Hydrological responses to climatic variability in a cold agricultural region

Abstract

Extended severe dry and wet periods are frequently observed in the northern continental climate of the Canadian Prairies. Prairie streamflow is mainly driven by spring snowmelt of the winter snowpack, whilst summer rainfall is an important control on evapotranspiration and thus seasonality affects the hydrological response to drought and wet periods in complex ways. A field-tested physically based model was used to investigate the influences of climatic variability on hydrological processes in this region. The model was set up to resolve agricultural fields and to include key cold regions processes. It was parameterized from local and regional measurements without calibration and run for the South Tobacco Creek basin in southern Manitoba, Canada. The model was tested against snow depth and streamflow observations at multiple scales and performed well enough to explore the impacts of wet and dry periods on hydrological processes governing the basin scale hydrological response. Four hydro-climatic patterns with distinctive climatic seasonality and runoff responses were identified from differing combinations of wet/dry winter and summer seasons. Water balance analyses of these patterns identified substantive multiyear subsurface soil moisture storage depletion during drought (2001)(2002)(2003)(2004)(2005) and recharge during a subsequent wet period (2009)(2010)(2011). The fractional percentage of heavy rainfall days was a useful metric to explain the contrasting runoff volumes between dry and wet summers. Finally, a comparison of modeling approaches highlights the importance of antecedent fall soil moisture, ice lens formation during the snowmelt period, and peak snow water equivalent in simulating snowmelt runoff.

Simulating cold-region hydrology in an intensively drained agricultural watershed in Manitoba, Canada, using the Cold Regions Hydrological Model

“…Simulated yearly cumulative evaporation from open water was compared to gross evaporation values published for Portage La Prairie between 1992 and 2000 (Martin, 2002), which is the closest location with available data corresponding to the time period of model simulations. Since both values were based on evaporation models and not measured, the evaporation comparison was made only to verify that CRHM simulations were in agreement with values expected for the study area.…”

“…This application of the CHRM has a very high proportion of the watershed used as cropland (87 %), extremely level topography, and modest depressional storage, which is unique and contrasts with other agricultural applications of the CRHM in land use proportions and topographic relief (Mahmood et al, 2017;Fang et al, 2010;. The land use split method, used in other modelling efforts in this watershed (Yang et al, 2014) to represent crop rotations in a static fashion in a multi-year model exercise, seemed to work well for the application in this study given the good simulations in wet years.…”

“…Simulations of SWE, evaporation, and VWC were validated by external datasets, indicating that the model state variables were in good agreement with the major hydrological processes in the agricultural landscape. Small yearly differences in evaporation were likely due to differences in input data, location, and method, where AAFC applies Meyer's revised formula (Martin, 2002) for lakes and the CRHM uses the PriestleyTaylor procedure for wetlands and small lakes. Differences in soil moisture values could be due to differences in soil properties used to calculate the water holding capacity of the soil.…”

“…The CRHM platform has been successfully used to simulate hydrological processes in a number of contrasting catchments in Canada (Dornes et al, 2008;Fang and Pomeroy, 2008;Ellis et al, 2010) and globally, in western China (Zhou et al, 2014), Patagonia (Krogh et al, 2015), the German Alps (Weber et al, 2016), and the Spanish Pyrenees (Rasouli et al, 2014), and in the intensively studied South Tobacco Creek watershed that drains from the more steeply sloped Pembina/Manitoba Escarpment feature into the Red River in Manitoba (Mahmood et al, 2017). However, the CRHM has not yet been used to simulate hydrological processes in the intensively managed lowland agricultural tributaries of the Red River such as the La Salle River.…”

Using an inverse modelling approach with equifinality control to investigate the dominant controls on snowmelt nutrient export

“…CRHM is a flexible, modular modelling framework suitable for cold and other climates that provides a range of numerical algorithms that can be critically and sensibly combined for representation of the most relevant hydrological processes of a basin. It has been extensively applied in the Canadian Prairies (Fang & Pomeroy, ; Fang et al, ; Mahmood, Pomeroy, Wheater, & Baulch, ). The reader is referred to Costa et al () for more details about the WINTRA model.…”