Modelling the energy exchange between the snowpack and the atmosphere is critical for many hydrological applications. Of the terms present in the snow energy balance, the turbulent fluxes of sensible and latent heat are the most challenging to estimate, particularly within mountain environments where the hydrological importance is great. Many of the flux estimation techniques, such as the bulk transfer method, are poorly adapted for use in complex terrain. In order to characterize the turbulence and to assess the suitability of flux estimation techniques, eddy covariance flux measurements and supporting meteorological data were collected from two mountain valley forest openings in Kananaskis Country, AB. These sites were generally calm, however wind gusts were frequently observed which markedly affected the turbulence characteristics and increased the rates of momentum and heat transfer. In order to successfully apply the bulk transfer technique at these sites, it was necessary to use environment-specific transfer coefficients to account for the effect of the surrounding complex terrain. These observations were compared with data collected on a treeless alpine ridge near Whitehorse, YT, where it was found that many of the turbulence characteristics were similar to flat sites. However, the boundary layer formed over the alpine ridge was very thin and the site was poorly suited for estimating surface fluxes. The mountain results were further contrasted with data collected over a homogeneous and flat prairie site located near Saskatoon, SK. This site included measurement of all of the snow energy terms, permitting an estimate of the energy balance closure obtainable over snow surfaces. The observed energy balance residual was very large, indicating that the eddy covariance technique was unable to capture all of the turbulent energy. It was concluded that an unmeasured transfer of sensible heat was occurring which was strongly correlated with the long-wave radiation balance. Mechanisms for this relationship were hypothesized. Two snow energy balance models were used to investigate the energy imbalance, where it was observed that the flux terms could be suitably simulated if effective parameters were used to augment the sensible heat transfer rate. The results from this thesis contribute to the understanding of heat transfer processes over snow surfaces during mid-winter conditions and improve the ability to model turbulent heat and mass fluxes from snow surfaces in complex environments.

Landslides are relatively frequent in Ottawa due to the presence of sensitive marine clays (Leda clay or Champlain Sea clay), and the presence of natural or climatic triggers such as rainfall or snowmelt. A geographic information system (GIS) based modeling tool has been developed to assess and predict climate (rainfall and snowmelt)-induced landslides in the sensitive marine clays of the Ottawa region. The Transient Rainfall Infiltration and Grid-based Regional Slope-Stability (TRIGRS) model is used in a GIS framework to investigate the influence of rainfall and snowmelt on shallow landslides through the Ottawa region, with respect to time and location. First, the GIS and TRIGRS models are combined to assess landslide susceptibility with respect to rainfall. The GIS-TRIGRS approach requires topographic, geologic, hydrologic, and geotechnical information of the study area. In addition to this technical information (input data), rainfall intensity data for different durations (5 minutes, and 6, 12, 18, and 24 hours), and historical data of the regional landslides is required. This data is used to verify the locations of predicted landslide-susceptible areas with respect to historical landslide maps in the area. The generated results from the GIS-TRIGRS model were verified by comparing the predicted and historical locations of shallow landslides induced by rainfall throughout the Ottawa region. The comparison results showed a high correlation between the predicted areas of landslides and the previously reported landslides. In addition, the results also indicated that not all previous landslides in Leda clays were triggered by rainfall. The second application of the developed GIS-TRIGRS approach was used to assess and predict snowmelt-induced landslides in areas of sensitive marine clay in the Ottawa region. Similar to the first analysis, the approach requires the following input data: topographic, geologic, hydrologic, geotechnical, snowmelt intensity data for various periods (6–48 hours, 3–15 days, 25 days, and 30 days), This approach also requires data indicating the location of historical landslides in the study area. Using this data, we examine both the timing and location of shallow landslides due to snowmelt in a GIS-based framework. The developed model was validated by comparing the predicted landslide-susceptible areas to historical landslide maps in the study area. A high correlation between predicted and historical landslide location trends was obtained, confirming that the developed GIS-TRIGRS model can predict the snowmelt-induced landslide susceptibility in the sensitive marine clays relatively well. The model results reinforced the conclusion that areas with high slopes and sensitive marine clays were more prone to snowmelt-induced landslides. Finally, in a Geographic Information System (GIS) the landslide occurrence susceptibility in the Ottawa area was modeled. Results of such models are presented as maps showing landslide susceptibility in Champlain Sea clays (Leda clays) in the Ottawa area due to both rainfall and snowmelt. Various input data was collected and entered into a GIS and TRIGRS model. The main categories of such inputs are climate, topography, geology, hydrology, and geotechnical data. The rainfall and snowmelt intensity data was extracted for 24 to 48 hour periods from Environment and Climate Change Canada historical climate records. Thereafter, the factor of safety was calculated in order to determine the stability of slopes across the study area. The model assesses the effects of rainfall and snowmelt on landslide occurrence, and based on the calculated factor of safety at each pixel of the study area, the model calculates the landslide susceptibility. The results presented in this thesis will provide a geotechnical basis for making appropriate engineering decisions during slope management and land use planning in the Ottawa region.

