests were conducted with varying combinations of inputs. Artificial Neural Networks (ANN) were implemented to determine if non-linear functions provided a better predictive performance. The MLR found that LSWT had a greater importance in the prediction of algal biomass parameters, while the ANN provided a stronger predictive performance overall. Northern lakes had relatively lower predictive error (median NRMSE = 0.72) using the ANN compared to that of the southern lakes (median NRMSE = 0.81). A random forest (RF) model was used to classify annual/seasonal algal bloom pixels using the lake ice and LSWT parameters (accuracy = 83.84% - 95.47%). The preliminary results indicate that the LSWT parameters had the highest reduction in mean accuracy when excluded from the annual RF, however when predicting HABs during early months (March – May), lake ice parameters typically had a higher importance. Through this analysis, ice parameter thresholds can be established to better understand its impact on algal biomass. This research has found that northern lakes typically had better predictive performance when using lake ice and LSWT parameters, and that lake ice parameters show a high importance in the classification of spring HABs.

Changes to ice cover on lakes throughout the northern landscape has been established as an indicator of climate change and variability, expected to have implications for both human and environmental systems. Monitoring lake ice cover is also required to enable more reliable weather forecasting across lake-rich northern latitudes. Currently, the Canadian Ice Service (CIS) monitors lakes using synthetic aperture radar (SAR) and optical imagery through visual interpretation, with total lake ice cover reported weekly as a fraction out of ten. An automated method of classification would allow for more detailed records to be delivered operationally. In this research, we present an automatic ice-mapping approach which integrates unsupervised segmentation from the Iterative Region Growing using Semantics (IRGS) algorithm with supervised random forest (RF) labeling. IRGS first locally segments homogeneous regions in an image, then merges similar regions into classes across the entire scene. Recently, these output regions were manually labeled by the user to generate ice maps, or were labeled using a Support Vector Machine (SVM) classifier. Here, three labeling methods (Manual, SVM, and RF) are applied after IRGS segmentation to perform ice-water classification on 36 RADARSAT-2 scenes of Great Bear Lake (Canada). SVM and RF classifiers are also tested without integration with IRGS. An accuracy assessment has been performed on the results, comparing outcomes with author-generated reference data, as well as the reported ice fraction from CIS. The IRGS-RF average classification accuracy for this dataset is 95.8%, demonstrating the potential of this automated method to provide detailed and reliable lake ice cover information operationally.

Lake Ice-Water Classification of RADARSAT-2 Images by Integrating IRGS Segmentation with Pixel-Based Random Forest Labeling

In this study we assess the total storage, landscape distribution, and vertical partitioning of soil organic carbon (SOC) stocks on the Brøgger Peninsula, Svalbard. This type of high Arctic area is underrepresented in SOC databases for the northern permafrost region. Physico-chemical, elemental, and radiocarbon (14C) dating analyses were carried out on thirty-two soil profiles. Results were upscaled using both a land cover classification (LCC) and a landform classification (LFC). Both LCC and LFC approaches provide weighted mean SOC 0–100 cm estimates for the study area of 1.0 ± 0.3 kg C m−2 (95% confidence interval) and indicate that about 68 percent of the total SOC storage occurs in the upper 30 cm of the soil, and about 10 percent occurs in the surface organic layer. Furthermore, LCC and LFC upscaling approaches provide similar spatial SOC allocation estimates and emphasize the dominant role of “vegetated area” (4.2 ± 1.6 kg C m−2) and “solifluction slopes” (6.7 ± 3.6 kg C m−2) in SOC 0–100 cm storage. LCC and LFC approaches report different and complementary information on the dominant processes controlling the spatial and vertical distribution of SOC in the landscape. There is no evidence for any significant SOC storage in the permafrost layer. We hypothesize, therefore, that the Brøgger Peninsula and similar areas of the high Arctic will become net carbon sinks, providing negative feedback on global warming in the future. The surface area that will have vegetation cover and incipient soil development will expand, whereas only small amounts of organic matter will experience increased decomposition due to active-layer deepening.

Warming has led to widespread impacts on the landscape in northern ecosystems. Areas underlain by discontinuous permafrost are as highly susceptible to significant change as a result of permafrost thaw leading to ground subsidence. The Baker Creek watershed is a typical Taiga Shield landscape on permafrost, just north of Yellowknife in the Northwest Territories. In this region, bedrock outcrops are interspersed with lakes, wetlands, soil-filled lowlands, and forests. Permafrost is discontinuous across the landscape – while bedrock and the terrain beneath lakes and streams are permafrost-free, permafrost can be found in forested parts of the landscape, usually associated with peatlands and glaciolacustrine sediments. Across this landscape, we took intensive ground-based measurements characterizing land cover types by their depth of organic soil and frost table. Using archival aerial photographs and recent satellite imagery, we assessed changes in land cover between 1972 and 2017. Strong associations between different land cover classes and the ground-based measurements in these sites allowed us to estimate change in permafrost extent using changes in land cover types as a proxy. While there is evidence of both development and thaw of permafrost within the Baker Creek watershed, our analysis suggests that there is net permafrost gain – a stark contrast with the patterns of change typically reported in discontinuous permafrost landscapes. We found that the aggradation we observed seems to be driven by a combination of local hydrology and climatic 'trigger years' that lead to colder, drier conditions favourable for the development of permafrost.

Deep neural networks are prevalent in supervised learning for large amounts of tasks such as image classification, machine translation and even scientific discovery. Their success is often at the sacrifice of interpretability and generalizability. The increasing complexity of models and involvement of the pre-training process make the inexplicability more imminent. The outstanding performance when labeled data are abundant while prone to overfit when labeled data are limited demonstrates the difficulty of deep neural networks' generalizability to different datasets. This thesis aims to improve interpretability and generalizability by restricting representations. We choose to approach interpretability by focusing on attribution analysis to understand which features contribute to prediction on BERT, and to approach generalizability by focusing on effective methods in a low-data regime. We consider two strategies of restricting representations: (1) adding bottleneck, and (2) introducing compression. Given input x, suppose we want to learn y with the latent representation z (i.e. x→z→y), adding bottleneck means adding function R such that L(R(z)) < L(z) and introducing compression means adding function R so that L(R(y)) < L(y) where L refers to the number of bits. In other words, the restriction is added either in the middle of the pipeline or at the end of it. We first introduce how adding information bottleneck can help attribution analysis and apply it to investigate BERT's behavior on text classification in Chapter 3. We then extend this attribution method to analyze passage reranking in Chapter 4, where we conduct a detailed analysis to understand cross-layer and cross-passage behavior. Adding bottleneck can not only provide insight to understand deep neural networks but can also be used to increase generalizability. In Chapter 5, we demonstrate the equivalence between adding bottleneck and doing neural compression. We then leverage this finding with a framework called Non-Parametric learning by Compression with Latent Variables (NPC-LV), and show how optimizing neural compressors can be used in the non-parametric image classification with few labeled data. To further investigate how compression alone helps non-parametric learning without latent variables (NPC), we carry out experiments with a universal compressor gzip on text classification in Chapter 6. In Chapter 7, we elucidate methods of adopting the perspective of doing compression but without the actual process of compression using T5. Using experimental results in passage reranking, we show that our method is highly effective in a low-data regime when only one thousand query-passage pairs are available. In addition to the weakly supervised scenario, we also extend our method to large language models like GPT under almost no supervision --- in one-shot and zero-shot settings. The experiments show that without extra parameters or in-context learning, GPT can be used for semantic similarity, text classification, and text ranking and outperform strong baselines, which is presented in Chapter 8. The thesis proposes to tackle two big challenges in machine learning --- "interpretability" and "generalizability" through restricting representation. We provide both theoretical derivation and empirical results to show the effectiveness of using information-theoretic approaches. We not only design new algorithms but also provide numerous insights on why and how "compression" is so important in understanding deep neural networks and improving generalizability.

Climate-driven permafrost degradation and an intensification of the hydrological cycle are rapidly altering the intricate ecohydrological processes of Arctic wetlands, threatening their long-term carbon sequestration capabilities. Addressing this concern through effective management holds immense potential for climate regulation, mitigation, and adaptation efforts. As such, there is growing need for timely spatial inventory data identifying Arctic wetlands with sufficient accuracy, resolution, and detail. Wetland mapping at large scales necessitates the processing of large volumes of Earth observation (EO) data, a challenge known as “Big Data”. Consequently, in this study, we present a cloud-based methodology exploiting the remarkable collection of EO data and computational power of Google Earth Engine (GEE) to map Arctic wetlands at 10 m spatial resolution. Our workflow evaluated temporally aggregated optical and radar satellite imagery and novel hydro-physiographic layers as inputs into a robust Random Forest (RF) machine learning (ML) algorithm. Both pixel and object-based classification approaches were assessed, whereby ML models were calibrated with a training dataset of sufficient and comprehensive samples. The study was conducted over Canada’s Southern Arctic ecozone (830,000 km2). GEE enabled the efficient preprocessing and classification of large volumes of EO data and resulted in excellent yet similar statistical performance for both pixel and object-based approaches, achieving overall accuracies of > 89 % and mean F1-scores of > 0.79. Moreover, McNemar tests indicated that these classifications were not statistically different, which has significant implications regarding computing time and processing efficiencies. These results demonstrate the efficacy and scalability of our cloud-based GEE methodology, and as such can support future endeavors around Pan-Arctic wetland mapping and monitoring.

The shoreline of Western Lake Ontario (WLO) comprises ecologically important wetlands and habitats for aquatic life. It is also borders to the Greater Toronto Area (GTA), one of Canada's largest and most densely populated regions. The WLO receives urban stormwater runoff and effluent from treatment plants that deliver nutrients and sediments to the nearshore zone. The WLO is also threatened by invasive species, including dreissenid mussels that are modifying biogeochemical nutrient cycling in the lake. Therefore, the potential re-eutrophication of Lake Ontario remains a major binational concern. In this study, we focus on reconstructing the eutrophication trajectory in WLO to provide essential knowledge to inform mitigation climate-adaptive strategies to protect the lake. We are using chlorophyll-a (Chl-a) concentration data obtained from remote sensing measurements as an indicator of phytoplankton biomass. One particular challenge in the nearshore waters of WLO is the presence of Cladophora, a submerged nuisance algae. We are currently assembling remote sensing and in situ data time series on surface Chl-a concentrations and submerged aquatic vegetation (SAV) coverage for WLO. Several retrieval algorithms for remote sensing Chl-a in the littoral zone of WLO are being compared and modified. To determine SAV coverage, we are developing a supervised classification approach applied to the available but limited in situ data. For the calibration and validation of the algorithms, we are relying on satellite and drone images (including ETM+, OLI and MSI sensors of Landsat 7,8 and Sentinel 2 satellites) taken over a period of 22 years (from 2000 to 2022). The in-situ dataset does not show a significant trend in Chl-a concentrations in the central WLO, which remains below the oligotrophic threshold (2.6 µg L-1). However, on the nearshore, there are hotspots, such as Hamilton Harbour, that experience eutrophic (20 µg L-1 < Chl-a < 56 µg L-1) or even hypertrophic (Chl-a > 56 µg L-1) conditions. In addition, some of these hotspots exhibit slight increases in Chl-a, emphasizing the importance of long-term monitoring and risk assessments of future blooms in this region.

Background and Aims: Stand-replacing crown fires are the most prevalent type of fire regime in boreal forests in North America. However, a substantial proportion of low-severity fires are found within fire perimeters. Here, we aimed to investigate the effects of low-severity fires on the reproductive potential and seedling recruitment in boreal forest stands in between stand-replacing fire events. Methods: We recorded site and tree characteristics from 149 trees within 12 sites dominated by mature black spruce [Picea mariana (Mill.) B.S.P.] trees in the Northwest Territories, Canada. The presence of fire-scarred trees supported classification of sites as unburned or affected by low-severity fires in recent history. We used non-parametric tests to evaluate differences in site conditions between unburned and low-severity sites. We used linear and additive statistical models to evaluate differences in tree age, size and reproductive traits among unburned trees and trees from low-severity sites. Key Results: The results showed a significantly higher density of dead black spruce trees in low-severity sites and marginally significant higher presence of permafrost. Trees from low-severity fire sites were significantly older, exhibited significantly lower tree growth and showed a tendency towards a higher probability of cone presence and percentage of open cones compared with trees from unburned sites. Surviving fire-scarred trees affected by more recent low-severity fires showed a tendency towards a higher probability of cone presence and cone production. The density of black spruce seedlings increased significantly with recent low-severity fires. Conclusions: Trees in low-severity sites appeared to have escaped mortality from up to three fires, as indicated by fire-scar records and their older ages. Shallow permafrost at low-severity sites might cause lower flammability, allowing areas to act as fire refugia. Low-severity surface fires temporarily enhanced the reproductive capacity of surviving trees and the density of seedlings, probably as a stress response to fire events.

This study developed an integrated biogeochemical and hydrological modelling system by incorporating the latest versions of the nitrogen coupled Canadian Land Surface Scheme-Canadian Terrestrial Ecosystem Model (CLASS-CTEM) into the Modelisation Environmentale Communautaire (MEC) Surface and Hydrology system (MESH), hereafter referred to as MESH-CTEM. The newly developed MESH-CTEM modelling system allows simulations of energy, water, carbon and nitrogen fluxes and their feedbacks on vegetation growth and exploration of impacts of future climatic changes on catchment-scale processes. Performance of the MESH-CTEM system was tested at the Big Creek watershed within Norfolk county, Ontario, Canada, which is a 573 km2 crop-dominated catchment with areas of broadleaf and needleleaf forests, using observed eddy covariance flux, meteorological and hydrological datasets from October 2004 to December 2017 at a grid resolution of 0.02o latitude × 0.02o longitude. MESH-CTEM showed a significant increase in the simulated streamflow as compared to MESH running with only CLASS, excluding dynamic vegetation growth and carbon fluxes, resulting in an overall increase in the accuracy of streamflow with Nash-Sutcliffe Efficiency (NSE) indices of 0.38 and 0.12 respectively. Significant improvements were also seen for each Plant Functional Type (PFT) within the catchment with respect to energy fluxes, evaporation and soil water regimes. Many of these improvements in simulated fluxes were due in part by changes in the canopy conductance formulation, more realistic soil heat and water processes due to the introduction of fine soil layers, inter-grid transfers of water and other spatial components and vegetation cover feedbacks on energy, water and carbon exchanges by using dynamic vegetation growth processes. Simulated averaged gross ecosystem productivity, ecosystem respiration, latent heat flux and sensible heat flux for the entire catchment were respectively 660 g C m−2 yr−1, 640 g C m−2 yr−1, 32.5 W m-2 and 27.1 W m-2. Application and use of MESH-CTEM will help to study the impact of climate change and extreme events on energy, water and carbon fluxes and associated feedbacks at the catchment scale. Additionally, this will help bridge a major gap in hydrologic modelling studies through integration of biogeochemical processes.

Land-atmosphere interactions are commonly quantified using eddy-covariance (EC) equipment. This technique provides fluxes which are attributed to an area-averaged two-dimensional flux footprint. Although source flux heterogeneity is present within these footprints, current EC footprint models are unable to distinguish the variable flux contributions (although they are incorporated into the bulk EC flux). A disaggregation method would increase the value of EC data by allowing users to isolate individual fluxes from features within the flux footprint; furthermore, this may extend the useability of the EC method to more complex terrain of mixed land classifications. It remains unexplored how high-resolution surface energy balance (SEB) models can be used to disaggregate these EC footprints. This thesis presents a SEB workflow using Unoccupied Aerial Vehicles (UAVs) to generate high-resolution patterns of evapotranspiration (ET) to disaggregate EC footprints in a novel disaggregated flux footprint prediction (disFFP) method. This workflow begins by using novel UAV Light Ranging And Detection (LiDAR) techniques to derive detailed maps of canopy height (h), effective leaf area index (LAI_e), and canopy viewing fraction (f_c), consistent with known vegetation patterns and field-average observations (LAI_e RMSE=0.08-0.81 m^2 m^(-2)). It then follows an atmospheric correction (path transmissivity and upward-welling radiance) and an ensemble emissivity adjustment (brightness to radiometric temperature), testing these with UAV thermal data to determine their absolute (magnitude) and relative (spatial pattern) effects on temperature. A modified t-test - considering autocorrelation - showed that a raw brightness temperature has similar spatial patterns (r>.99, p_val≫0.001) to brightness and radiometric temperatures. Combining these thermal and LiDAR UAV inputs with a high-resolution SEB model (HRMET), the model performance was then compared against EC fluxes. It followed that HRMET tended to overestimate latent heat over full canopies when surface-air temperature differences exceed 4-5°C. Overall, HRMET succeeded at replicating EC latent heat flux within RMSE of 79-136 W m^(-2) using raw brightness and ensemble emissivity corrected (observed radiometric) temperatures. The resultant relative ET maps (ET_R) provided a coherent chronology of the changing flux landscape. Furthermore, the ET_R trials using corrected temperatures (brightness, radiometric, and observed radiometric) had similar spatial patterns to those found using just raw brightness temperature (r>.93, p_val≫0.001). This implies that a raw brightness temperature is sufficient for determining relative patterns of evapotranspiration. The next part of the workflow uses ET_R to disaggregate a well-known parameterization of a backwards-Lagrangian flux footprint model. The proposed disaggregation method (disFFP) uses the concept of ET period to describe an interval of time (day scale) where the EC flux environment remains relatively constant (constant-rate cumulative ET rate). ET periods were determined using a piece-wise regression of the cumulative day-over-day EC latent flux rate. Each season was divided into five ET periods and compared with the two other seasons to discover potential metrics for further classifying ET periods. It was found that ET rates were consistent across comparable ET periods, 2-38 % (standard deviation as a percentage of sample mean), and that additional metrics (plant phenology, growing degree day, standard precipitation index) played an important role in characterizing ET periods for inter-seasonal work. Eddy-covariance and UAV data were coupled using climatology footprints of ET periods and the coinciding ET patterns (ET_R and coefficient of variation). The disFFP combination of these two products (footprint-weighted factoring) provided disaggregated footprints (EC bulk ET of 144 mm) that reflected increased contributions (180-200 mm) over high ET areas and diminished values (90-100 mm) over lower ET areas. These preliminary findings present an exciting new opportunity to connect discrete UAV data with continuous EC flux monitoring.

