Modeling crop water productivity using a coupled SWAT-MODSIM model

Abstract

This study examines the water productivity of irrigated wheat and maize yields in Karkheh River Basin (KRB) in the semi-arid region of Iran using a coupled modeling approach consisting of the hydrological model (SWAT) and the river basin water allocation model (MODSIM). Dynamic irrigation requirements instead of constant time series of demand were considered. As the cereal production of KRB plays a major role in supplying the food market of Iran, it is necessary to understand the crop yield-water relations for irrigated wheat and maize in the lower part of KRB (LKRB) where most of the irrigated agricultural plains are located. Irrigated wheat and maize yields (Y) and consumptive water use (AET) were modeled with uncertainty analysis at a subbasin level for 1990-2010. Simulated Y and AET were used to calculate crop water productivity (CWP). The coupled SWAT-MODSIM approach improved the accuracy of SWAT outputs by considering the water allocation derived from MODSIM. The results indicated that the highest CWP across this region was 1.31 kg·m-3 and 1.13 kg·m-3 for wheat and maize, respectively; and the lowest was less than 0.62 kg·m-3 and 0.58 kg·m-3. A close linear relationship was found for CWP and yield. The results showed a continuing increase for AET over the years while CWP peaks and then declines. This is evidence of the existence of a plateau in CWP as AET continues to increase and evidence of the fact that higher AET does not necessarily result in a higher yield.