Hydrological models have been developed to estimate snow accumulation, snowmelt and snowmelt runoff on the Canadian Prairies; however, their proper scale of application is unknown in the Prairie environment. The first objective of this thesis is to examine the proper scale for pre-melt snow accumulation as snow water equivalent (SWE) and snowmelt in a Prairie first order basin. Spatially distributed and spatially aggregated approaches were used to calculate SWE and snowmelt at St. Denis National Wildlife Area (SDNWA). Both approaches used models with similar physics, but differed in the model scale at which calculations were carried out. The simulated pre-melt SWE, cumulative seasonal SWE, and daily snowmelt from the two modelling approaches were compared to field observations of pre-melt SWE, cumulative seasonal SWE, and daily snowmelt; comparisons of areal cumulative seasonal SWE, areal snowmelt, snowmelt duration, and snow-covered area were also conducted between two modelling approaches. Results from these comparisons showed that both approaches had reasonable and similar accuracy in estimation of SWE and snowmelt. The spatially aggregated approach was more computationally efficient and was selected as a modelling scale for small-sized prairie basins. <p>Another objective of this thesis is to derive a snow hydrology model for the Canadian Prairies. Physically-based hydrological models were assembled in the Cold Regions Hydrological Model Platform (CRHM) using the aggregated approach. Tests of pre-melt SWE and surface snowmelt runoff were conducted at two basins in Saskatchewan Creighton Tributary of Bad Lake and Wetland 109, St. Denis. Results showed that the snow hydrology model had a reasonable capability to simulate SWE and snowmelt runoff to the stream and wetland. <p>Droughts are natural hazards that develop frequently on the Canadian Prairies. Analyzing the impact of drought on hydrological processes and water supply is another objective of this thesis. Synthetic drought scenarios were proposed for the Creighton Tributary of Bad Lake and the corresponding impacts on the snowmelt runoff-related processes were examined. Results indicated that wind redistribution of snow was very sensitive to drought conditions, sublimation of blowing snow and snow-covered period were sensitive to drought, but winter evaporation and infiltration did not show strong trend. The results also showed that drought conditions had magnified effects on the snowmelt runoff and could cause cessation of streamflow. Also, the impacts of the recent 1999-2005 drought on the snowmelt hydrology were investigated at St. Denis. Results illustrated that three-years (1999-2002) of severe winter drought were followed by a normal year (2002-03) and then a two-year (2003-05) recovery period, and then returning to normal (2005-06). Results showed that both snowfall and rainfall during hydrological winter were consistently low for severe drought and surface snowmelt runoff was very much lower during severe drought, about 45-65 mm less compared to that in the normal periods.