Organic chemical pollutants are delivered to riverine habitats via basin land use and hydrology interactions. Aquatic organisms eventually absorb these substances, where they might have negative consequences. However, our capacity to reliably predict potential future changes in pollutant concentrations is now constrained by information gaps relating to the links between hydrological, chemical, and biological processes. In the South Saskatchewan River, Canada, in the years 2020 and 2021, concentrations of three pesticide classes (organochlorines, organophosphates, and herbicides) in the water, sediments, and fish were examined. Organochlorine pesticides have been prohibited in Canada since the 1970s; however, methoxychlor and lindane were occasionally found in samples of sediment and fish that may have been contaminated in the past. Organophosphate pesticides, with the exception of malathion and parathion, were close to detection limit in both sampling years in all matrices, while neonicotinoids were below detection in all samples. On the other hand, for both sampling years, consistent levels of the herbicides 2,4-D and dicamba were found in water samples from all locations. Concentrations were on average three times higher in 2020, when river discharge was two times greater, possibly pointing to contaminated sediments being disturbed by high flows, or run-off from the nearby watershed. Of the trace metals, copper and zinc concentrations at several sampling locations exceeded standards for sediment quality. About 18% of the water and sediment samples that were examined had mercury concentrations that were above recommended levels. These discoveries fill in the gaps in monitoring datasets and show significant connections between hydrology and chemistry that can be further investigated in computational models to forecast pollutant trends in freshwater systems. Trace metal concentrations were used to model transport and fate in the South Saskatchewan River using an existing model developed for another freshwater system. The iv River Analysis System from the Hydrologic Engineering Center was paired with a well-known 1-D modelling technique (HEC-RAS). The stream transport module for the WASP (Water Quality Analysis Simulation Program), TOXI, can calculate the flow of water, sediment, and dissolved constituents through branching and ponded segments and is integrated with flow routing for free-flow streams, ponded segments, and backwater reaches. Two metals with primarily anthropogenic and geogenic origins were chosen: copper and nickel. The South Saskatchewan River was analysed in 2020 and 2021 at 10 distinct locations, both upstream and downstream of the City of Saskatoon. By comparing model predictions with copper and nickel concentrations obtained earlier, model performance was assessed. The model functioned reasonably well for sediment samples and did a good job of estimating the levels of copper and nickel in water samples. In both the water and sediment sample segments, the model overestimated concentrations. Diffuse pollutant loads were increased to enable the model to work more precisely. This work shows the predictive power of merging WASP-TOXI and HEC-RAS models for the prediction of contaminant loading, even though numerous default parameter values had to be employed because primary historical data was unavailable. This proof-of-concept study will be useful for future research, including studies on the effects of climate change on the quality of water in the Canadian prairies

Modeling of extreme events is important in many scientific fields, including environmental and civil engineering, and impacts and risk assessments. Among available methods, statistical models that allow estimating extremes’ frequency and intensity are regularly used in procedures to anticipate potential changes in extreme events. Extreme value theory provides a theoretical basis for statistical estimation of extreme events using frequency analysis. The challenge in modeling is knowing when to use the block maxima method or the peaks-over-threshold (POT) method. Each has its drawbacks. POT describes the main characteristics of the observed extreme series; the threshold selection is always challenging and might affect the accuracy of the simulated results and the credibility of changes in extreme values. To encompass this challenge, mixture models offer more flexibility to represent samples with nonhomogeneous data. This study presents the gamma generalized Pareto (GGP) mixture model for estimating risk occurrence of hydroclimatic extremes. The model was developed in its general form, whereas the observed hydrometeorological extreme events depend on multidimensional covariates. A maximum likelihood algorithm is proposed to estimate the parameters with a constraint on the shape parameter of the generalized Pareto (GP) distribution. A Monte Carlo (MC) simulation compared the proposed model with the classical POT approach, with a fixed threshold, and the annual maximum series of streamflow. The approach was applied using a daily hydrological data set from an observed station located in the Saint John River at Fort Kent (01AD002), New Brunswick, Canada. The results show a flexibility to model extremes for dependent or nonstationary time series and adequately describes the central part of the observed frequencies, as well as the tails.

The effects of wetland management of drainage and restoration on the hydrology of small basins over the Canadian Prairies, were investigated using a virtual basin modelling approach created through the Cold Regions Hydrological Modelling platform (CRHM) for seven biophysical basin classes that typify the region. The model included wind redistribution of snow, energy balance snowmelt, infiltration to frozen soils, Penman-Monteith actual evapotranspiration, soil and ground water dynamics and fill and spill of internally drained surface depressions that often form wetlands. Wetland loss induced by agricultural drainage was represented by scenarios that progressively reduced the depression area and storage capacity by increments of 10%, and ongoing wetland restoration policies were represented by expanding the existing depressions to historical sizes. Model simulations showed that, on average, both annual total streamflow and maximum daily streamflow had the largest sensitivities to wetland drainage in the pothole pond dominated basins. Every 10% loss of depression area resulted in a 15-22% increase in annual streamflow volume and a 13-18% increase in maximum daily streamflow from the pothole-dominated basins, whilst those increases were only 3-9% in other basin classes. Every 10% gain in depression area by restoration resulted in a 7% decrease in maximum daily streamflow from the pothole till basins, contrasted by a 0.2% decrease from the high elevation grasslands basins. Wetlands restoration represented by a 40% increase in depressional area, close to the historical maximum as indicated in the pothole till basins, reduced annual streamflow volume and peak daily streamflow by 5-36% and 1-28%, respectively, over the Prairies. Wetland restoration strategies of from large to small depressions and from bottom to top of the basin exerted a small influence on annual streamflow volume sensitivity, whilst restoring the wetlands closer to the basin outlet tended to be more effective in attenuating the maximum daily streamflow.

He, Z., Shook, K., Spence, C., Pomeroy, J.W., Whitfield, C. (2022) Modeling the Streamflow Sensitivity to Wetland Drainage and Restoration Over the Canadian Prairies Using a Basin Classification Approach. American Geophysical Union (AGU) Fall Meeting, Chicago, USA, December 12 to December 16, 2022. https://agu.confex.com/agu/fm22/meetingapp.cgi/Paper/1072974

This study investigated the effects of wetland loss caused by agricultural activities on basin streamflow in the Canadian Prairies. At the same time, we studied the extent to which wetland restoration policy could influence streamflow in the Prairies. To do that, we divided the entire Canadian Prairies into seven basin types based on a set of biophysical factors delineating land cover, soil, and topography. Responses of hydrological processes to wetland changes in the classified basins were physically represented in the Cold Regions Hydrological Modelling platform (CRHM). The results quantified the sensitivities of annual total streamflow and peak daily streamflow to per 10% loss or expansion in the wetlands. It is found that restoring wetland closer to the basin outlet is more effective in reducing the volume of peak streamflow. Our results could be used to inform agricultural practices and wetland management policies in the Canadian Prairies where a number of waterfowl and other grassland-associated animals live.

Modeling the Streamflow Sensitivity to Wetland Drainage and Restoration Over the Canadian Prairies Using a Basin Classification Approach

The Upper Columbia and Okanagan River basins are important mountain headwaters to provide biodiversity and ecosystem services and to supply water for hydropower dams and reservoir operations in British Colombia, Canada, and the northwest United States. However, the impact of forest disturbance by forest harvesting, disease, and wildfire and then recovery, and deglaciation on the basin hydrology has not been thoroughly studied. The impact of largescale forest disturbances in the basin on the magnitude of recent flooding has not been fully investigated. This study aims to develop a methodology to simulate forest disturbance and regrowth in a hydrological land surface model and then use the model to investigate the impact of forest disturbance on the basin hydrology. An enhanced version of MESH (Modélisation Environnementale communautaire - Surface Hydrology) that incorporated mountain hydrology and vector base routing was setup over a total of 2177 model sub-basins; 1822 for the Upper Columbia and 355 for the Okanagan and Similkameen River basins. Subbasin areas range from 0.005 to 366 km2. The Global Multiscale Model (GEM) with precipitation replaced by the Canadian Precipitation Analysis (CaPA) (~10 km), version RDRS v2 meteorological forcings were used to drive the model. As forest and glaciers are important to the hydrology of the study basin, a particular attention was given in the basin discretization to these. Glaciers were grouped into three classes based on elevation band: high, mid, and low elevation glacier and the albedo value of the three classes were separately computed from the available MODIS albedo remote sensing data. Forests were segregated into four classes based on their species into Hemlock, Fir, Pine, and Spruce. In addition, forest harvesting, regrowth and wildfire were treated separately from mature forest. The wildfire areas were variable from year to year but parameterized in a similar way as barren land. However, forest harvesting, and regrowth were further segregated into three groups by age as Fresh Clear-cut, Clear-cut up to 5 years old and Clear-cut aged 5 years and above and parametrized using MODIS LAI data. The methodology developed to model forest disturbance and regrowth is innovative in a continental scale hydrological model. The preliminary results of the forest disturbance and regrowth model in MESH will be demonstrated.

This study evaluated the effects of climate perturbations on snowmelt, soil moisture, and streamflow generation in small Canadian Prairies basins using a modelling approach based on classification of basin biophysical characteristics. Seven basin classes that encompass the entirety of the Prairies Ecozone in Canada were determined by cluster analysis of these characteristics. Individual semi-distributed virtual basin (VB) models representing these classes were parameterized in the Cold Regions Hydrological Model (CRHM) platform, which includes modules for snowmelt and sublimation, soil freezing and thawing, actual evapotranspiration (ET), soil moisture dynamics, groundwater recharge, and depressional storage dynamics including fill and spill runoff generation and variable connected areas. Precipitation (P) and temperature (T) perturbation scenarios covering the range of climate model predictions for the 21st century were used to evaluate climate sensitivity of hydrological processes in individual land cover and basin types across the Prairies Ecozone. Results indicated that snow accumulation in wetlands had a greater sensitivity to P and T than that in croplands and grasslands in all basin types. Wetland soil moisture was also more sensitive to T than the cropland and grassland soil moisture. Jointly influenced by land cover distribution and local climate, basin-average snow accumulation was more sensitive to T in the drier and grassland-characterized basins than in the wetter basins dominated by cropland, whilst basin-average soil moisture was most sensitive to T and P perturbations in basins typified by pothole depressions and broad river valleys. Annual streamflow had the greatest sensitivities to T and P in the dry and poorly connected Interior Grasslands (See Fig. 1) basins but the smallest in the wet and well-connected Southern Manitoba basins. The ability of P to compensate for warming-induced reductions in snow accumulation and streamflow was much higher in the wetter and cropland-dominated basins than in the drier and grassland-characterized basins, whilst decreases in cropland soil moisture induced by the maximum expected warming of 6 ∘C could be fully offset by a P increase of 11 % in all basins. These results can be used to (1) identify locations which had the largest hydrological sensitivities to changing climate and (2) diagnose underlying processes responsible for hydrological responses to expected climate change. Variations of hydrological sensitivity in land cover and basin types suggest that different water management and adaptation methods are needed to address enhanced water stress due to expected climate change in different regions of the Prairies Ecozone.

He, Z., Shook, K., Spence, C., Pomeroy, J. W., and Whitfield, C. (2023) Modelling the regional sensitivity of snowmelt, soil moisture, and streamflow generation to climate over the Canadian Prairies using a basin classification approach, Hydrol. Earth Syst. Sci., 27, 3525–3546, https://doi.org/10.5194/hess-27-3525-2023

Modelling the regional sensitivity of snowmelt, soil moisture, and streamflow generation to climate over the Canadian Prairies using a basin classification approach

We report on the characteristics of precipitation associated with three types of landfalling atmospheric rivers (ARs) over western North America in the winter season from 1980 to 2004. The ARs are classified according to three landfalling regions as southern, middle and northern types. Two main centers of precipitation are associated with the contributions by the ARs: one over Baja California linked to the southern type of the ARs, and the other over Washington State correlated with the northern and middle types of the ARs. ARs are seen to play a dominant role in the occurrences of extreme precipitation events, with a proportionately greater impact on more extreme events. Moisture flux convergence makes the dominant contribution to precipitation when ARs and extreme precipitation occur simultaneously in the studied areas. Moisture flux convergence in these cases is, in turn, dominated by the mean and transient moisture transported by the transient wind, with greater contribution from the latter, which is mainly concentrated in certain areas. The magnitude and direction of vertically integrated vapor transport (IVT) also play a role in determining the amount of precipitation received in the three regions considered. Larger IVT magnitude corresponds to more precipitation, while an IVT direction of about 220° (0° indicating east wind) is most favorable for high precipitation amount, which is especially obvious for the northern type of the ARs.

River ice monitoring is important for hydrological research and water resource management of the Tibetan Plateau but limited by the serious shortage of field observations, and remote sensing can be used as an effective supplementary means for monitoring river ice. However, remote sensing high-altitude river ice is scarce and a basin-scale understanding of river ice is lacking on the Tibetan Plateau. To ascertain the spatial and temporal distribution characteristics of high-altitude river ice at the basin scale, we selected the Babao River basin as the study area, which is a typical river basin located in the northeastern Tibetan Plateau. Utilizing 447 available Landsat images during the river ice period from 1999 to 2018 and the classical normalized difference snow index (NDSI) algorithm, we monitored the river ice in a long time series at the Babao River basin. The average Khat of accuracy validation reached 0.973. The average area of river ice in the river ice period of this basin showed a weak decreasing trend and was negatively correlated with air temperature. We also found that gentle slopes and high elevations are beneficial for the development of river ice. The melting of river ice supplements river discharge in spring. This study is the first to reveal the distribution characteristics and changing trend of river ice at the basin scale on the Tibetan Plateau, and the results provide a reference for river ice research in this region.

Lake Erie’s commercial and recreational walleye fishery is the largest of the Great Lakes, requiring effective management to maintain a sustainable and complex fishery. Lake Erie’s walleye fishery is composed of multiple spawning populations, which presents a management challenge. The movement patterns and recruitment of distinct walleye populations that make up the fishery must be considered by managers to avoid overexploitation and to maintain population diversity. The Grand River walleye population in Lake Erie’s eastern basin is considered a priority for rehabilitation due to blocked access to spawning habitat by a low-head dam and degraded habitat quality. The objectives of this study were to: i) investigate movement patterns of spawning walleye in the Grand River using acoustic telemetry; and, ii) investigate movement and habitat use of young-of-the-year (YOY) walleye in relation to the Dunnville Dam and surrounding habitat segments using stable isotope analysis. Between 2015 and 2018, 267 mature walleye were tracked in the Grand River using acoustic telemetry, and in fall of 2018 144 YOY walleye were sampled from the river via boat-mounted electrofishing. Both male and female mature walleye that were moved upstream of the Dunnville dam were found to actively migrate ~20-40 km up-river to areas with suspected suitable spawning substrate during the spring spawning season. Residence time of walleye above the Dunnville Dam and timing of return migrations suggest that the dam may be acting as an impediment to downstream movement. Of all the walleye tagged, 43% returned to the Grand River during at least one year subsequent to the initial spawning season during which they were tagged, and those that returned were detected at spawning habitat below the Dunnville Dam during March and April. Although differences in YOY walleye stable isotope signatures (carbon and nitrogen) were evident across sampling locations in the Grand River in fall of 2018, YOY walleye were not successfully sampled in 2019 and a description of the trophic baseline was needed to infer YOY walleye movements. Condition of YOY walleye sampled during the fall of 2018 was highest at the river mouth, which may indicate relatively favourable health conditions for YOY walleye at this location. The results of the biotelemetry study suggest that the removal of the Dunnville Dam or the construction of a functional fishway would increase access to potential additional spawning habitat, which may lead to an increase in successful spawning activity for the Grand River walleye population. Future research on YOY walleye in the southern Grand River will be necessary to enhance the understanding of how recruitment and year-class strength is impacted by movement barriers (i.e., Dunnville Dam) and variation in spawning and nursery habitat quality (i.e., abiotic and biotic stressors). Furthermore, additional analyses on mature walleye apparent annual survival and spawning site fidelity probabilities would further inform our understanding of Grand River walleye movement and support walleye management in Lake Erie’s eastern basin.

Unstructured triangular meshes are an efficient and effective landscape representation that are suitable for use in distributed hydrological and land surface models. Their variable spatial resolution provides similar spatial performance to high-resolution structured grids while using only a fraction of the number of elements. Many existing triangulation methods either sacrifice triangle quality to introduce variable resolution or maintain well-formed uniform meshes at the expense of variable triangle resolution. They are also generally constructed to only fulfil topographic constraints. However, distributed hydrological and land surface models require triangles of varying resolution to provide landscape representations that accurately represent the spatial heterogeneity of driving meteorology, physical parameters and process operation in the simulation domain. As such, mesh generators need to constrain the unstructured mesh to not only topography but to other important surface and sub-surface features. This work presents novel multi-objective unstructured mesh generation software that allows mesh generation to be constrained to an arbitrary number of important features while maintaining a variable spatial resolution. Triangle quality is supported as well as a smooth gradation from small to large triangles. Including these additional constraints results in a better representation of spatial heterogeneity than from classic topography-only constraints.

Seasonal measurements of glacier mass balance provide insight into the relation between climate forcing and glacier change. To evaluate the feasibility of using remotely sensed methods to assess seasonal balance, we completed tandem airborne laser scanning (ALS) surveys and field-based glaciological measurements over a 4-year period for six alpine glaciers that lie in the Columbia and Rocky Mountains, near the headwaters of the Columbia River, British Columbia, Canada. We calculated annual geodetic balance using coregistered late summer digital elevation models (DEMs) and distributed estimates of density based on surface classification of ice, snow, and firn surfaces. Winter balance was derived using coregistered late summer and spring DEMs, as well as density measurements from regional snow survey observations and our glaciological measurements. Geodetic summer balance was calculated as the difference between winter and annual balance. Winter mass balance from our glaciological observations averaged 1.95±0.09 m w.e. (meter water equivalent), 4 % larger than those derived from geodetic surveys. Average glaciological summer and annual balance were 3 % smaller and 3 % larger, respectively, than our geodetic estimates. We find that distributing snow, firn, and ice density based on surface classification has a greater influence on geodetic annual mass change than the density values themselves. Our results demonstrate that accurate assessments of seasonal mass change can be produced using ALS over a series of glaciers spanning several mountain ranges. Such agreement over multiple seasons, years, and glaciers demonstrates the ability of high-resolution geodetic methods to increase the number of glaciers where seasonal mass balance can be reliably estimated.

We merge classical extreme value methods to extract high (high temperatures (HT)) and low (low temperatures (LT)) temperatures and form time series having at least one extreme value per year. Observed daily maximum and minimum temperature records are used from 4,797 quality-controlled, global, surface stations over 1970–2019. We assess changes in the magnitude and frequency of extreme temperatures by introducing and applying novel methods that exploit the definition of stationarity. Analysis shows significant increasing (40.6% of the stations) and decreasing (41.1%) trends in the frequency of high and LT, respectively, and increasing trends in both high- and low-temperature values (35.6% and 49.7%). Globally, HT and LT frequencies are increasing and decreasing, respectively, by 0.9% and 1.1% per year, relative to the expected frequencies under the assumption of stationarity. The global mean annual HT and LT magnitudes are increasing by 0.004 and 0.016°C/year compared to the expected ones under stationarity. The results indicate that the assumption of stationarity fails to explain the observed changes. The proposed methods are an alternative approach to classical extreme value methods and a useful tool to reveal changes in extremes in the era of earth-system change.

