Simulating groundwater-surface water interactions in the Canadian Prairies using a coupled land-atmosphere model (ParFlow-CLM)

Wednesday, 16 December 2015
Poster Hall (Moscone South)
Andrew M Ireson and Melkamu Alebachew Ali, University of Saskatchewan, Saskatoon, SK, Canada
The Canadian prairies are cold and dry. Surface depressions are ubiquitous, and contain permanent or ephemeral ponds. The ponds are filled by snowmelt and precipitation on the ponds and lose a significant portion of their water to evaporation, but also, depending on their landscape position, may spill to other ponds or channels, recharge groundwater, or received groundwater discharge. Since precipitation and actual evaporation are closely balanced, the pond water balances are very sensitive to change in climate, and the prairies in general have been subject to damaging floods and droughts, in particular in the last decade or two. A 2.25 km2 field site at St Denis, central Saskatchewan, contains over 100 ponds, some permanent, some ephemeral, some saline, some fresh, some recharging groundwater, some receiving groundwater discharge. The site has been extensively studied for almost 50 years, with about one decade of continuous meteorological data, and three years of detailed pond level, soil moisture and temperature, and groundwater data. The objective of this study was to assess the performance of PARFLOW-CLM (a coupled land-atmosphere model) in simulating the pond-groundwater interactions at this site. Our conceptual model of the site includes soil properties that are progressively weathered with depth, and we implement this in a simplified dual permeability mathematical model of the soil hydraulic properties, whereby storage is dominated by the matrix and flow is dominated by macropores. The model performance was surprisingly good, doing quite a good job of capturing the observed groundwater and pond level dynamics. The soil freezing regime is also captured reasonably well, though the timing and pattern of the zero degree isotherm during soil thaw, which is critically important for runoff generation processes, was not captured as well. The model provides credible insights into the spatial patterns of evapotranspiration, and the seasonal dynamics of subsurface connectivity, as flow directions around the pond change throughout the year.