Simulation of actual evapotranspiration from agricultural landscapes in the Canadian Prairies

Abstract

Study region This study was carried out in southwestern Manitoba, in the prairie region of Canada. Study focus Mathematical models are routinely used to estimate evapotranspiration (ET) when measurements are lacking. This study was conducted to select the most relevant models for estimating ET in the Canadian Prairies. Eight reference ET models (i.e., Penman-Monteith, Priestley-Taylor, Makkink, Turc, Maulé et al., Blaney-Criddle, Hargreaves-Samani, and Hamon models) were evaluated. This study also assessed the applicability and transferability of the growing degree day (GDD)-based crop coefficients for estimating crop ET in the Canadian Prairies. New hydrological insights The equation developed by Maulé et al. (2006) was found to be the best reference ET alternative to the Penman-Monteith equation with a mean relative error of 11%. However, when models were validated against measured crop ET, the simpler radiation-based Turc and Makkink models were found to be the most useful models with daily mean relative errors ranging from 16% to 49%, outperforming the widely accepted Penman-Monteith model. Discrepancies in the GDD-based crop coefficients were found to also contribute to errors; however, results show the potential transferability of GDD-based coefficients across different locations and climatic conditions.