The Global Environmental Multiscale Model (GEM) is currently in operational use for data assimilation and forecasting at 25–15 km scales; regional 10 km scales over North America; and 2.5 km scales over Canada. To evaluate the GEM model for forecasting applications in Iran, global daily temperature and precipitation outputs of GEM at a 25 km scale were compared to data sets from hydrometeorological stations and the De Martonne climate classification method was used to demarcate climate zones for comparisons. GEM model outputs were compared to observations in each of these zones. The results show good agreement between GEM outputs and measured daily temperatures with Kling-Gupta efficiencies of 0.76 for the arid, 0.71 for the semiarid, and 0.78 for the humid regions. There is also an agreement between GEM outputs and measured annual precipitation with differences of 50% for the arid, 36% for the semiarid, and 15% for the humid region. There is a ~13% systematic difference between the elevation of stations and the average elevation of corresponding GEM grid cells; differences in elevation associated with forcing data sets can be potentially corrected using environmental lapse rates. Compared with hydrometeorological data sets, the GEM model precipitation outputs are less accurate than temperature outputs, and this may influence the accuracy of potential Iranian forecasting operations utilizing GEM. The results of this study provide an understanding of the operation and limitations of the GEM model for climate change and hydro-climatological studies.

Halogenated disinfection by-products (DBPs) are a diverse class of compounds formed during the treatment of drinking water through reactions between natural organic matter (NOM), inorganic precursors such as bromide, and applied disinfectants. Health Canada regulates a handful of DBPs, but there are over 700 unregulated DBPs that have been described and many of these are more toxic than the regulated DBPs. Here, a data-independent precursor isolation and characteristic fragment (DIPC-Frag) method operated on a Q ExactiveTM Hybrid Quadrupole-OrbitrapTM Mass Spectrometer equipped with a UHPLC system was adapted for the detection of brominated and iodinated DBPs (Br-DBPs and I-DBPs) in chlorinated water. Extraction and analytical conditions were optimized, chemometric strategies were applied, and a library of 553 Br-DBPs and 112 I-DBPs was established with structures predicted for the most abundant compounds. As the method exhibited good precision (~15% RSD), it was then used to study trends of formation and temporal trends of unregulated Br-DBPs in a year-long study that sampled raw, clearwell, and finished waters. While most Br-DBPs increased through the treatment process, cluster I Br-DBPs decreased between the clearwell and finished stages, a pattern significantly related to their chemical properties of low O/C and Br/C ratios. Correlation matrices were used to determine if quality parameters of the source waters (e.g. NOM, turbidity, river level, temperature, bromine (Br)) could explain monthly variations of Br-DBPs, but few significant relationships were found. Unexpectedly, total Br increased from 0.013-0.038 mg/L in raw water to 0.04-0.12 mg/L in finished water, which indicated introduction of Br during disinfection. Concentrations of Br in clearwell and finished water were significantly correlated to detection of 34/54 Br-DBPs at α=0.05 and 14/54 Br-DBPs at α=0.001. As few studies have evaluated toxicity of DBPs in mixtures, the next goal of this thesis was to explore temporal changes in whole mixture toxicity and to determine if raw water parameters could predict toxicity of finished water. By use of a 72 h CHO-K1 cytotoxicity assay and an Nrf2/ARE oxidative stress assay, results indicated cytotoxicity was greatest in finished water collected in November and March while oxidative stress was greatest in June and November, both of which could be related to seasonal trends in unregulated Br-DBPs. These toxic endpoints were correlated (R2 = 0.53, p = 7.4x10-3) and three classes of Br-DBPs (Br2, BrCl, S-DBPs) demonstrated significant correlations to both. The greatest predictors of mixture toxicity were concentration of Br and applied doses of chlorine at related stages. These were equally correlated to both cytotoxicity (R2 = 0.43, p = 0.002) and oxidative stress (R2 = 0.67, p = 0.001). This study is the first to explore temporal trends in whole mixture toxicity of DBPs. It is also the first to suggest that the concentration of Br may be a predictor of the occurrence of unregulated Br-DBPs as well as whole mixture toxicity.

Machine learning (ML) is increasingly considered the solution to environmental problems where only limited or no physico-chemical process understanding is available. But when there is a need to provide support for high-stake decisions, where the ability to explain possible solutions is key to their acceptability and legitimacy, ML can come short. Here, we develop a method, rooted in formal sensitivity analysis (SA), that can detect the primary controls on the outputs of ML models. Unlike many common methods for explainable artificial intelligence (XAI), this method can account for complex multi-variate distributional properties of the input-output data, commonly observed with environmental systems. We apply this approach to a suite of ML models that are developed to predict various water quality variables in a pilot-scale experimental pit lake. A critical finding is that subtle alterations in the design of an ML model (such as variations in random seed for initialization, functional class, hyperparameters, or data splitting) can lead to entirely different representational interpretations of the dependence of the outputs on explanatory inputs. Further, models based on different ML families (decision trees, connectionists, or kernels) seem to focus on different aspects of the information provided by data, although displaying similar levels of predictive power. Overall, this underscores the importance of employing ensembles of ML models when explanatory power is sought. Not doing so may compromise the ability of the analysis to deliver robust and reliable predictions, especially when generalizing to conditions beyond the training data.

Arctic and subarctic regions are amongst the most susceptible regions on Earth to global warming and climate change. Understanding and predicting the impact of climate change in these regions require a proper process representation of the interactions between climate, carbon cycle, and hydrology in Earth system models. This study focuses on land surface models (LSMs) that represent the lower boundary condition of general circulation models (GCMs) and regional climate models (RCMs), which simulate climate change evolution at the global and regional scales, respectively. LSMs typically utilize a standard soil configuration with a depth of no more than 4 m, whereas for cold, permafrost regions, field experiments show that attention to deep soil profiles is needed to understand and close the water and energy balances, which are tightly coupled through the phase change. To address this gap, we design and run a series of model experiments with a one-dimensional LSM, called CLASS (Canadian Land Surface Scheme), as embedded in the MESH (Modélisation Environmentale Communautaire – Surface and Hydrology) modelling system, to (1) characterize the effect of soil profile depth under different climate conditions and in the presence of parameter uncertainty; (2) assess the effect of including or excluding the geothermal flux in the LSM at the bottom of the soil column; and (3) develop a methodology for temperature profile initialization in permafrost regions, where the system has an extended memory, by the use of paleo-records and bootstrapping. Our study area is in Norman Wells, Northwest Territories of Canada, where measurements of soil temperature profiles and historical reconstructed climate data are available. Our results demonstrate a dominant role for parameter uncertainty, that is often neglected in LSMs. Considering such high sensitivity to parameter values and dependency on the climate condition, we show that a minimum depth of 20 m is essential to adequately represent the temperature dynamics. We further show that our proposed initialization procedure is effective and robust to uncertainty in paleo-climate reconstructions and that more than 300 years of reconstructed climate time series are needed for proper model initialization.

Permafrost is an important feature of cold-region hydrology, particularly in river basins such as the Mackenzie River basin (MRB), and it needs to be properly represented in hydrological and land surface models (H-LSMs) built into existing Earth system models (ESMs), especially under the unprecedented climate warming trends that have been observed. Higher rates of warming have been reported in high latitudes compared to the global average, resulting in permafrost thaw with wide-ranging implications for hydrology and feedbacks to climate. The current generation of H-LSMs is being improved to simulate permafrost dynamics by allowing deep soil profiles and incorporating organic soils explicitly. Deeper soil profiles have larger hydraulic and thermal memories that require more effort to initialize. This study aims to devise a robust, yet computationally efficient, initialization and parameterization approach applicable to regions where data are scarce and simulations typically require large computational resources. The study further demonstrates an upscaling approach to inform large-scale ESM simulations based on the insights gained by modelling at small scales. We used permafrost observations from three sites along the Mackenzie River valley spanning different permafrost classes to test the validity of the approach. Results show generally good performance in reproducing present-climate permafrost properties at the three sites. The results also emphasize the sensitivity of the simulations to the soil layering scheme used, the depth to bedrock, and the organic soil properties.

Permafrost is an important feature of cold-region hydrology and needs to be properly represented in hydrological and land surface models (H-LSMs), especially under the observed and expected climate warming trends. This study aims to devise a robust, yet computationally efficient, initialization and parameterization approach for permafrost. We used permafrost observations from three sites along the Mackenzie River valley spanning different permafrost classes to test the validity of the approach.

Le réservoir de carbone de sol est un élément clé du cycle global du carbone et donc du système climatique. Les sols et le carbone organique qu'ils contiennent constituent le plus grand réservoir de carbone des écosystèmes terrestres. Ce réservoir est également responsable du stockage d'une grande quantité de carbone prélevé de l'atmosphère par les plantes par la photosynthèse. C'est pourquoi les sols sont considérés comme une stratégie de mitigation viable pour réduire la concentration atmosphérique de CO2 dûe aux émissions globales de CO2 d'origine fossile. Malgré son importance, des incertitudes subsistent quant à la taille du réservoir global de carbone organique de sol et à ses dynamiques. Les modèles de biosphère terrestre sont des outils essentiels pour quantifier et étudier la dynamique du carbone organique de sol. Ces modèles simulent les processus biophysiques et biogéochimiques au sein des écosystèmes et peuvent également simuler le comportement futur du réservoir de carbone organique de sol en utilisant des forçages météorologiques appropriés. Cependant, de grandes incertitudes dans les projections faite par les modèles de biosphère terrestre sur les dynamiques du carbone organique de sol ont été observées, en partie dues au problème de l'équifinalité. Afin d'améliorer notre compréhension de la dynamique du carbone organique de sol, cette recherche visait à optimiser les paramètres du schéma de carbone de sol contenu dans le modèle de schéma canadien de surface terrestre incluant les cycles biogéochimiques (CLASSIC), afin de parvenir à une meilleure représentation de la dynamique du carbone organique de sol. Une analyse de sensibilité globale a été réalisée pour identifier lesquels parmis les 16 paramètres du schéma de carbone de sol, n'affectaient pas la simulation du carbone organique de sol et de la respiration du sol. L'analyse de sensibilité a utilisé trois sites de covariance des turbulences afin de représenter différentes conditions climatiques simulées par le schéma de carbone de sol et d'économiser le coût calculatoire de l'analyse. L'analyse de sensibilité a démontré que certains paramètres du schéma de carbone de sol ne contribuent pas à la variance des simulations du carbone organique de sol et de la respiration du sol. Ce résultat a permis de réduire la dimensionnalité du problème d'optimisation. Ensuite, quatre scénarios d'optimisation ont été élaborés sur la base de l'analyse de sensibilité, chacun utilisant un ensemble de paramètres. Deux fonctions coûts ont été utilisées pour l'optimisation de chacun des scénarios. L'optimisation a également démontré que la fonction coût utilisée avait un impact sur les ensembles de paramètres optimisés. Les ensembles de paramètres obtenus à partir des différents scénarios et fonctions coûts ont été comparés à des ensembles de données indépendants et à des estimations globales du carbone organique de sol à l'aide de métrique tel la racine de l'erreur quadratique moyenne et le bias, afin d'évaluer l'effet des ensembles de paramètres sur les simulations effectuées par le schéma de carbone de sol. Un ensemble de paramètres a surpassé les autres ensembles de paramètres optimisés ainsi que le paramétrage par défaut du modèle. Ce résultat a indiqué que la structure d'optimisation était en mesure de produire un ensemble de paramètres qui simulait des valeurs de carbone organique de sol et de respiration du sol qui étaient plus près des valeurs observées que le modèle CLASSIC par défaut, améliorant la représentation de la dynamique du carbone du sol. Cet ensemble de paramètres optimisés a ensuite été utilisé pour effectuer des simulations futures (2015-2100) de la dynamique du carbone organique de sol afin d'évaluer son impact sur les projections de CLASSIC. Les simulations futures ont montré que l'ensemble de paramètres optimisés simulait une quantité de carbone organique de sol 62 % plus élevée que l'ensemble de paramètres par défaut tout en simulant des flux de respiration du sol similaires. Les simulations futures ont également montré que les ensembles de paramètres optimisés et par défaut prévoyaient que le réservoir de carbone organique de sol demeurerait un puits de carbone net d'ici 2100 avec des sources nettes régionales. Cette étude a amélioré globalement la représentation de la dynamique du carbone organique de sol dans le schéma de carbone de sol de CLASSIC en fournissant un ensemble de paramètres optimisés. Cet ensemble de paramètres devrait permettre d'améliorer notre compréhension de la dynamique du carbone du sol. The soil carbon pool is a vital component of the global carbon cycle and, therefore, the climate system. Soil organic carbon (SOC) is the largest carbon pool in terrestrial ecosystems. This pool stores a large quantity of carbon that plants have removed from the atmosphere through photosynthesis. Because of this, soils are considered a viable climate change mitigation strategy to lower the global atmospheric CO2 concentration that is presently being driven higher by anthropogenic fossil CO2 emissions. Despite its importance, there are still considerable uncertainties around the size of the global SOC pool and its response to changing climate. Terrestrial biosphere models (TBM) simulate the biogeochemical processes within ecosystems and are critical tools to quantify and study SOC dynamics. These models can also simulate the future behavior of SOC if carefully applied and given the proper meteorological forcings. However, TBM predictions of SOC dynamics have high uncertainties due in part to equifinality. To improve our understanding of SOC dynamics, this research optimized the parameters of the soil carbon scheme contained within the Canadian Land Surface Scheme Including Biogeochemical Cycles (CLASSIC), to better represent SOC dynamics. A global sensitivity analysis was performed to identify which of the 16 parameters of the soil carbon scheme did not affect simulated SOC stocks and soil respiration (Rsoil). The sensitivity analysis used observations from three eddy covariance sites for computational efficiency and to encapsulate the climate represented by the global soil carbon scheme. The sensitivity analysis revealed that some parameters of the soil carbon scheme did not contribute to the variance of simulated SOC and Rsoil. These parameters were excluded from the optimization which helped reduce the dimensionality of the optimization problem. Then, four optimization scenarios were created based on the sensitivity analysis, each using a different set of parameters to assess the impact the number of parameters included had on the optimization. Two different loss functions were used in the optimization to assess the impact of accounting for observational error. Comparing the optimal parameters between the optimizations performed using the different loss functions showed that the loss functions impacted the optimized parameter sets. To determine which optimized parameter set obtained by each loss function was most skillful, they were compared to independent data sets and global estimates of SOC, which were not used in the optimization using comparison metrics based on root-mean-square-deviation and bias. This study generated an optimal parameter set that outperformed the default parameterization of the model. This optimal parameter set was then applied in future simulations of SOC dynamics to assess its impact upon CLASSIC's future projections. These future simulations showed that the optimal parameter set simulated future global SOC content 62 % higher than the default parameter set while simulating similar Rsoil fluxes. The future simulations also showed that both the optimized and default parameter sets projected that the SOC pool would be a net sink by 2100 with regional net sources, notably tropical regions.

Gauthier, Charles (2023) Optimisation des paramètres de carbone de sol dans le modèle CLASSIC à l'aide d'optimisation bayésienne et d'observations, Umontreal Papyrus - Thèses et mémoires, https://hdl.handle.net/1866/31992

Optimisation des paramètres de carbone de sol dans le modèle CLASSIC à l'aide d'optimisation bayésienne et d'observations

Dallosch, M. A., & Creed, I. F. (2021). Optimization of Landsat Chl-a Retrieval Algorithms in Freshwater Lakes through Classification of Optical Water Types. Remote Sensing, 13(22), 4607. https://doi.org/10.3390/rs13224607

Optimization of Landsat Chl-a Retrieval Algorithms in Freshwater Lakes through Classification of Optical Water Types

Qu, B., Roy, A., Melton, J.R., Baltzer, J.L., Ryu, Y., Detto, M., Sonnentag, O. (2024) Optimizing maximum carboxylation rate for North America’s boreal forests in the Canadian Land Surface Scheme Including biogeochemical Cycles (CLASSIC) v.1.3, Geoscientific Model Development

Optimizing maximum carboxylation rate for North America’s boreal forests in the Canadian Land Surface Scheme Including biogeochemical Cycles (CLASSIC) v.1.3

Arctic soils store large amounts of organic carbon and other elements, such as amorphous silicon, silicon, calcium, iron, aluminum, and phosphorous. Global warming is projected to be most pronounced in the Arctic, leading to thawing permafrost which, in turn, changes the soil element availability. To project how biogeochemical cycling in Arctic ecosystems will be affected by climate change, there is a need for data on element availability. Here, we analyzed the amorphous silicon (ASi) content as a solid fraction of the soils as well as Mehlich III extractions for the bioavailability of silicon (Si), calcium (Ca), iron (Fe), phosphorus (P), and aluminum (Al) from 574 soil samples from the circumpolar Arctic region. We show large differences in the ASi fraction and in Si, Ca, Fe, Al, and P availability among different lithologies and Arctic regions. We summarize these data in pan-Arctic maps of the ASi fraction and available Si, Ca, Fe, P, and Al concentrations, focusing on the top 100?cm of Arctic soil. Furthermore, we provide element availability values for the organic and mineral layers of the seasonally thawing active layer as well as for the uppermost permafrost layer. Our spatially explicit data on differences in the availability of elements between the different lithological classes and regions now and in the future will improve Arctic Earth system models for estimating current and future carbon and nutrient feedbacks under climate change (https://doi.org/10.17617/3.8KGQUN, Schaller and Goeckede, 2022)

Long-term atmospheric CO2 concentration records have suggested a reduction in the positive effect of warming on high-latitude carbon uptake since the 1990s. A variety of mechanisms have been proposed to explain the reduced net carbon sink of northern ecosystems with increased air temperature, including water stress on vegetation and increased respiration over recent decades. However, the lack of consistent long-term carbon flux and in situ soil moisture data has severely limited our ability to identify the mechanisms responsible for the recent reduced carbon sink strength. In this study, we used a record of nearly 100 site-years of eddy covariance data from 11 continuous permafrost tundra sites distributed across the circumpolar Arctic to test the temperature (expressed as growing degree days, GDD) responses of gross primary production (GPP), net ecosystem exchange (NEE), and ecosystem respiration (ER) at different periods of the summer (early, peak, and late summer) including dominant tundra vegetation classes (graminoids and mosses, and shrubs). We further tested GPP, NEE, and ER relationships with soil moisture and vapor pressure deficit to identify potential moisture limitations on plant productivity and net carbon exchange. Our results show a decrease in GPP with rising GDD during the peak summer (July) for both vegetation classes, and a significant relationship between the peak summer GPP and soil moisture after statistically controlling for GDD in a partial correlation analysis. These results suggest that tundra ecosystems might not benefit from increased temperature as much as suggested by several terrestrial biosphere models, if decreased soil moisture limits the peak summer plant productivity, reducing the ability of these ecosystems to sequester carbon during the summer

Wildfire is the largest disturbance affecting peatlands, with northern peat reserves expected to become more vulnerable to wildfire as climate change enhances the length and severity of the fire season. Recent research suggests that high water table positions after wildfire are critical to limit atmospheric carbon losses and enable the re-establishment of keystone peatland mosses (i.e. Sphagnum). Post-fire recovery of the moss surface in Sphagnum-feathermoss peatlands, however, has been shown to be limited where moss type and burn severity interact to result in a water repellent surface. While in situ measurements of moss water repellency in peatlands have been shown to be greater for feathermoss in both a burned and unburned state in comparison to Sphagnum moss, it is difficult to separate the effect of water content from species. Consequently, we carried out a laboratory based drying experiment where we compared the water repellency of two dominant peatland moss species, Sphagnum and feathermoss, for several burn severity classes including unburned samples. The results suggest that water repellency in moss is primarily controlled by water content, where a sharp threshold exists at gravimetric water contents (GWC) lower than ∼1.4 g g−1. While GWC is shown to be a strong predictor of water repellency, the effect is enhanced by burning. Based on soil water retention curves, we suggest that it is highly unlikely that Sphagnum will exhibit strong hydrophobic conditions under field conditions.

