Variability in evaporation across the Canadian Prairie region during drought and non-drought periods

- Hydrologic simulations were run at 15 point locations across the Canadian Prairies.
- Evaporation distributions were examined during drought/non-drought periods.
- Statistical distributions differed widely for sub-humid and semi-arid locations.
- Large variability ensued under drought but was dampened under increased wetting.
- Information provided may be useful for validating land surface parameterizations.

SummaryKnowledge of changes in spatial and temporal distributions of actual evaporation would be useful for land surface parameterizations in the Prairie region of Canada. Yet challenges persist for examining the variability of evaporation from land surfaces and vegetation over such a large region. This is due in part to the existence of numerous methods of varying complexity for obtaining estimates of evaporation and a general lack of sufficient measurements to drive detailed models. Integrated approaches may be applied for distributing evaporation over vast regions using energy and mass balance methods that integrate remote sensing imagery and surface reference data. Whilst informative, previous studies have not considered the variability of actual evaporation under drought and above normal moisture conditions. Continuous physically-based simulations were conducted for a 46 year period using the Cold Regions Hydrological Model (CRHM) platform. The Penman-Monteith model was applied in this platform to calculate estimates of actual evaporation at point locations which had sufficient hourly measurements. Variations in the statistical properties and mapped distributions derived from point-scale modelling via CRHM were instructional for understanding how evaporation varied spatially and temporally for a baseline normal period (1971-2000) and the years 1999-2005 which included both drought and above normal moisture conditions. The modelling approach was applied successfully for examining the historical variability of evaporation and can be applied to constrain land surface parameterization schemes; validate more empirical predictive model outputs; inform operational agrometeorological and hydrological applications in the Canadian Prairies.