Prediction of rainfed corn evapotranspiration and soil moisture using the STICS crop model in eastern Canada

Abstract

The accurate simulation of crop evapotranspiration (ET) and volumetric soil moisture (SM) is key for agro-environmental simulation of rainfed crops. We compared two ET calculation methods available in the STICS crop model, i.e., the crop coefficient (FAO-56, CC) and the resistance (Shuttleworth and Wallace, SW) methods, over six contrasting rainfed grain corn growing seasons. The ET predictions using the SW method were in good agreement with eddy covariance ET measurements and overall ET predictions were slightly higher than observations by 3% while the CC method overestimated ET by 18%. STICS better predicted daily ET for wet years than for dry years. Ten-day corn ET showed that the SW method performed better when transpiration was dominant (nRMSE of 21.3%) than otherwise (nRMSE of 47.2%). The choice of ET calculation methods had little impact on SM predictions, which are computed in STICS using a tipping bucket approach. SM predictions over the entire soil profile were good for both wet (nRMSE of 15.1%) and dry (nRMSE of 18.2%) years. Larger discrepancies between measured and predicted SM were observed in the surface layer (0–10 cm). A new method was proposed to improve crop model simulations of the soil water transport in the shrinkage cracks during dry years, named Improving Shrinkage Cracks Impact (ISCI). The ISCI method enhanced the performance of SM predictions in the surface soil layer, for which nRMSE decreased to 29.7% for 2002, 18.1% for 2010 and 10.4% for 2018 in comparison with SM results without the use of ISCI method. The steps for implementing ISCI method in crop models using the tipping bucket approach for computing soil moisture are presented. • Two evapotranspiration calculation methods using STICS crop model were compared. • The resistance method predicted corn evapotranspiration better than the crop coefficient one. • Surface soil moisture predicted by STICS departed from observations during dry growing season. • We proposed a new method to account for the impact of shrinkage cracks on surface soil moisture.