The availability and quality of quad-pol synthetic aperture radar (SAR) datasets has increased substantially since the early 2000s, allowing for polarimetrically complete investigations of freshwater ice. These investigations have lead to improved ice classification methods, new understanding of microwave-ice scattering processes, and the potential for new methods to extract ice observables. Such analyses are predicated on the decomposition of the target’s polarimetric properties along mathematical or physical lines. This paper comprehensively reviews the underlying theory and contemporary application of radar polarimetric decomposition as it applies to freshwater ice systems. Modelling and investigation of lake ice, river ice, and glacial systems are discussed. We conclude with recommendations for further research, discussing the value of further development of freshwater-ice models, their use in characterization of the scattering process, and the potential for new methods to extract environmental observables.

Although there is increasing consensus that river restoration should focus on restoring processes rather than form, proven techniques to design and monitor projects for sediment transport processes are lacking. This study monitors bedload transport and channel morphology in a rural, an urban unrestored, and an urban restored reach. Objectives are to compare bedload transport regimes, assess the stability and self-maintenance of constructed riffle-pool sequences, and evaluate the impact of the project on coarse sediment continuity in the creek. Sediment tracking is done using radio frequency identification tracers and morphologic change is assessed from repeated cross-section surveys. Mean annual velocity is used to quantify the average downstream velocity of tracers, defined as the mean overall tracer travel length divided by the total study duration. The channel reconstruction slows down the downstream velocity of particles in the D75 and D90 size classes, but does not significantly change the velocity of particles in the D50 size class or smaller. Surveys show that riffle features remain stable and that pool depths are maintained or deepened, while tracer paths match with what has been observed in natural riffle-pools. However, the slowdown of coarse sediment and increase in channel slope may lead to future failures related to over-steepening of the banks and a disruption in the continuity of sediment transport in the creek. This study demonstrates how bedload tracking and morphological surveys can be used to assess river restoration projects, and highlights the importance of incorporating coarse sediment connectivity into restoration design and monitoring.

Cyanobacteria, also known as blue-green algae for a long time, are the most ancient and problematic bloom-forming phylum on earth. An alert levels framework has been established by World Health Organization(WHO) to prevent the potential harmful cyanobacterial blooms. Normally, low cyanobacteria levels are found in surface water. 2000 cyanobacterial cells/mL and 100,000 cyanobacterial cells/mL are established for WHO Alert Level 1 and 2, respectively. However, eutrophication, climate change and other factors may promote the spread of cyanobacteria and increase the occurrence of harmful cyanobacterial blooms in water on a global scale. Hence, a rapid real time cyanobacteiral monitoring system is required to protect public health from the cyanotoxins produced by toxic cyanobacterial species. Current methods to control or prevent the development of harmful cyanobacterial blooms are either expensive, time consuming or not effective in the long term. The best method to control the blooms is to prevent the formation of the blooms at the very beginning. Although emerging advanced autofluorescence-based sensors, imaging flow cytometry applications, and remote sensing have been utilized for rapid real-time enumeration and classification of cyanobacteria, the need to accurately monitor low-level cyanobacterial species in water remains unsolved. Microflow cytometry has been employed as a functional cell analysis technique in past decades, and it can provide real-time, accurate results. The autofluorescence of cyanobacterial pigments can be used for determination and counting of cyanobacterial density in water. A pre-concentration system of an automated cyanobacterial concentration and recovery system (ACCRS) based on tangential flow filtration and back-flushing technique was applied to reduce the sample assay volume and increase the concentration of target cells for further cell capture and detection. In this project, a microflow cytometry platform with a microfluidic device and an automated pre-concentration system was established to monitor cyanobacteria and provide early warning alerts for potential harmful blooms. In this work, quantification of low-level cyanobacterial samples (∼ 5 cyanobacterial cells/mL) in water has been achieved by using a microflow cytometer together with a pre-concentration system (ACCRS). Meanwhile, this platform can also provide early warning alerts for potential harmful cyanobacterial blooms at least 15 days earlier before reaching WHO Alert Level 1. Results have shown that this platform can be applied for rapid determination of cyanobacteria and early warning alerts can be triggered for authorities to protect the public and the environment.

In seasonally frozen environments, hydrological processes are highly dynamic during and following the melt period in the spring, and this is when most of the runoff and groundwater re- charge happens. This is also when evapotranspiration (ET) fluxes start to increase in response to higher solar radiation, and a resumption of photosynthesis in evergreen species. This thesis applies the Canadian Land Surface Scheme (CLASS) to three Boreal Ecosystem Research and Monitoring Sites (BERMS) in the boreal forest in Saskatchewan; Old Jack Pine, Old Black Spruce and Old Aspen. CLASS was used to simulate the energy and water balance of the vegetation, soil and snowpack at the three sites. Consistent with previous studies, it was shown that ET is overestimated in the model during the melt/thaw period. A series of numerical experiments were undertaken to investigate in detail the controls on simulated fluxes within the CLASS model and explore the model behaviour. The phenomenon of freezing point depression, where water freezes below 0 °C in soils, is not represented in the CLASS model. Consequently, the model predicted a significant amount of transpiration to occur during the melt period while the soil was at 0 °C and ice was still present in the soil pores. Subtracting the transpiration that occurred from soil layers containing ice improved the simulated ET, compared with flux tower estimates. Therefore, it is suggested that implementing freezing point depression in the model and including a water stress function to shut down transpiration when the soil temperature is ≤ 0 °C would improve the simulated evapotran- spiration during the melt period. The study also showed that calibration of the model parameters improved the simulations but is unable to uniquely constrain the infiltration and soil drainage fluxes by either single objective (ET) or multi-objective (soil moisture and ET) calibration. Further research is needed to explore the hypothesis that root water uptake does not occur in soils where the soil temperature ≤ 0 °C would.

Major rivers that originate in mountainous areas provide the main water supply for more than one third of the world's population. These rivers typically exhibit a four-to-five-month high flow period driven by snowmelt and rain, followed by a seven-to-eight-month low flow period sustained by groundwater discharge from mountain headwaters. Recent small-scale and field-based studies have identified talus slopes, moraines, rock glaciers, and alpine meadows as the main landforms responsible for storing and discharging groundwater in these headwater environments and have further classified them as alpine aquifers. However, there has not been much progress upscaling our current small-scale understanding of alpine aquifers to the watershed-scale. This study aims to upscale our knowledge of alpine aquifers by developing a geospatial modelling approach that can 1) map the spatial extent and distribution of different aquifers that are common in alpine watersheds and 2) employ a numerical groundwater flow model to simulate annual groundwater storage and discharge for a given watershed. The Opabin sub-watershed, located within the Lake O'Hara watershed in British Columbia will be used as the pilot site. Over the past 15 years, numerous research studies have thoroughly characterized the aquifers present and developed an extensive library of relevant hydrological data. Present aquifers will be identified and differentiated from one another using an object-oriented classification technique that incorporates remote sensing imagery and a digital elevation model, that cover the extent of the watershed. After, the extent of the aquifers will be extracted from the classified map and discretized in a numeric groundwater flow model. Representative storage and discharge characteristics derived from the existing library of field-data will then be assigned to each aquifer, enabling the model to simulate the annual propagation of snowmelt and rain through the aquifers present and in doing so, effectively quantify annual groundwater storage and discharge in the Opabin sub-watershed. The resulting model will be calibrated and validated using measured stream discharge data from Opabin creek, and the model efficacy will be evaluated by applying it to two other watersheds in the Canadian Rockies, specifically within the Bow River basin. The new modelling approach will provide an efficient tool to quantify and predict the annual groundwater contributions from mountain headwaters to major rivers, which will in turn help downstream populations sustainably manage their water supply.

The role of hummocky terrain in governing runoff routing and focussing groundwater recharge in the Northern Prairies of North America is widely recognised. However, most hydrological studies in the region have not effectively utilised information on the surficial geology and associated landforms in large-scale hydrological characterization. The present study uses an automated digital elevation model (DEM) analysis of a 6500-km2 area in the Northern Prairies to quantify hydrologically relevant terrain parameters for the common types of terrains in the prairies with different surficial deposits widespread in the prairies, namely, moraines and glaciolacustrine deposits. Runoff retention (and storage) capacity within depressions varies greatly between different surficial deposits and is comparable in magnitude with a typical amount of seasonal snowmelt runoff generation. The terrain constraint on potential runoff retention varies from a few millimetres in areas classified as moraine to tens of millimetres in areas classified as stagnant ice moraine deposits. Fluted moraine and glaciolacustrine deposits have intermediate storage capacity values. The study also identified the probability density function describing a number of immediate upstream neighbours for each depression in a fill-and-spill network. A relationship between depression parameters and surficial deposits, as well as identified depression network structure, allows parametrisation of hydrologic models outside of the high-resolution DEM coverage, which can still account for terrain variation in the Prairies.

Identical or nearly similar code fragments in a software system's code-base are known as code clones. Code clones from the same clone class have a tendency of co-changing (changing together) consistently during evolution. Focusing on this co-change tendency, existing studies have investigated prediction and ranking co-change candidates of regular clones. However, a recent study shows that micro-clones which are smaller than the minimum size threshold of regular clones might also need to be co-changed consistently during evolution. Thus, identifying and ranking co-change candidates of micro-clones is also important. In this paper, we investigate factors that influenc the co-change tendency of the co-change candidates of a target micro-clone fragment. We mine fil level evolutionary coupling from thousands of revisions of our subject systems through mining association rules and analyze this coupling for the purpose of ranking. According to our finding on six open-source subject systems written in Java and C, consistent co-change tendency of micro-clones is influenc d by fil proximity of the micro-clone fragments as well as evolutionary coupling of the file containing those micro-clone fragments. On the basis of our finding we propose a composite ranking mechanism by incorporating both fil proximity and file coupling for ranking co-change candidates for micro-clones and fin that our proposed mechanism performs significantl better than File Proximity Ranking mechanism. We believe that our proposed ranking mechanism has the potential to help programmers in updating micro-clones consistently with less effort

As a consequence of increasing temperatures, a rapid increase in shrub vegetation has occurred throughout the circumpolar North and is expected to continue. Rates of shrub expansion are highly variable, both at the regional scale and within local study areas. This study uses repeat airborne LiDAR and field surveys to measure changes in shrub vegetation cover along with landscape-scale variations in a well-studied subarctic headwater catchment in Yukon Territory, Canada. Airborne LiDAR surveys were conducted in August 2007 and 2018, whereas vegetation surveys were conducted in summer 2019. Machine learning classification algorithms were used to predict shrub presence/absence in 2018 based on rasterized LiDAR metrics, with the best-performing model applied to the 2007 LiDAR to create binary shrub cover layers to compare between survey years. Results show a 63.3% total increase in detectable shrub cover >= 0.45 m in height between 2007 and 2018, with an average yearly expansion of 5.8%. These changes were compared across terrain derivatives to quantify the influence of topography on shrub expansion. Terrain comparisons show that shrubs are located in and are preferentially expanding into lower and flatter areas near stream networks, at lower slope positions and with a higher potential for topographic wetness. Overall, the findings from this research reinforce the documented increase in pan-Arctic shrub vegetation in recent years, quantify the variation in shrub expansion over terrain derivatives at the landscape scale, and demonstrate the feasibility of using LiDAR to compare changes in shrub properties over time.

The real-time polymerase chain reaction (PCR), commonly known as quantitative PCR (qPCR), is increasingly common in environmental microbiology applications. During the COVID-19 pandemic, qPCR combined with reverse transcription (RT-qPCR) has been used to detect and quantify SARS-CoV-2 in clinical diagnoses and wastewater monitoring of local trends. Estimation of concentrations using qPCR often features a log-linear standard curve model calibrating quantification cycle (Cq) values obtained from underlying fluorescence measurements to standard concentrations. This process works well at high concentrations within a linear dynamic range but has diminishing reliability at low concentrations because it cannot explain “non-standard” data such as Cq values reflecting increasing variability at low concentrations or non-detects that do not yield Cq values at all. Here, fundamental probabilistic modeling concepts from classical quantitative microbiology were integrated into standard curve modeling approaches by reflecting well-understood mechanisms for random error in microbial data. This work showed that data diverging from the log-linear regression model at low concentrations as well as non-detects can be seamlessly integrated into enhanced standard curve analysis. The newly developed model provides improved representation of standard curve data at low concentrations while converging asymptotically upon conventional log-linear regression at high concentrations and adding no fitting parameters. Such modeling facilitates exploration of the effects of various random error mechanisms in experiments generating standard curve data, enables quantification of uncertainty in standard curve parameters, and is an important step toward quantifying uncertainty in qPCR-based concentration estimates. Improving understanding of the random error in qPCR data and standard curve modeling is especially important when low concentrations are of particular interest and inappropriate analysis can unduly affect interpretation, conclusions regarding lab performance, reported concentration estimates, and associated decision-making.

Developers often search for relevant code examples on the web for their programming tasks. Unfortunately, they face two major problems. First, the search is impaired due to a lexical gap between their query (task description) and the information associated with the solution. Second, the retrieved solution may not be comprehensive, i.e., the code segment might miss a succinct explanation. These problems make the developers browse dozens of documents in order to synthesize an appropriate solution. To address these two problems, we propose CROKAGE (Crowd Knowledge Answer Generator), a tool that takes the description of a programming task (the query) and provides a comprehensive solution for the task. Our solutions contain not only relevant code examples but also their succinct explanations. Our proposed approach expands the task description with relevant API classes from Stack Overflow Q&A threads and then mitigates the lexical gap problems. Furthermore, we perform natural language processing on the top quality answers and then return such programming solutions containing code examples and code explanations unlike earlier studies. We evaluate our approach using 97 programming queries, of which 50% was used for training and 50% was used for testing, and show that it outperforms six baselines including the state-of-art by a statistically significant margin. Furthermore, our evaluation with 29 developers using 24 tasks (queries) confirms the superiority of CROKAGE over the state-of-art tool in terms of relevance of the suggested code examples, benefit of the code explanations and the overall solution quality (code + explanation).

Quantifying the behavior and performance of hydrologic models is an important aspect of understanding the underlying hydrologic systems. We argue that classical error measures do not offer a complete picture for building this understanding. This study demonstrates how the information theoretic measure known as transfer entropy can be used to quantify the active transfer of information between hydrologic processes at various timescales and facilitate further understanding of the behavior of these systems. To build a better understanding of the differences in dynamics, we compare model instances of the Structure for Unifying Multiple Modeling Alternatives (SUMMA), the Variable Infiltration Capacity (VIC) model, and the Precipitation Runoff Modeling System (PRMS) across a variety of hydrologic regimes in the Columbia River Basin in the Pacific Northwest of North America. Our results show differences in the runoff of the SUMMA instance compared to the other two models in several of our study locations. In the Snake River region, SUMMA runoff was primarily snowmelt driven, while VIC and PRMS runoff was primarily influenced by precipitation and evapotranspiration. In the Olympic mountains, evapotranspiration interacted with the other water balance variables much differently in PRMS than in VIC and SUMMA. In the Willamette River, all three models had similar process networks at the daily time scale but showed differences in information transfer at the monthly timescale. Additionally, we find that all three models have similar connectivity between evapotranspiration and soil moisture. Analyzing information transfers to runoff at daily and monthly time steps shows how processes can operate on different timescales. By comparing information transfer with correlations, we show how transfer entropy provides a complementary picture of model behavior.

Assessing extreme precipitation events is of high importance to hydrological risk assessment, decision making, and adaptation strategies. Global gridded precipitation products, constructed by combining various data sources such as precipitation gauge observations, atmospheric reanalyses and satellite estimates, can be used to estimate extreme precipitation events. Although these global precipitation products are widely used, there has been limited work to examine how well these products represent the magnitude and frequency of extreme precipitation. In this work, the five most widely used global precipitation datasets (MSWEP, CFSR, CPC, PERSIANN-CDR and WFDEI) are compared to each other and to GHCN-daily surface observations. The spatial variability of extreme precipitation events and the discrepancy amongst datasets in predicting precipitation return levels (such as 100- and 1000-year) were evaluated for the time period 1979-2017. The behaviour of extremes, that is the frequency and magnitude of extreme precipitation, was quantified using indices of the heaviness of the upper tail of the probability distribution. Two parameterizations of the upper tail, the power and stretched-exponential, were used to reveal the probabilistic behaviour of extremes. The analysis shows strong spatial variability in the frequency and magnitude of precipitation extremes as estimated from the upper tails of the probability distributions. This spatial variability is similar to the Köppen-Geiger climate classification. The predicted 100- and 1000-year return levels differ substantially amongst the gridded products, and the level of discrepancy varies regionally, with large differences in Africa and South America and small differences in North America and Europe. The results from this work reveal the shortcomings of global precipitation products in representing extremes. The work shows that there is no single global product that performs best for all regions and climates.

Software developers often look for solutions to their code-level problems by submitting questions to technical Q&A websites like Stack Overflow (SO). They usually include example code segments with questions to describe the programming issues. SO users prefer to reproduce the reported issues using the given code segments when they attempt to answer the questions. Unfortunately, such code segments could not always reproduce the issues due to several unmet challenges (e.g., external library not found) that might prevent questions from receiving prompt and appropriate solutions. A previous study produced a catalog of potential challenges that hinder the reproducibility of issues reported at SO questions. However, it is unknown how the practitioners (i.e., developers) perceive the challenge catalog. Understanding the developers' perspective is inevitable to introduce interactive tool support that promotes reproducibility. We thus attempt to understand developers' perspectives by surveying 53 users of SO. In particular, we attempt to -- (1) see developers' viewpoints on the agreement to those challenges, (2) find the potential impact of those challenges, (3) see how developers address them, and (4) determine and prioritize tool support needs. Survey results show that about 90% of participants agree to the already exposed challenges. However, they report some additional challenges (e.g., error log missing) that might prevent reproducibility. According to the participants, too short code segment and absence of required Class/Interface/Method from code segments severely prevent reproducibility, followed by missing important part of code. To promote reproducibility, participants strongly recommend introducing tool support that interacts with question submitters with suggestions for improving the code segments if the given code segments fail to reproduce the issues.

