Simulation of irrigation in southern Ukraine incorporating soil moisture state in evapotranspiration assessments

Abstract

The paper studies the accuracy of modeling moisture transport under the conditions of sprinkler irrigation using evapotranspiration assessment methods that take into account the soil moisture conditions. Appropriate modifications of the Penman-Monteith and the Priestley-Taylor models are considered. Moisture transport modeling is performed using the Richards equation in its integer- and fractional-order forms. Parameters identification is performed by the particle swarm optimization algorithm based on the readings of suction pressure sensors. Results for the two periods of 11 and 50 days demonstrate the possibility of up to ~20% increase in the simulation accuracy by using a modified Priestley-Taylor model when the maintained range of moisture content in the root layer is 70%-100% of field capacity. When irrigation maintained the range of 80%-100% of field capacity, moisture content consideration within evapotranspiration assessment models did not enhance simulation accuracy. This confirms the independence of evapotranspiration from soil moisture content at its levels above 80% of field capacity as in this case actual evapotranspiration reaches a level close to the potential one. Scenario modeling of the entire growing season with the subsequent estimation of crop (maize) yield showed that irrigation regimes generated using evapotranspiration models, which take into account soil moisture data, potentially provide higher yields at lower water supply.