Abstract
Soil water content (SWC) plays a critical role in crop yield, irrigation scheduling, and water resources management. In the Canadian Prairies, the water content in the rootzone replenished by rainfall is rarely sufficient to satisfy crop water requirements. Thus, the need for robust and effective water management. Hydrologic modelling provides the opportunity to understand the underlying processes controlling and affecting soil water movement and distribution. Evapotranspiration (ET) is an important input of hydrologic models; thus, the estimation of ET could have significant consequences on modelling outcome and inference. The FAO Penman-Monteith (PM) is the recommended model for estimating the reference crop evapotranspiration (ETo). However, it is limited by requiring too many weather variables that are not readily available. Simple empirical ETo models have been developed as an alternative. In this study, six ETo models with different inputs were used to simulate soil water dynamics in a rainfed potato farm in Winkler, Manitoba, using the HYDRUS-1D model. The results showed that when used to simulate SWC, all the models followed a similar pattern, although a significant difference was observed at shallow depth (20 cm). Specifically, a significant difference (p < 0.05) was observed between observed and simulated SWC from Hargreaves Samani, Romanenko, Penman, and FAO-PM (missing) models. When used to simulate the crop evapotranspiration (ETc), there was no significant difference (p > 0.05) between observed and simulated ETc from FAO PM, Irmak, and Priestly – Taylor models. Hence, ETo models with fewer data inputs such as Irmak and Priestly – Taylor models can provide accurate and reliable results for water management in southern Manitoba.