Hydrological models have been developed to estimate snow accumulation, snowmelt and snowmelt runoff on the Canadian Prairies; however, their proper scale of application is unknown in the Prairie environment. The first objective of this thesis is to examine the proper scale for pre-melt snow accumulation as snow water equivalent (SWE) and snowmelt in a Prairie first order basin. Spatially distributed and spatially aggregated approaches were used to calculate SWE and snowmelt at St. Denis National Wildlife Area (SDNWA). Both approaches used models with similar physics, but differed in the model scale at which calculations were carried out. The simulated pre-melt SWE, cumulative seasonal SWE, and daily snowmelt from the two modelling approaches were compared to field observations of pre-melt SWE, cumulative seasonal SWE, and daily snowmelt; comparisons of areal cumulative seasonal SWE, areal snowmelt, snowmelt duration, and snow-covered area were also conducted between two modelling approaches. Results from these comparisons showed that both approaches had reasonable and similar accuracy in estimation of SWE and snowmelt. The spatially aggregated approach was more computationally efficient and was selected as a modelling scale for small-sized prairie basins. <p>Another objective of this thesis is to derive a snow hydrology model for the Canadian Prairies. Physically-based hydrological models were assembled in the Cold Regions Hydrological Model Platform (CRHM) using the aggregated approach. Tests of pre-melt SWE and surface snowmelt runoff were conducted at two basins in Saskatchewan Creighton Tributary of Bad Lake and Wetland 109, St. Denis. Results showed that the snow hydrology model had a reasonable capability to simulate SWE and snowmelt runoff to the stream and wetland. <p>Droughts are natural hazards that develop frequently on the Canadian Prairies. Analyzing the impact of drought on hydrological processes and water supply is another objective of this thesis. Synthetic drought scenarios were proposed for the Creighton Tributary of Bad Lake and the corresponding impacts on the snowmelt runoff-related processes were examined. Results indicated that wind redistribution of snow was very sensitive to drought conditions, sublimation of blowing snow and snow-covered period were sensitive to drought, but winter evaporation and infiltration did not show strong trend. The results also showed that drought conditions had magnified effects on the snowmelt runoff and could cause cessation of streamflow. Also, the impacts of the recent 1999-2005 drought on the snowmelt hydrology were investigated at St. Denis. Results illustrated that three-years (1999-2002) of severe winter drought were followed by a normal year (2002-03) and then a two-year (2003-05) recovery period, and then returning to normal (2005-06). Results showed that both snowfall and rainfall during hydrological winter were consistently low for severe drought and surface snowmelt runoff was very much lower during severe drought, about 45-65 mm less compared to that in the normal periods.

Reliable hydrological model simulations are the result of numerous complex interactions among hydrological inputs, landscape properties, and initial conditions. Determination of the effects of these factors is one of the main challenges in hydrological modelling. This situation becomes even more difficult in cold regions due to the ungauged nature of subarctic and arctic environments.<p> This research work is an attempt to apply a new approach for modelling snowcover ablation and snowmelt runoff in complex subarctic environments with limited data while retaining integrity in the process representations. The modelling strategy is based on the incorporation of both detailed process understanding and inputs along with information gained from observations of basin-wide streamflow phenomenon; essentially a combination of deductive and inductive approaches. The study was conducted in the Wolf Creek Research Basin, Yukon Territory, using three models, a small-scale physically based hydrological model, a land surface scheme, and a land surface hydrological model. The spatial representation was based on previous research studies and observations, and was accomplished by incorporating landscape units, defined according to topography and vegetation, as the spatial model elements.<p> Comparisons between distributed and aggregated modelling approaches showed that simulations incorporating distributed initial snowcover and corrected solar radiation were able to properly simulate snowcover ablation and snowmelt runoff whereas the aggregated modelling approaches were unable to represent the differential snowmelt rates and complex snowmelt runoff dynamics. Similarly, the inclusion of spatially distributed information in a land surface scheme clearly improved simulations of snowcover ablation. Application of the same modelling approach at a larger scale using the same landscape based parameterisation showed satisfactory results in simulating snowcover ablation and snowmelt runoff with minimal calibration. Verification of this approach in an arctic basin illustrated that landscape based parameters are a feasible regionalisation framework for distributed and physically based models. In summary, the proposed modelling philosophy, based on the combination of an inductive and deductive reasoning, is a suitable strategy for reliable predictions of snowcover ablation and snowmelt runoff in cold regions and complex environments.