Hydrological models, which have become increasingly complex in the last half century due to the advances in computing capabilities and data collection, have been extensively utilized to facilitate decision-making in water resources management. Such computer-based models generally contain considerable parameters that cannot be directly measured, and hence calibration and validation are required to ensure model transferability and robustness in model building (development). The most widely used method used for assessing model transferability in time is the split-sample test (SST) framework, which has even been a paradigm in the hydrological modeling community for decades. However, there is no clear guidance or empirical/numerical evidence that supports how a dataset should be split into the calibration and validation subsets. The SST decisions usually appear to be unclear and even subjective in literature. Even though past studies have spared tremendous efforts to investigate possible ways to improve model performance by adopting various data splitting methods; however, such problem of data splitting still remain as a challenge and no consensus has achieved on which splitting method may be optimal in hydrological modeling community. One of the key reasons is lacking a robust evaluation framework to objectively compare different data splitting methods in the “out-of-sample” model application period. To mitigate these gaps, this thesis aims at assessing different data splitting methods using the large-sample hydrology approach to identify optimal data splitting methods under different conditions, as well as exploring alternative validation methods to improve model robustness that is usually done by the SST method. First, the thesis introduces a unique and comprehensive evaluation framework to compare different data splitting methods. This evaluation framework defines different model build years, as such models can be built in various data availability scenarios. Years after the model build year are retained as model testing period, which acts as an “out-of-sample” data beyond the model building period and matches how models are applied in operational use. The evaluation framework allows to incorporate various data splitting methods into comparison, as the comparison of model performance is performed in the common testing period no matter how calibration and validation data are split in model building period. Moreover, a reference climatology, which is purely observation data-based, is applied to benchmark our model simulations. Model inadequacy is properly handled by considering the possible decisions modelers may make when faced with bad model simulations. As such, the model building can be more robust and realistic. Example approaches which cover a wide range of aspects modelers may care about in practice are provided to assess large-sample modeling results. Two large-sample modeling experiments are performed in the proposed evaluation framework to compare different data splitting methods. In the first experiment, two conceptual hydrological models are applied in 463 catchments across the United States to evaluate 50 different continuous calibration sub-periods (CSPs) for model calibration (varying data period length and recency) across five different model build year scenarios, which ensures robust results across three testing period conditions. Model performance in testing periods are assessed from three independent aspects: frequency of each short-period CSP being better than its corresponding full-period CSP; central tendency of the objective function metric as computed in model testing period; and frequency that a CSP correctly classifies model testing period failure and success. The second experiment assesses 44 representative continuous and discontinuous data splitting methods using a conceptual hydrological model in 463 catchments across the United States. These data splitting methods consist of all the ways hydrological model calibration split-sampling is currently done when only a single split sample is evaluated and one method found in data-driven modeling. This results in over 0.4 million model calibration-validation and 1.7 million model testing exercises for an extensive analysis. Model performance in testing periods are assessed in similar ways in the first experiment except that all model optimization trials are utilized to draw even more robust conclusions. Three SST recommendations are made based on the strong empirical evidence. Calibrating models to older data and then validating models on newer data produces inferior model testing period performance in every single analysis conducted and should be avoided. Calibrating a model to the full available data period and skipping temporal model validation entirely is the most robust choice. It is recommended that hydrological modelers rebuild models after their validation experiments, but prior to operational use of the model, by calibrating models to all available data. Last but not least, alternative model validation methods are further tested to enhance model robustness based on the above large-sample modeling results. A proxy validation is adopted to replace the traditional validation period in the SST method by using Split Kling-Gupta Efficiency (KGE) and Split Reference KGE in calibration to identify unacceptable models. The proxy validation is demonstrated to have some promise to enhance model robustness when all data are used in calibration.

Both climate and statistical models play an essential role in the process of demonstrating that the distribution of some atmospheric variable has changed over time and in establishing the most likely causes for the detected change. One statistical difficulty in the research field of detection and attribution resides in defining events that can be easily compared and accurately inferred from reasonable sample sizes. As many impacts studies focus on extreme events, the inference of small probabilities and the computation of their associated uncertainties quickly become challenging. In the particular context of event attribution, the authors address the question of how to compare records between the counterfactual “world as it might have been” without anthropogenic forcings and the factual “world that is.” Records are often the most important events in terms of impact and get much media attention. The authors will show how to efficiently estimate the ratio of two small probabilities of records. The inferential gain is particularly substantial when a simple hypothesis-testing procedure is implemented. The theoretical justification of such a proposed scheme can be found in extreme value theory. To illustrate this study’s approach, classical indicators in event attribution studies, like the risk ratio or the fraction of attributable risk, are modified and tailored to handle records. The authors illustrate the advantages of their method through theoretical results, simulation studies, temperature records in Paris, and outputs from a numerical climate model.

This book exposes practitioners and students to the theory and application of river and lake ice processes to gain a better understanding of these processes for modelling and forecasting. It focuses on the following processes of the surface water ice: freeze-up, ice cover thickening, ice cover breakup and ice jamming. The reader will receive a fundamental understanding of the physical processes of each component and how they are applied in monitoring and modelling ice covers during the winter season and forecasting ice floods. Exercises accompany each component to reinforce the theoretical principles learned. These exercises will also expose the reader to different tools to process data, such a space-borne remote sensing imagery for ice cover classification. A thread supporting numerical modelling of river ice and lake ice processes runs through the book.

Aksamit, N. O., & Pomeroy, J. W. (2015). Saltating snow mechanics: Three species classification from high speed videography. Proceedings of the 72nd Eastern Snow Conference, 56-66. http://www.usask.ca/hydrology/papers/Aksamit_Pomeroy_2015.pdf

Saltating snow mechanics: Three species classification from high speed videography

first_pagesettingsOrder Article Reprints Open AccessArticle Satellite-Observed Soil Moisture as an Indicator of Wildfire Risk by Jaison Thomas Ambadan,Matilda Oja,Ze’ev GedalofORCID andAaron A. Berg *ORCID Department of Geography, Environment and Geomatics, University of Guelph, Guelph, ON N1G 2W1, Canada * Author to whom correspondence should be addressed. Remote Sens. 2020, 12(10), 1543; https://doi.org/10.3390/rs12101543 Received: 20 February 2020 / Revised: 6 May 2020 / Accepted: 7 May 2020 / Published: 12 May 2020 (This article belongs to the Section Remote Sensing in Agriculture and Vegetation) Download Browse Figures Versions Notes Abstract Wildfires are a concerning issue in Canada due to their immediate impact on people’s lives, local economy, climate, and environment. Studies have shown that the number of wildfires and affected areas in Canada has increased during recent decades and is a result of a warming and drying climate. Therefore, identifying potential wildfire risk areas is increasingly an important aspect of wildfire management. The purpose of this study is to investigate if remotely sensed soil moisture products from the Soil Moisture and Ocean Salinity (SMOS) satellite can be used to identify potential wildfire risk areas for better wildfire management. We used the National Fire Database (NFDB) fire points and polygons to group the wildfires according to ecozone classifications, as well as to analyze the SMOS soil moisture data over the wildfire areas, between 2010–2017, across fourteen ecozones in Canada. Timeseries of 3-day, 5-day, and 7-day soil moisture anomalies prior to the onset of each wildfire occurrence were examined over the ecozones individually. Overall, the results suggest, despite the coarse-resolution, SMOS soil moisture products are potentially useful in identifying soil moisture anomalies where wildfire hot-spots may occur.

Recent advancements in language models, particularly for high-resource languages, have not been paralleled in low-resource languages spoken across Africa. This thesis addresses this gap by scaling pre-training data and developing improved language models for African languages. We introduce Wura, a high-quality, document-level pre-training dataset encompassing 16 African languages along with four high-resource languages commonly spoken on the continent: Arabic, English, French, and Portuguese. Leveraging Wura, we pre-train new versions of the AfriBERTa (encoder-only) and AfriTeVa (encoder-decoder) model families. These new models demonstrate superior performance across a variety of natural language understanding and generation tasks compared to existing baselines. Notably, AfriTeVa V2 Large (1B) stands as the largest sequence-to-sequence model pre-trained for African languages to date. Our methodology includes a meticulous three-stage curation process for Wura --- auditing and filtering existing web crawls, initiating new web crawls, and integrating existing language resources. The experimental setup and evaluation encompass tasks like text classification, information retrieval, translation, summarization, and cross-lingual question answering. Our new models outperform their predecessors and other established models, even those with significantly more parameters, highlighting the efficacy of high-quality pre-training data. Furthermore, we study the generalization of our models to languages not deliberately included in their pre-training data.

Lake ice is a significant component of the cryosphere due to its large spatial coverage in high-latitude regions during the winter months. The Laurentian Great Lakes are the world’s largest supply of freshwater and their ice cover has a major impact on regional weather and climate, ship navigation, and public safety. Ice experts at the Canadian Ice Service (CIS) have been manually producing operational Great Lakes image analysis charts based on visual interpretation of the synthetic aperture radar (SAR) images. In that regard, we have investigated the performance of the semi-automated segmentation algorithm “glocal” Iterative Region Growing with Semantics (IRGS) for lake ice classification using dual polarized RADARSAT-2 imagery acquired over Lake Erie. Analysis of various case studies indicated that the “glocal” IRGS algorithm could provide a reliable ice-water classification using dual polarized images with a high overall accuracy of 90.4%. However, lake ice types that are based on stage of development were not effectively identified due to the ambiguous relation between backscatter and ice types. The slight improvement of using dual-pol as opposed to single-pol images for ice-water discrimination was also demonstrated.

Climate extremes such as heat waves and droughts are projected to occur more frequently with increasing temperature and an intensified hydrological cycle. It is important to understand and quantify how forest carbon fluxes respond to heat and drought stress. In this study, we developed a series of daily indices of sensitivity to heat and drought stress as indicated by air temperature (Ta ) and evaporative fraction (EF). Using normalized daily carbon fluxes from the FLUXNET Network for 34 forest sites in North America, the seasonal pattern of sensitivities of net ecosystem productivity (NEP), gross ecosystem productivity (GEP) and ecosystem respiration (RE) in response to Ta and EF anomalies were compared for different forest types. The results showed that warm temperatures in spring had a positive effect on NEP in conifer forests but a negative impact in deciduous forests. GEP in conifer forests increased with higher temperature anomalies in spring but decreased in summer. The drought-induced decrease in NEP, which mostly occurred in the deciduous forests, was mostly driven by the reduction in GEP. In conifer forests, drought had a similar dampening effect on both GEP and RE, therefore leading to a neutral NEP response. The NEP sensitivity to Ta anomalies increased with increasing mean annual temperature. Drier sites were less sensitive to drought stress in summer. Natural forests with older stand age tended to be more resilient to the climate stresses compared to managed younger forests. The results of the Classification and Regression Tree analysis showed that seasons and ecosystem productivity were the most powerful variables in explaining the variation of forest sensitivity to heat and drought stress. Our results implied that the magnitude and direction of carbon flux changes in response to climate extremes are highly dependent on the seasonal dynamics of forests and the timing of the climate extremes.

Software architectural changes involve more than one module or component and are complex to analyze compared to local code changes. Development teams aiming to review architectural aspects (design) of a change commit consider many essential scenarios such as access rules and restrictions on usage of program entities across modules. Moreover, design review is essential when proper architectural formulations are paramount for developing and deploying a system. Untangling architectural changes, recovering semantic design, and producing design notes are the crucial tasks of the design review process. To support these tasks, we construct a lightweight tool [4] that can detect and decompose semantic slices of a commit containing architectural instances. A semantic slice consists of a description of relational information of involved modules, their classes, methods and connected modules in a change instance, which is easy to understand to a reviewer. We extract various directory and naming structures (DANS) properties from the source code for developing our tool. Utilizing the DANS properties, our tool first detects architectural change instances based on our defined metric and then decomposes the slices (based on string processing). Our preliminary investigation with ten open-source projects (developed in Java and Kotlin) reveals that the DANS properties produce highly reliable precision and recall (93-100%) for detecting and generating architectural slices. Our proposed tool will serve as the preliminary approach for the semantic design recovery and design summary generation for the project releases.

Software architectural changes involve more than one module or component and are complex to analyze compared to local code changes. Development teams aiming to review architectural aspects (design) of a change commit consider many essential scenarios such as access rules and restrictions on usage of program entities across modules. Moreover, design review is essential when proper architectural formulations are paramount for developing and deploying a system. Untangling architectural changes, recovering semantic design, and producing design notes are the crucial tasks of the design review process. To support these tasks, we construct a lightweight tool [4] that can detect and decompose semantic slices of a commit containing architectural instances. A semantic slice consists of a description of relational information of involved modules, their classes, methods and connected modules in a change instance, which is easy to understand to a reviewer. We extract various directory and naming structures (DANS) properties from the source code for developing our tool. Utilizing the DANS properties, our tool first detects architectural change instances based on our defined metric and then decomposes the slices (based on string processing). Our preliminary investigation with ten open-source projects (developed in Java and Kotlin) reveals that the DANS properties produce highly reliable precision and recall (93-100%) for detecting and generating architectural slices. Our proposed tool will serve as the preliminary approach for the semantic design recovery and design summary generation for the project releases.

Lake ice is a significant component of the cryosphere due to its large spatial coverage in high-latitude regions during the winter months. The Laurentian Great Lakes are the world’s largest supply of freshwater and their ice cover has a major impact on regional weather and climate, ship navigation, and public safety. Ice experts at the Canadian Ice Service (CIS) have been manually producing operational Great Lakes image analysis charts based on visual interpretation of the synthetic aperture radar (SAR) images. In that regard, we have investigated the performance of the semi-automated segmentation algorithm “glocal” Iterative Region Growing with Semantics (IRGS) for lake ice classification using dual polarized RADARSAT-2 imagery acquired over Lake Erie. Analysis of various case studies indicated that the “glocal” IRGS algorithm could provide a reliable ice-water classification using dual polarized images with a high overall accuracy of 90.4%. However, lake ice types that are based on stage of development were not effectively identified due to the ambiguous relation between backscatter and ice types. The slight improvement of using dual-pol as opposed to single-pol images for ice-water discrimination was also demonstrated.

Semi-automated classification of lake ice cover using dual polarization RADARSAT-2 imagery

Simulating catchment response to climate and land use change using catchment classification and virtual basin modelling, paper

Simulating historical and future boreal forest net ecosystem production using CLASSIC

Permafrost thaw is causing a rapid evolution of the lowland discontinuous permafrost regions of the Taiga Plains in Northern Canada and elsewhere in the Northern Hemisphere. Notably, this thaw is changing the spatial distribution of the dominant hydrologic land cover types (permafrost plateaus, fens, and isolated bogs) in parts of the Northwest Territories (NWT), Canada. Here, we develop a multinomial time series land cover model (TSLCM) to simulate historical land cover transitions, model spatial patterns of transition, and predict the long term evolution of land cover in the areas surrounding the Scotty Creek Research Station (SCRS), NWT, and similar discontinuous permafrost landscapes. The machine learning-based TSLCM is informed by a set of observed spatio-temporal variables. The independent variables represent driving factors of change, and include the estimated summertime land surface temperature anomaly (LST), the distance and a custom cost distance to land cover interfaces, time increment between initial and final states, and time-accumulated temperature; the dependent variable is classified land cover maps from 1970 to 2008. First, we applied both random forest (RF) and Multinomial Log-Linear Regression (MLR) methods to train a synthetic data model; the model which improves the performance of the TSLCM in extrapolating time series change by adding new data instances to the initial data set. We boosted the initial data by combining the predicted land cover change maps from synthetic data model and the real data set. Then, we evaluated a MLR, RF, and an extreme gradient boosting (XGBoost) model in their ability to simulate land cover change. The final results of this study show that the Ensemble Learning (EL) based approaches are capable of effectively representing historical land cover change and can produce physically consistent and plausible future land cover scenarios. Deterministic predictions from the TSLCM indicate that the permafrost plateaus’ coverage will continue to decrease, with corresponding decreases in isolated bogs’ coverage and their secondary runoff contributing areas.

Bo, Q., Roy, A., Melton, J., & Sonnentag, O. (2020). Site-level simulations of water and carbon balances in North American boreal forests across permafrost-free and permafrost regions using CLASSIC. Colloque annuel du Centre d'études nordiques. Université du Québec à Montréal, Québec, Canada. Conference Presentation

Site-level simulations of water and carbon balances in North American boreal forests across permafrost-free and permafrost regions using CLASSIC

Snow is an essential component in hydrology systems, and more speci cally, monitoring variations in snow cover can provide valuable information about water supply, wildlife habitats, and climate changes. In recent years, the potential of wide band snow radar (e.g. 2-8 GHz) has been discovered with a series of campaigns in Operation IceBrdige (OIB). Due to data from OIB mainly focusing on sea ice, most of the algorithms were also developed for snow on sea ice. As a result, this thesis aimed to test the applicability of the interface-based pulse peakiness snow depth retrieval method to snow on land. In addition, due to the common usage of radar altimeter in sea ice classi cation, this thesis also explored the possibility of adapting some of the ideas in sea ice classi cation to develop another retrieval method. Both approaches were tested on the 6 major vegetation types (tree, tall shrub, riparian shrub, dwarf shrub, tussock, and lichen) in the study area. Snow depth derived from Airborne Laser Scanner (ALS) point clouds was used as the reference for snow depth retrieval. Running the recalibrated pulse peakiness algorithm yielded a Mean Absolute Error (MAE) of 12 cm, 27 cm, 29 cm, 13 cm, 9 cm, and 10 cm for tree, tall shrub, riparian shrub, dwarf shrub, tussock and lichen respectively. It was concluded that the principles behind the pulse peakiness approach is valid for snow on land. The presence of surface vegetation and the hummocky terrain of the study area did present some considerable challenges. In comparison, the classi cation approach using K-means while produced some accurate results under speci c situations, was not as robust as the existing pulse peakiness approach. Although, it was argued that the classi cation approach was in early stages and there were some potential for better results.

Software is solely designed, implemented, tested, and inspected by expert people, unlike other engineering projects where they are mostly implemented by workers (non-experts) after designing by engineers. Researchers and practitioners have linked software bugs, security holes, problematic integration of changes, complex-to-understand codebase, unwarranted mental pressure, and so on in software development and maintenance to inconsistent and complex design and a lack of ways to easily understand what is going on and what to plan in a software system. The unavailability of proper information and insights needed by the development teams to make good decisions makes these challenges worse. Therefore, software design documents and other insightful information extraction are essential to reduce the above mentioned anomalies. Moreover, architectural design artifacts extraction is required to create the developer’s profile to be available to the market for many crucial scenarios. To that end, architectural change detection, categorization, and change description generation are crucial because they are the primary artifacts to trace other software artifacts. However, it is not feasible for humans to analyze all the changes for a single release for detecting change and impact because it is time-consuming, laborious, costly, and inconsistent. In this thesis, we conduct six studies considering the mentioned challenges to automate the architectural change information extraction and document generation that could potentially assist the development and maintenance teams. In particular, (1) we detect architectural changes using lightweight techniques leveraging textual and codebase properties, (2) categorize them considering intelligent perspectives, and (3) generate design change documents by exploiting precise contexts of components’ relations and change purposes which were previously unexplored. Our experiment using 4000+ architectural change samples and 200+ design change documents suggests that our proposed approaches are promising in accuracy and scalability to deploy frequently. Our proposed change detection approach can detect up to 100% of the architectural change instances (and is very scalable). On the other hand, our proposed change classifier’s F1 score is 70%, which is promising given the challenges. Finally, our proposed system can produce descriptive design change artifacts with 75% significance. Since most of our studies are foundational, our approaches and prepared datasets can be used as baselines for advancing research in design change information extraction and documentation.

