Assessment the AquaCrop and DSSAT Models in Simulating Potato Yield and Water Productivity under Different Water Stress Conditions

Simulation of Potato growth helps researchers to plan crop's response to the amount of irrigation water. AquaCrop and DSSAT are the most widely used crop models but have not yet been evaluated under the same conditions for potato simulation. Therefore, in the present study, biennial data collected from a research farm located in Kermanshah, Iran, were used to evaluate these two models under different irrigation water stresses. Irrigation treatments include: 100% supply; T2: 75% supply and T3: 50% water supply in surface drip irrigation. The results showed that both AquaCrop and DSSAT had overestimated errors in simulating potato yield and water productivity (MBE<0). The accuracy of these two models was excellent for simulating yield and water productivity (NRMSE<0.10). In calibration stage, the error of AquaCrop to simulate yield and water productivity of potato water was equal to 1.3 ton.ha-1 and 0.22 kg.m-3, respectively. While the error values for these parameters in the validation stage were equal to 1.5 ton.ha-1 and 0.23 kg.m-3, respectively. DSSAT model error in the calibration stage to simulate water yield and productivity is equal to 2.4 ton.ha-1 and 0.34 kg.m-3, respectively, and in the validation stage is equal to 0.61 ton.ha-1 and 0.11 kg.m-3, respectively. Therefore, both crop models used had acceptable error. The EF values for the yield simulated by AquaCrop and DSSAT models were 0.92 and 0.98, respectively. The EF values for water productivity of these two models were determined to be -2.6 and 0.17, respectively. According to these results, the use of both models for potato simulation is suggested. However, the results obtained for the DSSAT were better accurate and efficient.