Saskatchewan producers traditionally overwinter their cattle in pens in the yard. The practice of winter feeding of cattle directly in the field is increasing in popularity leading to concerns about increased nutrients being deposited in soil and potentially lost in runoff water and to ground water. In 2008/2009 an experiment was conducted to observe the effect of in-field winter feeding of cows on the nutrients in spring snowmelt run-off. Approximately 100 cattle were baled grazed on a Russian wild ryegrass pasture at a stocking rate of 2240 cow-days ha-1 for 88 d during the winter at the Western Beef Development Center at Lanigan, SK. The spring 2009 ponded water was sampled from four basins in the control (no cattle were present) and four basins in the winter feeding treatment from the end of March to mid-April. Ground water samples from two piezometers in the control and two in the winter feeding area were gathered from the start of runoff until the middle of summer. Soil samples (0-10 cm) were collected in the fall 2008 before winter feeding and again in the spring 2009 after winter feeding on both the control and treated areas to examine the influence of winter feeding on soil nutrients. Orthophosphate-P and ammonium-N concentrations were elevated to levels up to 19.9 mg PO4-P L-1 and 102.3 mg NH4-N L-1 respectively in run-off from the winter feed treatment basins compared to the controls (2.1 mg PO4-P L-1 and 1.72 mg NH4-N L-1). Nitrate-N concentrations in snowmelt run-off water were similar from the winter-fed areas (0.008 mg NO3-N L-1 to 0.739 mg NO3-N L-1) and the control (0.001 mg NO3-N L-1 to 1.046 mg NO3-N L-1). This is explained by lack of sufficient time and temperature for organic N, urea and ammonium in the urine and fecal matter to convert to nitrate. In the ground water there was a slight increase in nutrient ion concentration in the winter feed basins compared to the control. Soil sampled in the spring from the winter feeding site had higher soluble nitrate, ammonium and phosphorus compared to the control. The soluble and exchangeable forms of phosphorus in the soil were lower compared to the fall soil samples for the control and winter feeding site, possibly due to immobilization by plant and microbial uptake in the spring. Caution should be used when selecting sites for in-field winter feeding system so the run-off water does not reach sensitive water bodies.

The presence of nutrients in snowmelt runoff from agricultural watersheds has been reported by previous studies. However, no study has answered the most important question what areas of the watershed contribute nutrients to snowmelt runoff? or addressed the factors that control snowmelt runoff water quality. This study was designed to (1) find the areas that contribute nutrient to snowmelt runoff (termed as critical source areas, CSA), and (2) understand the source and transport factors that control the snowmelt runoff water quality in the Canadian prairies. The findings of this study will provide vital information to understand snowmelt runoff water quality and for sustainable management of soil nutrients and snowmelt runoff water quality in the Canadian prairies. <p>Source and transport factors and snowmelt runoff water quality were studied for two years on shoulder, backslope and footslope landform segments. The distribution of fall soil nutrients in the top 5 cm soil layer (available soil P [ASP], nitrate [NO3-] and ammonium [NH4+]), snow depth, snow water equivalent (SWE), snowmelt runoff and snowmelt runoff water quality (total P [TP], total dissolved P [TDP], NO3-N and sediment) were studied using closed and open plots placed on each landform segment. The influence of source and transport factors was evaluated in relation to snowmelt runoff water quality. <p>The ASP had a distribution pattern of backslope < shoulder < footslope in 2003 before manure application (bma) and shoulder = backslope = footslope in 2004. The NO3- distributed as shoulder = backslope = footslope in 2003 (bma) and shoulder < backslope < footslope in 2004. However, NH4+ had a stable distribution of shoulder = backslope < footslope in 2003 bma and in 2004. The pre-melt SWE increased in the down slope direction having the lowest in