Drainage leads to trade-offs between crop production efficiency and wetland conservation, with complex impacts on ecosystem services. In North America’s Prairie Pothole Region, wetland drainage is widespread, often to increase the available land for cultivation, prevent crop loss due to flooding, and manage soil salinity. Wetlands are known for providing key ecosystem services such as improved water quality, flood mitigation, and carbon storage. There is limited research on how changes to soil hydrology and soil redistribution through wetland drainage can impact soil carbon storage and persistence in this region. This research evaluates factors that contribute to soil carbon storage in drained prairie pothole wetland based on 33 drained wetlands in Saskatchewan, Canada. These analyses showed regional differences in the response of soil carbon storage to drainage that are driven by environmental factors such as annual precipitation, temperature, and wetland permanence. We observed increasing soil carbon storage from the Dark Brown to Black soil zones, as well as with longer wetland pond permanence. The sampling depth used for calculating soil carbon storage was especially important when comparing geographically across the soil zones as the Black soil zone had greater soil carbon stored at depth. Soil carbon was also intensively monitored over 2 years following installation of surface drainage across a wetland complex (8 drained wetlands) where water was partially directed to a consolidation wetland. We further assessed changes in soil carbon dynamics and protection from microbial decomposition based on three soil organic matter fractions, ATR-FTIR for organic matter functional groups, and phospholipid fatty acid analysis to understand the microbial community abundance and structure. After 2 years following drainage, ephemeral wetlands with short pond permanence were found to be most sensitive to drainage and the only wetland class with decreases in soil carbon. The temporary and seasonal wetland classes showed no significant differences in soil carbon content but there were changes in the organic matter with depth due to soil redistribution during drainage implementation. Jointly, this research provides region-specific estimates of soil carbon storage in drained prairie pothole wetlands that can be used to inform wetland soil carbon management in cultivated fields.

Increased fire frequency, extent and severity are expected to strongly affect the structure and function of boreal forest ecosystems. In this study, we examined 213 plots in boreal forests dominated by black spruce (Picea mariana) or jack pine (Pinus banksiana) of the Northwest Territories, Canada, after an unprecedentedly large area burned in 2014. Large fire size is associated with high fire intensity and severity, which would manifest as areas with deep burning of the soil organic layer (SOL). Our primary objectives were to estimate burn depth in these fires and then to characterise landscapes vulnerable to deep burning throughout this region. Here we quantify burn depth in black spruce stands using the position of adventitious roots within the soil column, and in jack pine stands using measurements of burned and unburned SOL depths. Using these estimates, we then evaluate how burn depth and the proportion of SOL combusted varies among forest type, ecozone, plot-level moisture and stand density. Our results suggest that most of the SOL was combusted in jack pine stands regardless of plot moisture class, but that black spruce forests experience complete combustion of the SOL only in dry and moderately well-drained landscape positions. The models and calibrations we present in this study should allow future research to more accurately estimate burn depth in Canadian boreal forests.

Study region Italy. Study focus Knowing magnitude and frequency of extreme precipitation is necessary to reduce their impact on vulnerable areas. Here we investigate the performance of the Generalized Extreme Value () distribution, using a fine-resolution satellite-based gridded product, to analyze 13,247 daily rainfall annual maxima samples. A non-extreme value distribution with a power-type behavior, that is, the Burr Type XII (), is also evaluated and used to test the reliability of the in describing extreme rainfall. New hydrological insights for the region (1) in 44.9 % of the analyzed samples the predicts an upper rainfall limit; we deem this is an artifact due to sample variations; (2) we suggest the distribution, that is, the with shape parameters restricted only to positive values as a more consistent model complying with the nature of extreme precipitation; (3) , , and performed equally well in describing the observed annual precipitation, yet all distributions underestimate the observed sample maximum; (4) the , for large return periods, predicts larger rainfall amounts compared to indicating that estimates could underestimate the risk of extremes; and (5) the correlation between the predicted rainfall and the elevation is investigated. Based on the results of this study, we suggest instead of using the classical to use the and non-extreme value distributions such as the to describe precipitation extremes.

In an effort to feed a growing world population, agriculture has rapidly intensified over the last six decades, relying heavily on agrochemicals (fertilizers, insecticides, fungicides, and herbicides) to increase and maintain desired crop yields. Despite environmental concerns in Canada’s agricultural regions, long-term patterns of changing crops and the associated trends in the proportion of cropland treated with agrochemicals are poorly documented. Using the Canadian Census of Agriculture, we compiled historical data over 35 years (eight census periods: 1981–2016) on agrochemical applications, measured as the proportion of cropland treated with pesticides and fertilizers and the associated crop classes, to identify and interpret spatial and temporal trends in Canada’s agricultural practices across 260 census units. Due to differences in agricultural practices, soil, and climatic conditions across the country, the Pacific (British Columbia), Prairie (Alberta, Saskatchewan, Manitoba), Central (Ontario, Quebec), and Atlantic (Nova Scotia, New Brunswick, Newfoundland/Labrador, Prince Edward Island) regions were analyzed separately. Most of the agrochemicals in Canada were applied in the Prairie and Central regions, which combined comprise 97% of the total cropland. Fertilizers were the dominant agrochemicals across Canada applied on 48% (Pacific) to 78% (Prairie) of the total cropland area, followed by herbicides, which were applied on 30% (Pacific) to 81% (Prairie) of the total cropland area in 2016. Notably, we observed significant changes between 1996 and 2016 in area treated with fungicides and insecticides, which increased by 412% and 50% in the Prairie region and by 291% and 149% in the Central region, respectively. The proportion and distribution of crops shifted in favor of more oilseeds and soybeans in the most intensive Prairie and Central regions, whereas cereals decreased over the same time period. Our analysis of past and current trends of agrochemicals and cropping patterns within Canada indicates a rapid and systemic increase in chemical use, and policies that promote a shift toward lower chemical reliance through sustainable agricultural practices are urgently needed.

Scientific workflows have emerged as well-established pillars of large-scale computational science and appeared as torchbearers to formalize and structure a massive amount of complex heterogeneous data and accelerate scientific progress. Scientists of diverse domains can analyze their data by constructing scientific workflows as a useful paradigm to manage complex scientific computations. A workflow can analyze terabyte-scale datasets, contain numerous individual tasks, and coordinate between heterogeneous tasks with the help of scientific workflow management systems (SWfMSs). However, even for expert users, workflow creation is a complex task due to the dramatic growth of tools and data heterogeneity. Scientists are now more willing to publicly share scientific datasets and analysis pipelines in the interest of open science. As sharing of research data and resources increases in scientific communities, scientists can reuse existing workflows shared in several workflow repositories. Unfortunately, several challenges can prevent scientists from reusing those workflows, which hurts the purpose of the community-oriented knowledge base. In this thesis, we first identify the repositories that scientists use to share and reuse scientific workflows. Among several repositories, we find Galaxy repositories have numerous workflows, and Galaxy is the mostly used SWfMS. After selecting the Galaxy repositories, we attempt to explore the workflows and encounter several challenges in reusing them. We classify the reusability status (reusable/nonreusable). Based on the effort level, we further categorize the reusable workflows (reusable without modification, easily reusable, moderately difficult to reuse, and difficult to reuse). Upon failure, we record the associated challenges that prevent reusability. We also list the actions upon success. The challenges preventing reusability include tool upgrading, tool support unavailability, design flaws, incomplete workflows, failure to load a workflow, etc. We need to perform several actions to overcome the challenges. The actions include identifying proper input datasets, updating/upgrading tools, finding alternative tools support for obsolete tools, debugging to find the issue creating tools and connections and solving them, modifying tools connections, etc. Such challenges and our action list offer guidelines to future workflow composers to create better workflows with enhanced reusability. A SWfMS stores provenance data at different phases of a workflow life cycle, which can help workflow construction. This provenance data allows reproducibility and knowledge reuse in the scientific community. But, this provenance information is usually many times larger than the workflow and input data, and managing provenance data is growing in complexity with large-scale applications. In our second study, we document the challenges of provenance management and reuse in e-science, focusing primarily on scientific workflow approaches by exploring different SWfMSs and provenance management systems. We also investigate the ways to overcome the challenges. Creating a workflow is difficult but essential for data-intensive complex analysis, and the existing workflows have several challenges to be reused, so in our third study, we build a recommendation system to recommend tool(s) using machine learning approaches to help scientists create optimal, error-free, and efficient workflows by using existing reusable workflows in Galaxy workflow repositories. The findings from our studies and proposed techniques have the potential to simplify the data-intensive analysis, ensuring reliability and efficiency.

Reading through code, finding relevant methods, classes and files takes a significant portion of software development time. Having good tool support for this code browsing activity can reduce human effort and increase overall developer productivity. To help with program comprehension activities, building an abstract code summary of a software system from its call graph is an active research area. A call graph is a visual representation of the caller-callee relationships between different methods of a software system. Call graphs can be difficult to comprehend for a large code-base. Previous work by Gharibi et al. on abstract code summarizing suggested using the Agglomerative Hierarchical Clustering (AHC) tree for understanding the codebase. Each node in the tree is associated with the top five method names. When we replicated the previous approach, we observed that the number of nodes in the AHC tree is burdensome for developers to explore. We also noticed only five method names for each node is not sufficient to comprehend an abstract node. We propose a technique to transform the AHC tree using cluster flattening for natural grouping and reduced nodes. We also generate a natural text summary for each abstract node derived from method comments. In order to evaluate our proposed approach, we collected developers’ opinions about the abstract code summary tree based on their codebase. The evaluation results confirm that our approach can not only help developers get an overview of their codebases but also could assist them in doing specific software maintenance tasks.

Colloidally-stable zero valent iron nanoparticles (nZVI) were synthesized through a classical redox reaction of iron sulfate with minor modifications using cellulose nanocrystals (CNCs) as stabilizers. We obtained spherical nZVI particles with high surface roughness and a mean size of 130nm. Particles remain colloidally stable after more than two months. Cellulose nanocrystals play a dual role in nZVI stability: a foreign surface to encourage stable nucleation over fast aggregation and a stabilizer to prevent iron nanoparticles aggregating into fractal colloids. Our results highlight the impact of the presence of CNCs on the rates and mechanisms of nucleation, growth, aggregation, and aging of nZVI particles, indicating promise in controlling size and morphology of similarly redox-generated nanoparticles. Cellulose nanocrystal-stabilized nZVI nanoparticles demonstrate properties well-suited for enhanced soil and groundwater remediation. //Nanocomposites composed of carboxylated cellulose nanocrystals and iron (Fe-oxCNC) were prepared through a classical redox reaction of iron sulfate using TEMPO-oxidized cellulose nanocrystals (oxCNCs) as a template and stabilizer. Morphological control over Fe-oxCNC nanoparticles was realized by varying the amount of oxCNC added to the redox process. As the molar ratio between oxCNC and Fe was increased from 1 to 8, the morphology of Fe-oxCNC nanoparticles evolved from rounded iron aggregates supported by cellulose nanocrystals to thin film iron-coatings on cellulose nanocrystals. Transmission electron microscopy (TEM), Inductively Coupled Plasma Optical Emission Spectrometry (ICP-OES), and chemical analyses (EDX, EELS) revealed that oxCNCs were coated by iron. Small changes to the density and type of functional groups on the CNC surface have large impacts on the morphology and the oxidation state of adsorbed iron nanoparticles.

Methane emissions from boreal and arctic wetlands, lakes, and rivers are expected to increase in response to warming and associated permafrost thaw. However, the lack of appropriate land cover datasets for scaling field-measured methane emissions to circumpolar scales has contributed to a large uncertainty for our understanding of present-day and future methane emissions. Here we present the Boreal–Arctic Wetland and Lake Dataset (BAWLD), a land cover dataset based on an expert assessment, extrapolated using random forest modelling from available spatial datasets of climate, topography, soils, permafrost conditions, vegetation, wetlands, and surface water extents and dynamics. In BAWLD, we estimate the fractional coverage of five wetland, seven lake, and three river classes within 0.5 × 0.5∘ grid cells that cover the northern boreal and tundra biomes (17 % of the global land surface). Land cover classes were defined using criteria that ensured distinct methane emissions among classes, as indicated by a co-developed comprehensive dataset of methane flux observations. In BAWLD, wetlands occupied 3.2 × 106 km2 (14 % of domain) with a 95 % confidence interval between 2.8 and 3.8 × 106 km2. Bog, fen, and permafrost bog were the most abundant wetland classes, covering ∼ 28 % each of the total wetland area, while the highest-methane-emitting marsh and tundra wetland classes occupied 5 % and 12 %, respectively. Lakes, defined to include all lentic open-water ecosystems regardless of size, covered 1.4 × 106 km2 (6 % of domain). Low-methane-emitting large lakes (>10 km2) and glacial lakes jointly represented 78 % of the total lake area, while high-emitting peatland and yedoma lakes covered 18 % and 4 %, respectively. Small (<0.1 km2) glacial, peatland, and yedoma lakes combined covered 17 % of the total lake area but contributed disproportionally to the overall spatial uncertainty in lake area with a 95 % confidence interval between 0.15 and 0.38 × 106 km2. Rivers and streams were estimated to cover 0.12  × 106 km2 (0.5 % of domain), of which 8 % was associated with high-methane-emitting headwaters that drain organic-rich landscapes. Distinct combinations of spatially co-occurring wetland and lake classes were identified across the BAWLD domain, allowing for the mapping of “wetscapes” that have characteristic methane emission magnitudes and sensitivities to climate change at regional scales. With BAWLD, we provide a dataset which avoids double-accounting of wetland, lake, and river extents and which includes confidence intervals for each land cover class. As such, BAWLD will be suitable for many hydrological and biogeochemical modelling and upscaling efforts for the northern boreal and arctic region, in particular those aimed at improving assessments of current and future methane emissions. Data are freely available at https://doi.org/10.18739/A2C824F9X (Olefeldt et al., 2021).

Climate change is raising average temperatures, altering rain and snowfall patterns, thawing permafrost, increasing the frequency and intensity of extreme weather, and raising sea levels within Canada. This is leading to specific impacts for water resources, which affect water availability and quality. The nature and timing of water-related hazards such as floods and droughts are also exacerbated. Climate change, combined with human developments, have led to a range of impacts and risks that intersect with water systems, including damaged infrastructure, deteriorating water quality, and ecosystem impacts. These impacts and risks affect Canadians in a different ways depending on their geographic location, sector of interest, and time of year. Canadian organizations and institutions that manage water-related risks require easily accessible climate data necessary to understand these impacts and risks and prioritize adaptation efforts. The Canadian Centre for Climate Services (CCCS) offers climate data, tools and expert advice, free to all Canadians. This presentation will introduce the CCCS and the array of “best-in-class” climate data made available by researchers from across the country that can support impact and risk assessment and adaptation planning. The presentation will focus on the CCCS’ products and services including a demonstration of the CCCS website and ClimateData.ca, which provides the ability to view and download climate data by watershed. Attendees of this presentation will leave with a better understanding of why the consideration of climate information is important and where and how they can access the data themselves.

The Influence of Shade and Complex Topography on the Classification, Climate, and Energy Balance of a Sub-alpine Wetland

Snowmelt contributions to streamflow in mid-latitude mountain basins typically dominate other runoff sources on annual and seasonal timescales. Future increases in temperature and changes in precipitation will affect both snow accumulation and seasonal runoff timing and magnitude, but the underlying and fundamental roles of mountain basin geometry and hypsometry on snowmelt sensitivity have received little attention. To investigate the role of basin geometry in snowmelt sensitivity, a linear snow accumulation model and the Cold Regions Hydrological Modeling (CRHM) platform driven are used to estimate how hypsometry affects basin-wide snow volumes and snowmelt runoff. Area-elevation distributions for fifty basins in western Canada were extracted, normalized according to their elevation statistics, and classified into three clusters that represent top-heavy, middle, and bottom-heavy basins. Prescribed changes in air temperature alter both the snow accumulation gradient and the total snowmelt energy, leading to snowpack volume reductions (10–40%), earlier melt onsets (1–4 weeks) and end of melt season (3 weeks), increases in early spring melt rates and reductions in seasonal areal melt rates (up to 50%). Basin hypsometry controls the magnitude of the basin response. The most sensitive basins are bottom-heavy, and have a greater proportion of their area at low elevations. The least sensitive basins are top-heavy, and have a greater proportion of their area at high elevations. Basins with similar proportional areas at high and low elevations fall in between the others in terms of sensitivity and other metrics. This work provides context for anticipating the impacts of ongoing hydrological change due to climate change, and provides guidance for both monitoring networks and distributed modeling efforts.

Quantifying water and energy fluxes are critical to understand how water is moved and stored on the landscape. These measurements are important for flood and drought forecasting, water resources management, and large-scale numerical weather prediction models. Moreover, land surface model’s (LSMs) which are hydrological tools used to predict and forecast water and energy fluxes, rely on these measurements to calibrate and validate their predictions. To evaluate hydrological fluxes and in turn water storage, representative observations are needed to capture the temporal and spatial dynamics of water on the landscape. However, hydrological fluxes are often difficult to measure and are limited to specific fluxes and spatial resolutions. Geological Weighing Lysimeters (GWL) are novel instruments that provide measurements of total integrated water storage at scales of 102 m2 and 106 m2 (field-scale). These tools use a saturated formations response to changes in mechanical loading, to estimate the change of water storage on the land surface. This research assessed the efficacy of a GWL in a deep confined aquifer at a research site in Duck Lake, Saskatchewan, to measure total water storage and partition individual stores from field-scale water balance. We found when coupled with supplementary observations of shallow groundwater and snow storage, the GWL provided a reliable record of temporal storage dynamics observed in point scale dielectric probes. Inconsistencies in soil moisture storage were from the dielectric probes inability to measure ice content in the soils and different estimates of hydrological fluxes between scales. We then used these storage estimates to critically assess the performance of two LSMs: the Canadian Land Surface Scheme (CLASS) and the Structure for Unifying Multiple Modeling Alternatives: (SUMMA). We found each LSM was able to reproduce total water storage and subsurface storage dynamics well, however they both had major inconsistencies simulating snowpack dynamics and hydrological fluxes. We speculate these inconsistencies are the result of differences in soil hydraulic property representations. The outcome of this research is two-fold. First, GWL and supplementary observations can be used to partition individual storage components from the water balance providing insight into hydrological fluxes; and secondly, small differences in soil hydraulic properties may largely influence Land Surface Schemes (LSSS’s) simulated fluxes, more research is needed to assess the influence have.

Aufeis, also known as an icing or naled, is an accumulation of ice that forms primarily during winter when water is expelled onto frozen ground or ice surfaces and freezes in layers. Process-oriented aufeis research initially expanded in the 20th century, but recent interest in changing hydrological conditions in permafrost regions has rejuvenated this field. Despite its societal relevance, the controls on aufeis distribution and dynamics are not well defined and this impedes projections of variation in aufeis size and distribution expected to accompany climate change. This paper reviews the physical controls on aufeis development, current broad-scale aufeis distribution and anticipated change, and approaches to aufeis investigation. We propose an adjustment to terminology to better distinguish between the formation process and resulting ice bodies, a clarification of the aufeis classification approach based on source water, and a size threshold for broad-scale aufeis inventory to facilitate collaborative research. We identify additional objectives for future research including advancing process knowledge at fine spatial scales, describing broad-scale distribution using current remote sensing capabilities, and improving our understanding and predictive capacity over the interactions between aufeis and landscape-scale permafrost, hydrogeological, geotectonic, and climate conditions.

Performance in simulating atmospheric rivers (ARs) over western North America based on AR frequency and landfall latitude is evaluated for 10 models from phase 5 of the Coupled Model Intercomparison Project among which the CanESM2 model performs well. ARs are classified into southern, northern, and middle types using self-organizing maps in the ERA-Interim reanalysis and CanESM2. The southern type is associated with the development and eastward movement of anomalous lower pressure over the subtropical eastern Pacific, while the northern type is linked with the eastward movement of anomalous cyclonic circulation stimulated by warm sea surface temperatures over the subtropical western Pacific. The middle type is connected with the negative phase of North Pacific Oscillation–west Pacific teleconnection pattern. CanESM2 is further used to investigate projected AR changes at the end of the twenty-first century under the representative concentration pathway 8.5 scenario. AR definitions usually reference fixed integrated water vapor or integrated water vapor transport thresholds. AR changes under such definitions reflect both thermodynamic and dynamic influences. We therefore also use a modified AR definition that isolates change from dynamic influences only. The total AR frequency doubles compared to the historical period, with the middle AR type contributing the largest increases along the coasts of Vancouver Island and California. Atmospheric circulation (dynamic) changes decrease northern AR type frequency while increasing middle AR type frequency, indicating that future changes of circulation patterns modify the direct effect of warming on AR frequency, which would increase ARs (relative to fixed thresholds) almost everywhere along the North American coastline.

East of the Continental Divide in the cold interior of Western Canada, the Mackenzie and Nelson River basins have some of the world's most extreme and variable climates, and the warming climate is changing the landscape, vegetation, cryosphere, and hydrology. Available data consist of streamflow records from a large number (395) of natural (unmanaged) gauged basins, where flow may be perennial or temporary, collected either year-round or during only the warm season, for a different series of years between 1910 and 2012. An annual warm-season time window where observations were available across all stations was used to classify (1) streamflow regime and (2) seasonal trend patterns. Streamflow trends were compared to changes in satellite Normalized Difference Indices. Clustering using dynamic time warping, which overcomes differences in streamflow timing due to latitude or elevation, identified 12 regime types. Streamflow regime types exhibit a strong connection to location; there is a strong distinction between mountains and plains and associated with ecozones. Clustering of seasonal trends resulted in six trend patterns that also follow a distinct spatial organization. The trend patterns include one with decreasing streamflow, four with different patterns of increasing streamflow, and one without structure. The spatial patterns of trends in mean, minimum, and maximum of Normalized Difference Indices of water and snow (NDWI and NDSI) were similar to each other but different from Normalized Difference Index of vegetation (NDVI) trends. Regime types, trend patterns, and satellite indices trends each showed spatially coherent patterns separating the Canadian Rockies and other mountain ranges in the west from the poorly defined drainage basins in the east and north. Three specific areas of change were identified: (i) in the mountains and cold taiga-covered subarctic, streamflow and greenness were increasing while wetness and snowcover were decreasing, (ii) in the forested Boreal Plains, particularly in the mountainous west, streamflows and greenness were decreasing but wetness and snowcover were not changing, and (iii) in the semi-arid to sub-humid agricultural Prairies, three patterns of increasing streamflow and an increase in the wetness index were observed. The largest changes in streamflow occurred in the eastern Canadian Prairies.

Using only warm season streamflow records, regime and change classifications were produced for ~ 400 watersheds in the Nelson and Mackenzie River basins, and trends in water storage and vegetation were detected from satellite imagery. Three areas show consistent changes: north of 60° (increased streamflow and basin greenness), in the western Boreal Plains (decreased streamflow and basin greenness), and across the Prairies (three different patterns of increased streamflow and basin wetness).

Model calibration and validation are critical in hydrological model robustness assessment. Unfortunately, the commonly used split-sample test (SST) framework for data splitting requires modelers to make subjective decisions without clear guidelines. Unlike most SST studies that use two sub-periods (i.e., calibration and validation) to build models, this study incorporates an independent model testing period in addition to calibration and validation periods. Two hydrological models are calibrated and tested in 463 CAMELS catchments across the United States using 50 different data splitting schemes. These schemes are established regarding the data availability, length, and data recentness of the continuous calibration sub-periods (CSPs). A full-period CSP is also included in the experiment, which skips model validation entirely. The results are synthesized regarding the large sample of catchments and are comparatively assessed in multiple novel ways, including how model building decisions are framed as a decision tree problem and viewing the model validation process as a formal testing period classification problem, aiming to accurately predict model success/failure in the testing period. Results span different climate and catchments make conclusions generalizable. Strong patterns show that calibrating models to older data and then validating models on newer data produces inferior model testing period performance and should hence be avoided. Calibrating to the full available data and skipping model validation entirely is the most robust split-sample decision. Results strongly support revising the traditional split-sample approach in hydrological modeling.

Model calibration and validation are critical in hydrological model robustness assessment. Unfortunately, the commonly used split-sample test (SST) framework for data splitting requires modelers to make subjective decisions without clear guidelines. This large-sample SST assessment study empirically assesses how different data splitting methods influence post-validation model testing period performance, thereby identifying optimal data splitting methods under different conditions. This study investigates the performance of two lumped conceptual hydrological models calibrated and tested in 463 catchments across the United States using 50 different data splitting schemes. These schemes are established regarding the data availability, length and data recentness of continuous calibration sub-periods (CSPs). A full-period CSP is also included in the experiment, which skips model validation. The assessment approach is novel in multiple ways including how model building decisions are framed as a decision tree problem and viewing the model building process as a formal testing period classification problem, aiming to accurately predict model success/failure in the testing period. Results span different climate and catchment conditions across a 35-year period with available data, making conclusions quite generalizable. Calibrating to older data and then validating models on newer data produces inferior model testing period performance in every single analysis conducted and should be avoided. Calibrating to the full available data and skipping model validation entirely is the most robust split-sample decision. Experimental findings remain consistent no matter how model building factors (i.e., catchments, model types, data availability, and testing periods) are varied. Results strongly support revising the traditional split-sample approach in hydrological modeling. Key Points - A unique split-sample experiment is performed across 463 catchments to provide guidance on split sample decision-making in model calibration - Calibrating models to the full available data period and skipping model validation entirely is the most robust choice - Calibrating models to older data and then validating models on newer data, a very common approach in literature, is an inferior choice Plain Language Summary Hydrological model calibration is a critical model building process that infers key model parameter values from observed system response data. Conventionally, this process requires the historical period to be split into a calibration period for tuning parameters and a validation period for testing model robustness (i.e., the split-sample). Unfortunately, there is a lack of empirical evidence supporting how exactly to define the split-sample. We designed an exhaustive and novel experiment comparing the range of possible split-sampling schemes, including calibrating to older/recent years, calibrating to a short/long period, and calibrating to the full period of available system response data. Each scheme was evaluated based on performance, assessed in three different ways, in numerous post-validation model testing periods for each of the 926 calibration case studies (two different hydrological models applied in 463 catchments). Results show that using older data for model calibration and then using newer data for validation, which is the typical practice in the literature, is an inferior choice and should be avoided. The results also show that calibrating to the full historical data and skipping model validation entirely is the most robust choice. Therefore, the split-sample approach applied in this community for decades should be revised.

Wildfires, which constitute the most extensive natural disturbance of the boreal biome, produce a broad range of ecological impacts to vegetation and soils that may influence post-fire vegetation assemblies and seedling recruitment. We inventoried post-fire understory vascular plant communities and tree seedling recruitment in the northwestern Canadian boreal forest and characterized the relative importance of fire effects and fire history, as well as non-fire drivers (i.e., the topoedaphic context and climate), to post-fire vegetation assemblies. Topoedaphic context, pre-fire forest structure and composition, and climate primarily controlled the understory plant communities and shifts in the ranked dominance of tree species (***8% and **13% of variance explained, respectively); however, fire and fire-affected soils were significant secondary drivers of post-fire vegetation. Wildfire had a significant indirect effect on understory vegetation communities through post-fire soil properties (**5%), and fire history and burn severity explained the dominance shifts of tree species (*7%). Fire-related variables were important explanatory variables in classification and regression tree models explaining the dominance shifts of four tree species (R2 = 0.43–0.65). The dominance of jack pine (Pinus banksiana Lamb.) and trembling aspen (Populus tremuloides Michx.) increased following fires, whereas that of black spruce (Picea mariana (Mill.) BSP.) and white spruce (Picea glauca (Moench) Voss) declined. The overriding importance of site and climate to post-fire vegetation assemblies may confer some resilience to disturbed forests; however, if projected increases in fire activity in the northwestern boreal forest are borne out, secondary pathways of burn severity, fire frequency, and fire effects on soils are likely to accelerate ongoing climate-driven shifts in species compositions.

Existing software engineering tools have proved useful in automating some aspects of the code review process, from uncovering defects to refactoring code. However, given that software teams still spend large amounts of time performing code reviews despite the use of such tools, much more research remains to be carried out in this area. This dissertation present two major contributions to this field. First, we perform a text classification experiment over thirty thousand GitHub review comments to understand what code reviewers typically discuss in reviews. Next, in an attempt to offer an innovative, data-driven approach to automating code reviews, we leverage probabilistic models of source code and graph embedding techniques to perform human-like code inspections. Our experimental results indicate that the proposed algorithm is able to emulate human-like code inspection behaviour in code reviews with a macro f1-score of 62%, representing an impressive contribution towards the relatively unexplored research domain of automated code reviewing tools.

Soulis, E.D., K.R. Snelgrove, N. Kouwen, F. Seglenieks and D. Verseghy, (2000). Towards closing the vertical water balance in Canadian atmospheric models: Coupling of the land surface scheme CLASS with the distributed hydrological model WATFLOOD. Atmosphere-Ocean 38, 251-269.

Towards closing the vertical water balance in Canadian atmospheric models: Coupling of the land surface scheme CLASS with the distributed hydrological model WATFLOOD

Glacier melt is an important fresh water source. Seasonal changes can have impacting consequences on downstream water resources management. Today’s glacier monitoring lacks an observation-based tool for regional, sub-seasonal observation of glacier mass balance and a quantification of associated meltwater release at high temporal resolution. The snowline on a glacier marks the transition between the ice and snow surface, and is, at the end of the summer, a proxy for the annual glacier mass balance. It was shown that glacier mass balance model simulations closely tied to sub-seasonal snowline observations on optical satellite sensors are robust for the observation date. Recent advances in remote sensing permit efficient and extensive snowline mapping. Different methods automatically discriminate snow over ice on high- to medium-resolution optical satellite images. Other studies rely on lower ground resolution optical imagery to retrieve snow cover fraction at pixel level and produce regional maps of snow cover extent. However, state-of-the-art methods using optical sensors still have important shortcomings, such as cloud-cover related issues. Images acquired by Synthetic Aperture Radar (SAR), which are almost insensitive to cloud coverage, have proofed suitable for transient snowline delineation. The combination of SAR and optical data in a complementary way carries a unique potential for a better monitoring of snow depletion on high temporal and spatial resolution. The aim of this work is to map snow cover over glaciers by combining Sentinel-1 SAR, Sentinel-2 multispectral and lower resolution MODIS images. Consecutively, we developed an approach that can automatically handle classification of multi-source and multi-resolution satellite image stacks. This provides a unique solution for continuous snowline mapping since the beginning of the century. With the provided close-to-daily transient snow cover fractions on glacier level, we provide the basis for a new strategy to directly integrate multi-source satellite image classification into glacier mass balance monitoring.

Land surface models have many parameters that have a spatially variable impact on model outputs. In applying these models, sensitivity analysis (SA) is sometimes performed as an initial step to select calibration parameters. As these models are applied to large domains, performing sensitivity analysis across the domain is computationally prohibitive. Here, using a Variable Infiltration Capacity model (VIC) deployment to a large domain as an example, we show that watershed classification based on climatic attributes and vegetation land cover helps to identify the spatial pattern of parameter sensitivity within the domain at a reduced cost. We evaluate the sensitivity of 44 VIC model parameters with regard to streamflow, evapotranspiration and snow water equivalent over 25 basins with a median size of 5078 km2. Basins are clustered based on their climatic and land cover attributes. Performance in transferring parameter sensitivity between basins of the same cluster is evaluated by the F1 score. Results show that two donor basins per cluster are sufficient to correctly identify sensitive parameters in a target basin, with F1 scores ranging between 0.66 (evapotranspiration) and 1 (snow water equivalent). While climatic attributes are sufficient to identify sensitive parameters for streamflow and evapotranspiration, including the vegetation class significantly improves skill in identifying sensitive parameters for the snow water equivalent. This work reveals that there is opportunity to leverage climate and land cover attributes to greatly increase the efficiency of parameter sensitivity analysis and facilitate more rapid deployment of land surface models over large spatial domains.

Uncertainties of snowpack models and of their meteorological forcings limit their use by avalanche hazard forecasters, or for glaciological and hydrological studies. The spatialized simulations currently available for avalanche hazard forecasting are only assimilating sparse meteorological observations. As suggested by recent studies, their forecasting skills could be significantly improved by assimilating satellite data such as snow reflectances from satellites in the visible and the near-infrared spectra. Indeed, these data can help constrain the microstructural properties of surface snow and light absorbing impurities content, which in turn affect the surface energy and mass budgets. This paper investigates the prerequisites of satellite data assimilation into a detailed snowpack model. An ensemble version of Météo-France operational snowpack forecasting system (named S2M) was built for this study. This operational system runs on topographic classes instead of grid points, so-called ‘semi-distributed’ approach. Each class corresponds to one of the 23 mountain massifs of the French Alps (about 1000 km2 each), an altitudinal range (by step of 300 m) and aspect (by step of 45°). We assess the feasability of satellite data assimilation in such a semi-distributed geometry. Ensemble simulations are compared with satellite observations from MODIS and Sentinel-2, and with in-situ reflectance observations. The study focuses on the 2013–2014 and 2016–2017 winters in the Grandes-Rousses massif. Substantial Pearson R2 correlations (0.75–0.90) of MODIS observations with simulations are found over the domain. This suggests that assimilating it could have an impact on the spatialized snowpack forecasting system. However, observations contain significant biases (0.1–0.2 in reflectance) which prevent their direct assimilation. MODIS spectral band ratios seem to be much less biased. This may open the way to an operational assimilation of MODIS reflectances into the Météo-France snowpack modelling system.

Accurate and frequent mapping of transient wetland inundation in the boreal region is critical for monitoring the ecological and societal functions of wetlands. Satellite Synthetic Aperture Radar (SAR) has long been used to map wetlands due to its sensitivity to surface inundation and ability to penetrate clouds, darkness, and certain vegetation canopies. Here, we track boreal wetland inundation by developing a two-step modified decision-tree algorithm implemented in Google Earth Engine using Sentinel-1 C-band SAR and Sentinel-2 Multispectral Instrument (MSI) time-series data as inputs. This approach incorporates temporal as well as spatial characteristics of SAR backscatter and is evaluated for the Peace-Athabasca Delta, Alberta (PAD), and Yukon Flats, Alaska (YF) from May 2017 to October 2019. Within these two boreal study areas, we map spatiotemporal patterns in wetland inundation classes of Open Water (OW), Floating Plants (FP), Emergent Plants (EP), and Flooded Vegetation (FV). Temporal variability, frequency, and maximum extents of transient wetland inundation are quantified. Retrieved inundation estimates are compared with in-situ field mapping obtained during the NASA Arctic-Boreal Vulnerability Experiment (ABoVE), and a multi-temporal Landsat-derived surface water map. Over the 2017–2019 study period, we find that fractional inundation area ranged from 18.0% to 19.0% in the PAD, and from 10.7% to 12.1% in the YF. Transient wetland inundation covered ~595 km2 of the PAD, comprising ~9.1% of its landscape, and ~102 km2 of the YF, comprising ~3.6%. The implications of these findings for wetland function monitoring, and estimating landscape-scale methane emissions are discussed, together with limitations and uncertainties of our approach. We conclude that time series of Sentinel-1 C-band SAR backscatter, screened with Sentinel-2 MSI optical imagery and validated by field measurements, offer a valuable tool for tracking transient boreal wetland inundation.

Microflow cytometers and many other miniaturized microfluidic devices have shown great potential in many fields, such as, particle detection, cell sorting and classification. A reliable signal analysis method is required to improve the measurement accuracy of the emerging microfluidic devices. In this paper, a novel method is presented to analyze the signal from microspheres with different diameters based on transit time and amplitude. Experimental results show that transit time threshold plays a more important role at lower flow rate for particle differentiation and can be used to improve the performance of a microflow cytometer.

Transit time and amplitude thresholds in a microflow cytometer for particle classification

Microflow cytometers and many other miniaturized microfluidic devices have shown great potential in many fields, such as, particle detection, cell sorting and classification. A reliable signal analysis method is required to improve the measurement accuracy of the emerging microfluidic devices. In this paper, a novel method is presented to analyze the signal from microspheres with different diameters based on transit time and amplitude. Experimental results show that transit time threshold plays a more important role at lower flow rate for particle differentiation and can be used to improve the performance of a microflow cytometer.

Zhang, Y., Xu, C. Q. (2020) The impact of transit time on a microflow cytometer for particle classification. 2020 Photonics North (PN), pp. 1-1, https://doi.org/10.1109/PN50013.2020.9167024.