the shoulder and backslope and the highest in the footslope in 2005. The average daily snowmelt runoff from 1 m2 plots did not vary between the shoulder and the backslope. Infiltration was dominant in 2004 while runoff was dominant in 2005. Of the three landform segments, the shoulder was the greatest contributor of runoff to the depression. The backslope contributed the least. <p>Hog manure injection did not seem to influence snowmelt runoff water quality. Most nutrients and sediments were from the land surface. Analysis revealed that fall soil nutrient concentrations were not a dominant factor controlling the nutrients in the snowmelt runoff in this site. However, snowmelt runoff volume controlled snowmelt runoff water quality. Snowmelt runoff water quality did not vary between the landform segments. However, as a result of the dominance of shoulder in this landscape, the total transport of nutrients and sediment was the highest from shoulder. Where landform characteristics are similar to the study watershed, it may be argued that all landform segments are CSA. Runoff volume is the most influential factor in determining the importance of CSA and controlling the snowmelt runoff water quality.

The presence of nutrients in snowmelt runoff from agricultural watersheds has been reported by previous studies. However, no study has answered the most important question what areas of the watershed contribute nutrients to snowmelt runoff? or addressed the factors that control snowmelt runoff water quality. This study was designed to (1) find the areas that contribute nutrient to snowmelt runoff (termed as critical source areas, CSA), and (2) understand the source and transport factors that control the snowmelt runoff water quality in the Canadian prairies. The findings of this study will provide vital information to understand snowmelt runoff water quality and for sustainable management of soil nutrients and snowmelt runoff water quality in the Canadian prairies. <p>Source and transport factors and snowmelt runoff water quality were studied for two years on shoulder, backslope and footslope landform segments. The distribution of fall soil nutrients in the top 5 cm soil layer (available soil P [ASP], nitrate [NO3-] and ammonium [NH4+]), snow depth, snow water equivalent (SWE), snowmelt runoff and snowmelt runoff water quality (total P [TP], total dissolved P [TDP], NO3-N and sediment) were studied using closed and open plots placed on each landform segment. The influence of source and transport factors was evaluated in relation to snowmelt runoff water quality. <p>The ASP had a distribution pattern of backslope < shoulder < footslope in 2003 before manure application (bma) and shoulder = backslope = footslope in 2004. The NO3- distributed as shoulder = backslope = footslope in 2003 (bma) and shoulder < backslope < footslope in 2004. However, NH4+ had a stable distribution of shoulder = backslope < footslope in 2003 bma and in 2004. The pre-melt SWE increased in the down slope direction having the lowest in the shoulder and backslope and the highest in the footslope in 2005. The average daily snowmelt runoff from 1 m2 plots did not vary between the shoulder and the backslope. Infiltration was dominant in 2004 while runoff was dominant in 2005. Of the three landform segments, the shoulder was the greatest contributor of runoff to the depression. The backslope contributed the least. <p>Hog manure injection did not seem to influence snowmelt runoff water quality. Most nutrients and sediments were from the land surface. Analysis revealed that fall soil nutrient concentrations were not a dominant factor controlling the nutrients in the snowmelt runoff in this site. However, snowmelt runoff volume controlled snowmelt runoff water quality. Snowmelt runoff water quality did not vary between the landform segments. However, as a result of the dominance of shoulder in this landscape, the total transport of nutrients and sediment was the highest from shoulder. Where landform characteristics are similar to the study watershed, it may be argued that all landform segments are CSA. Runoff volume is the most influential factor in determining the importance of CSA and controlling the snowmelt runoff water quality.