Transit time and amplitude thresholds in a microflow cytometer for particle classification

Accurate classification of the phase of precipitation has important implications for understanding the vulnerability of future water resources in warming cold regions, where a substantial shift from snow to rain fraction is expected to occur over this century. In hydrological models, precipitation is typically partitioned into rainfall and snowfall using empirical near-surface air temperature thresholds methods. Surface air temperature methods require site-specific calibration and lack the physical basis to simulate this partitioning realistically and confidently for future climates. Advanced algorithms using the psychrometric energy balance of falling hydrometeors have demonstrated a solid physical basis and appear scalable across time and space but have only been applied in small basin studies. This study evaluated the impact of the choice of precipitation phase prediction technique on the historical and future hydrology of the 406,000 km2 Saskatchewan River Basin (SRB), from its headwaters in the Canadian Rockies to its lower reaches in the Canadian Prairies and boreal forest. It did so using meteorological data (1971-2100) to force a continental scale hydrological land surface scheme with water management, MESH. Using the psychrometric energy balance phase partitioning method, the changing characteristics of snowfall and rainfall were shown to cause substantial changes in runoff generation mechanisms and streamflow over time that depended on elevation and ecozone. The historical period results using the energy balance-based phase partitioning approach resulted in high snow fractions, particularly at high elevations, that are consistent with observations. To evaluate the uncertainty introduced by using air temperature phase methods, the MESH model of the SRB was falsified and run with different versions of these methods with various parameters. Only one of the surface-temperature-based methods showed adequate partitioning results at high elevations. The uncertainty introduced by temperature-based phase partitioning methods was substantial for the historical period and became even more pronounced in the future, suggesting that empirical phase partitioning methods are a large and unnecessary contributor of uncertainty for large scale hydrological predictions of climate change impacts.

Lake ice coverage products are a requirement identified by the climate community for improving numerical weather prediction and atmospheric reanalysis products, as well as for climate monitoring as determined by the Global Climate Observing System (GCOS). There are many suitable sources of observations available for mapping and monitoring lake ice coverage such as optical satellite data with the most practical ones from the Moderate Resolution Imaging Spectroradiometer (MODIS) over the last two decades. Considering the limitation of the presence of cloud cover and daylight dependency to capture imagery by optical sensors, the high revisit time of NASA’s Terra and Aqua satellites that carry MODIS allows for the production of lake ice maps required for operational and research-based projects. Building on our previous research findings concluded from a GWF-supported project on lake ice cover mapping of Lake Erie from RADARSAT data, we are proposing a method to characterize inherent uncertainties (aleatoric) and model uncertainties (epistemic) for the production of daily lake ice maps. Random Forest (RF) is used for classifying lake ice, water, and cloud and for measuring and quantifying predictive uncertainty. As RF is an ensemble-based approach, it allows learning different hypotheses (different trees); and therefore, it provides different expected outcome. The total uncertainty in a prediction can be calculated by the (Shannon) entropy of the predictive posterior distribution, whereas calculating the entropy of each probability distribution and then computing the average gives the aleatoric uncertainty. Epistemic uncertainty is then calculated by subtracting aleatoric from total uncertainties. Uncertainty estimates expands product usability, making researchers aware of aleatoric and epistemic uncertainty when incorporating ice fractions in their physical/numerical lake models in the form of direct integration of observation error variance or as a quality control flag.

Cavaliere, E., Baulch, H., Basu, N., Wolfe, J., Hoggarth, C., & Hergott, A. (2019). Understanding nutrient retention in prairie wetlands using catchment classes and nutrient regimes. Global Water Futures Annual Science Meeting. Saskatoon, SK Canada. May 15-17, 2019 Conference Poster

Understanding nutrient retention in prairie wetlands using catchment classes and nutrient regimes

Discontinuous permafrost regions are experiencing a change in land cover distribution as a result of permafrost thaw. In wetlands interspersed with discontinuous permafrost, climate change is particularly problematic because temperature increases can result in significant permafrost thaw, thaw-driven landscape changes, and resultant changes in watershed hydrologic responses. The influence of land cover change on the short- and long-term hydrological responses of wetland-peatland complexes is poorly understood. A better understanding of the impacts of climate-related land cover evolution on the hydrology of wetland-covered watersheds requires information about the distribution of hydrologically important lands, their pattern, and the rate at which they change over time. Here, we first developed a machine learning-based land cover evolution model (TSLCM) to estimate the long-term evolution of dominant land covers for application to the discontinuous permafrost regions of Northern Canada. This model is applied to replicate historical land cover and estimate future land cover scenarios at the Scotty Creek Research Basin in the Northwest Territories, Canada. A significant challenge when analyzing land cover change effects on hydrological properties is generating time-dependent classified maps of the region of interest, and the challenges associated with preprocessing remotely sensed data for discriminating between wetlands and forest-covered regions. In this work, we focus on two important objectives supporting the improved classification of wetlands in discontinuous permafrost regions: the exclusive use of only RGB imagery, and the use of an image segmentation method to accelerate the automatic classification of land cover. A semantic segmentation neural network, a multi-layer perceptron (MLP), and watershed function algorithms are applied to develop the taiga wetland identification neural network (TWINN) for the hydrological classification of wetlands. TWINN is here demonstrated to accurately classify high-resolution imagery of discontinuous permafrost regions within the Northwest Territories into the water, forest, and wetlands, and also able to delineate the runoff area of wetlands. To study the effect of land cover evolution on runoff generation in the Scotty Creek basin, the products of TWINN and TSLCM are used to inform a process-based hydrological model where land cover change is represented explicitly. According to simulation results, land cover transitions can modify annual mean streamflow by as much as 7%, in addition to influences due to changing precipitation regimes alone.

After oil spills occur, regulators require adequate information to select best practices to minimize impacts on environments and to remediate target freshwater ecosystems. Zooplankton are valuable indicators of structure and function of aquatic ecosystems since they play pivotal roles in biochemical cycles while stabilizing food webs. Traditional identification of zooplankton can be costly and time-consuming, while also being difficult to standardize. Compared with classification of individuals by identification, based on visual inspection of morphology, metabarcoding of DNA and or RNA has promise for cost-effective high-throughput and benchmarkable biomonitoring of zooplankton communities. These identification methods were applied in the context of assessing responses of the zooplankton community exposed to simulated spills of diluted bitumen (dilbit), with concurrent exposure of experimental remediation practices of enhanced monitored natural recovery and shoreline cleaner application. The objective of this study was also to apply DNA and RNA metabarcoding of zooplankton for ecotoxicological assessment and compare it with traditional morphological identification in experimental shoreline enclosures in a boreal lake. Metabarcoding detected 77.4% of the morphologically identified boreal zooplankton taxa down to the genus level, with a total of 24 shared genera. Metabarcoding-based relative abundance of shared genus also served as an acceptable proxy for biomass inferred by morphological identification at the genera-level. Overall, both DNA and RNA metabarcoding determined significant differences between genera richness between the no treatment enclosure and shoreline cleaner application, while morphological identification determined no difference. DNA metabarcoding determined overall differences in community composition between no treatment and treatments, shoreline cleaner application and enhanced monitored natural recovery, while RNA metabarcoding and morphological identification determined differences between one or the other. Shoreline cleaner application overall seemed to have the greatest effect on zooplankton communities relative to enhanced monitored natural recovery, regardless of zooplankton identification method. Both metabarcoding and morphological identification were able to discern the differences between the two experimental remediation practices. Metabarcoding of zooplankton can provide informative results for ecotoxicological assessment of remediation practices of dilbit, advancing our knowledge of best practices for remediating oil-impacted aquatic ecosystems while serving to accelerate the assessment of at-risk freshwater ecosystems.

Space-based, global-extent digital elevation models (DEMs) are key inputs to many Earth sciences applications. However, many of these applications require the use of a ‘bare-earth’ DEM versus a digital surface model (DSM), the latter of which may include systematic positive biases due to tree canopies in forested areas. Critical topographic features may be obscured by these biases. Vegetation-free datasets have been created by using statistical relationships and machine learning to train on local-scale datasets (e.g., lidar) to debias the global-extent datasets. Recent advances in satellite platforms coupled with an increased availability of computational resources and lidar reference products has allowed for a new generation of vegetation- and urban-canopy removals. One of these is the Forest And Buildings removed Copernicus DEM (FABDEM), based on the most recent and most accurate global DSM Copernicus-30. Amongst the more challenging landscapes to quantify surface elevations are dense forested mountain catchments where even airborne lidar applications struggle to capture surface returns. The increasing affordability and availability of UAV-based lidar platforms has resulted in new capacity to fly modest spatial extents with unrivalled point densities. These data allow an unprecedented ability to validate global sub-canopy DEMs against representative UAV-based lidar data. In this work, the FABDEM is validated against an up-scaled lidar data in a steep and forested mountain catchment considering elevation, slope, and Terrain Position Index (TPI) metrics. Comparisons of FABDEM with SRTM, MERIT, and the Copernicus-30 dataset are made. It was found that the FABDEM had a 24% reduction in elevation RMSE and 135% reduction in bias compared to the Copernicus-30 dataset. Overall, the FABDEM provides a clear improvement over existing de-forested DEM products in complex mountain topography such as the MERIT DEM. This study supports the use of FABDEM in forested mountain catchments as the current best-in-class data product.

Variation among land classification units in the NWT

Peatlands in the Rocky Mountains most commonly occur in valley bottoms and are classified as fens. Understanding how fens influence water storage and water release is essential for better predicting water availability as the climate changes. Peatlands located in mountain regions tend to have a complex soil profile due to the geomorphologically dynamic environment. There is little information on the water storage capacity of mountain peatlands. To address this knowledge gap, the water storage capacity of a fen peatland with a complex soil profile in the Canadian Rocky Mountains, Alberta, Canada, was studied. Using the water table fluctuation method, vertical variations in specific yield were estimated. The influence of several factors – soil profile complexity, vegetation cover, water table depth, and seasonality – on specific yield were determined. Results showed that soil profile complexity plays a vital role in determining the spatial variability of vertical specific yield. The effect of stratigraphy on specific is important because it demonstrates that active geomorphic environments (often found in mountain regions) are a crucial piece of information required to determine the water storage capacity of mountain fens. The seasonality analysis results show that the overall wetness of a given year or time during the growing season influences the water table depth and response to rainfall events, thus exerting a control on specific yield. The impact of seasonality is also important because it reveals that even small changes to weather patterns can impact water storage in mountain peatlands. Overall, the research yielded new insights into how much water is stored in and released from mountain fens, information which is useful to improving regional hydrological models and predicting hydrological impacts of climate change or geomorphic events.

Flow management has the potential to significantly affect ecosystem condition. Shallow lakes in arid regions are especially susceptible to flow management changes, which can have important implications for the formation of cyanobacterial blooms. Here, we reveal water quality shifts associated with changing source water inflow management. Using in situ monitoring data, we studied a seven-year time span during which inflows to a shallow, eutrophic drinking water reservoir transitioned from primarily natural landscape runoff (2014–2015) to managed flows from a larger upstream reservoir (Lake Diefenbaker; 2016–2020) and identified significant changes in cyanobacteria (as phycocyanin) using generalized additive models to classify cyanobacterial bloom formation. We then connected changes in water source with shifts in chemistry and the occurrence of cyanobacterial blooms using principal components analysis. Phycocyanin was greater in years with managed reservoir inflow from a mesotrophic upstream reservoir (2016–2020), but dissolved organic matter (DOM) and specific conductivity, important determinants of drinking water quality, were greatest in years when landscape runoff dominated lake water source (2014–2015). Most notably, despite changing rapidly, it took multiple years for lake water to return to a consistent and reduced level of DOM after managed inflows from the upstream reservoir were resumed, an observation that underscores how resilience may be hindered by weak resistance to change and slow recovery. Environmental flows for water quality are rarely defined, yet we show that trade-offs exist between poor water quality via elevated conductivity and DOM and higher bloom risk, depending on water source. Our work highlights the importance of source water quality, not just quantity, to water security, and our findings have important implications for water managers who must protect ecosystem services while adapting to projected hydroclimatic change.

Hassan, Amina (2024) Watershed Classification in the Great Lakes Basin: Implications for Water Quality and Agricultural Management Practices, UWSpace - Theses, http://hdl.handle.net/10012/20623

Watershed Classification in the Great Lakes Basin: Implications for Water Quality and Agricultural Management Practices

Wetlands around the world are increasingly impacted by a shift in environmental conditions due to climate change, land use development, resource extraction, urbanization, and sea level rise, to name a few external pressures (Meng et al., 2016; Walpole and Davidson, 2018). These environmental changes can alter the hydrological regime, impacting the biogeochemical processes that govern important wetland ecosystem services, such as carbon sequestration and water storage. Biogeochemical processes in wetlands are highly dynamic (Reddy et al., 2010; Jackson et al., 2014) and involve complex interactions between hydrological processes, mineralogical transformations, bacterial and vegetation communities, and soil stores of carbon and nutrients (Cherry, 2011; U.S. EPA, 2015). Currently, our understanding of biogeochemical properties of wetlands are derived from mechanistic and statistical links between biological, geological, and chemical processes. However, how climatic and hydrological processes interact with wetland biogeochemical functions is still not well-understood. Wetland ecosystems maintain a fragile balance between soil, water, plant, microbial, and atmospheric processes, which regulates water flow and water quality (Reddy and Delaune, 2008). Even minor gradients (naturally or anthropogenically induced) in hydrological and climatic parameters (e.g., wetting and drying, flooding, freezing, and thawing, groundwater-surface water interactions, etc.) can change the ecology and (bio)geochemistry of wetlands. These changes can have profound impacts on globally important processes, such as greenhouse gas emissions. Within a wetland, there is a high degree of spatial and temporal heterogeneity of chemical properties, temperature, and water-saturation that regulates the transport and transformation of carbon, nutrients, and redox-active elements (Reddy et al., 2010; Cherry, 2011; Jackson et al., 2014). The heterogeneity results in both spatial and temporal pulses of biogeochemical activity, primarily associated with aerobic or anaerobic microbial respiration. Thus, wetlands are considered “biogeochemical hotspots” in the landscape, with an enhanced cycling of nutrients, carbon and trace metals (Megonigal, 2008; Reddy et al., 2010; Cherry, 2011). Quantifying the variability in process intensity remains challenging but is, however, critical to unravel the linkages between forcing environmental boundary conditions and biogeochemical responses. This Research Topic brings together wetland (bio)geochemists, hydrologists, biologists, ecologists, and soil scientists to share research in various areas of wetland biogeochemistry that addresses how current and future hydroclimatic conditions and land use change modulate (bio)geochemical processes in wetlands. The resultant collection of papers covers a broad snapshot of our understanding of how biogeochemical transformations and the movement of water in wetlands impacts the concentration and mobility of nutrients and contaminants, microbial community dynamics, greenhouse gas emissions, carbon cycling, and sequestration. The studies cover a range of freshwater wetland classifications (i.e., marsh, peatland, etc.) in natural, disturbed, and restored environments. The assembled papers provide important new information that addresses critical knowledge gaps on how wetland biogeochemistry is impacted by environmental change. The impact of environmental change on vegetation and subsequent impacts on biogeochemical cycling are apparent and several papers in this Research Topic showcase the interconnected nature of wetland biogeochemistry, ecology, and carbon cycling. Yuckin and Rooney quantified the effect of Phragmites australis invasion on carbon and macronutrient standing stocks in a freshwater coastal marsh, highlighting that plant invasions may create trade-offs between ecosystem processes. Yavitt et al. examined the importance of leaf litter in carbon cycling in peatlands and why vegetation, plant species composition, and peatland type must be determined to put peatland ecosystems into the context of global carbon budgets. Turner et al. demonstrated how wind sheltering influences land-atmosphere fluxes of carbon in wetlands and plays an important role on wetland ecosystem characteristics and energy balance. Koop-Jakobsen and Gutbrod observed distinctive spatio-temporal oxygen dynamics in an open tidal marsh that differs from the surrounding shaded vegetated marsh. The authors highlight the role of vascular plant canopies for biogeochemical processes in wetlands. It is evident from these papers that the feedbacks between vegetation and biogeochemical processes in wetlands provide an important mechanism that regulates essential ecosystem services. Several authors highlighted the complex controls and interactions between microbiological communities, biogeochemical processes, and hydrological setting. In coastal peatlands, Gosch et al. assessed the change in the decomposition of organic matter and releasing organic and inorganic solutes from peat under seawater intrusion. While in inland peatlands, Negassa et al. discussed the lack of empirical data on the high variability of chemical and biochemical properties in rewetted peatlands. Zak et al. examined if long-term peat mineralization during decades of drainage in minerotrophic fens causes an enrichment or a decline of enzyme-inhibiting polyphenols. While, Alshehri et al. evaluated the potential of adding phenolic compounds to peatland soils to inhibit extracellular enzyme activities in order to reduce the flux of CO2 from peatlands; the authors recommend phenolic enrichment as a potential peatland restoration strategy. Limpert et al. examined the CO2 and CH4 emissions and microbial community diversity during a wetland rehabilitation process (rewetting) and observed a clear succession of microbial communities during the dry-wet phases, suggesting that wetland hydrology plays a significant role in the microbial community structure. All of the publications in this Research Topic elucidate strong linkages between the hydro-physical setting and biogeochemical processes and recognized that understanding the feedbacks associated with these linkages requires further studies. With 42 authors from six countries in Europe, North America, and Australia/Oceania this Research Topic identifies key priorities for future research in wetlands biogeochemical transformation and processes. This Research Topic highlights the need for a more detailed understanding of interactions between and cycling of carbon, oxygen, nitrogen, phosphorus, and sulfur. The authors emphasize the role of redox-pH conditions, organic matter, microbial-mediated processes that drive nutrient and carbon transformations in wetlands, plant responses, and adaptation to wetland soil conditions. Importantly, the contributions emphasize the need for more integrated research efforts into the physical, hydrological and climatic processes that regulate wetland biogeochemical processes. The further development of interdisciplinary linkages is considered essential for a process-based understanding of wetland functions and the successful restoration and management of wetland ecosystems.

High-latitude forests of North America are characterized by their natural dependence on large and severe wildfires. However, these wildfires also pose a range of social, economic, and environmental risks, with growing concern regarding persistent effects on stream flow volume, seasonal timing of flow, water quality, aquatic ecosystem health, and downstream community drinking water treatment. Here, we present the outcomes of a comprehensive scoping review of post-fire hydrologic studies in high-latitude forests of North America (Canada and Alaska). Our objectives were to (1) create an inventory of studies on post-fire hydrologic effects on surface water; (2) analyze those studies in terms of watershed characteristics and the type and duration of hydrologic effects; (3) identify and evaluate the link between upstream hydrologic effects with hydrologic ecosystem services; and (4) propose a research agenda addressing the link between wildfire science and hydrologic ecosystem services. We screened 2935 peer-reviewed articles and selected 82 studies to include based on their relevance according to a systematic, multi-step selection process. Next, we classified the papers into five themes: (a) runoff volume and flow regimes, (b) erosion and sediment transport, (c) water chemistry, (d) hydromorphology, and (e) aquatic food webs. For each study, we documented location, fire regime, watershed characteristics, and ecosystem services. The annual number of published studies on post-fire hydrology in high-latitude forests and, in particular, those addressing hydrologic ecosystem services, has increased steadily in recent years. Descriptions of wildfire characteristics, watershed characteristics, and effects on hydrologic ecosystem services were highly variable across studies, hindering cross-study comparisons. Moreover, there were limited efforts to extend study results to implications for forest or water management decisions regarding ecosystem services from source watersheds. Most studies focused on fire impacts on aquatic habitats and water chemistry while services of direct concern to communities, such as drinking water, were rarely addressed. We contend that study standardization, further use of geospatial technologies, and more studies directly addressing ecosystem services will help mitigate the increasing risks to water resources in northern forests.