Reliable hydrological model simulations are the result of numerous complex interactions among hydrological inputs, landscape properties, and initial conditions. Determination of the effects of these factors is one of the main challenges in hydrological modelling. This situation becomes even more difficult in cold regions due to the ungauged nature of subarctic and arctic environments.<p> This research work is an attempt to apply a new approach for modelling snowcover ablation and snowmelt runoff in complex subarctic environments with limited data while retaining integrity in the process representations. The modelling strategy is based on the incorporation of both detailed process understanding and inputs along with information gained from observations of basin-wide streamflow phenomenon; essentially a combination of deductive and inductive approaches. The study was conducted in the Wolf Creek Research Basin, Yukon Territory, using three models, a small-scale physically based hydrological model, a land surface scheme, and a land surface hydrological model. The spatial representation was based on previous research studies and observations, and was accomplished by incorporating landscape units, defined according to topography and vegetation, as the spatial model elements.<p> Comparisons between distributed and aggregated modelling approaches showed that simulations incorporating distributed initial snowcover and corrected solar radiation were able to properly simulate snowcover ablation and snowmelt runoff whereas the aggregated modelling approaches were unable to represent the differential snowmelt rates and complex snowmelt runoff dynamics. Similarly, the inclusion of spatially distributed information in a land surface scheme clearly improved simulations of snowcover ablation. Application of the same modelling approach at a larger scale using the same landscape based parameterisation showed satisfactory results in simulating snowcover ablation and snowmelt runoff with minimal calibration. Verification of this approach in an arctic basin illustrated that landscape based parameters are a feasible regionalisation framework for distributed and physically based models. In summary, the proposed modelling philosophy, based on the combination of an inductive and deductive reasoning, is a suitable strategy for reliable predictions of snowcover ablation and snowmelt runoff in cold regions and complex environments.

Saskatchewan producers traditionally overwinter their cattle in pens in the yard. The practice of winter feeding of cattle directly in the field is increasing in popularity leading to concerns about increased nutrients being deposited in soil and potentially lost in runoff water and to ground water. In 2008/2009 an experiment was conducted to observe the effect of in-field winter feeding of cows on the nutrients in spring snowmelt run-off. Approximately 100 cattle were baled grazed on a Russian wild ryegrass pasture at a stocking rate of 2240 cow-days ha-1 for 88 d during the winter at the Western Beef Development Center at Lanigan, SK. The spring 2009 ponded water was sampled from four basins in the control (no cattle were present) and four basins in the winter feeding treatment from the end of March to mid-April. Ground water samples from two piezometers in the control and two in the winter feeding area were gathered from the start of runoff until the middle of summer. Soil samples (0-10 cm) were collected in the fall 2008 before winter feeding and again in the spring 2009 after winter feeding on both the control and treated areas to examine the influence of winter feeding on soil nutrients. Orthophosphate-P and ammonium-N concentrations were elevated to levels up to 19.9 mg PO4-P L-1 and 102.3 mg NH4-N L-1 respectively in run-off from the winter feed treatment basins compared to the controls (2.1 mg PO4-P L-1 and 1.72 mg NH4-N L-1). Nitrate-N concentrations in snowmelt run-off water were similar from the winter-fed areas (0.008 mg NO3-N L-1 to 0.739 mg NO3-N L-1) and the control (0.001 mg NO3-N L-1 to 1.046 mg NO3-N L-1). This is explained by lack of sufficient time and temperature for organic N, urea and ammonium in the urine and fecal matter to convert to nitrate. In the ground water there was a slight increase in nutrient ion concentration in the winter feed basins compared to the control. Soil sampled in the spring from the winter feeding site had higher soluble nitrate, ammonium and phosphorus compared to the control. The soluble and exchangeable forms of phosphorus in the soil were lower compared to the fall soil samples for the control and winter feeding site, possibly due to immobilization by plant and microbial uptake in the spring. Caution should be used when selecting sites for in-field winter feeding system so the run-off water does not reach sensitive water bodies.