Assessment of SIMETAW model in simulation of climate parameters and investigation of climate change effect on evapotranspiration in four different climates

According to the water crisis and the need for careful planning, prediction of potential evapotranspiration can be helpful to plan water resources and adopt appropriate management methods. Furthermore, knowing the exact amount of evapotranspiration is essential for estimation of water use and irrigation system design. Different models have been provided by researchers for estimation of evapotranspiration, which vary considerably from each other in terms of the number of meteorological parameters they take into account. But due to the lack of the reliable long-term data in some stations, use of model with little data is needed and will be necessary to help planners. In this study the Simulation of Evapotranspiration of Applied Water model (SIMETAW) was evaluated for generation of daily weather data from the long-term monthly values of four synoptic stations (Rasht: Very wet, Shahrekord: Semi-arid, Ahvaz: Arid and Sanandaj: Moderate), also estimation of potential evapotranspiration in the base period (1961-2000) and future period (2040–2080). In the first stage, SIMETAW model was run using long-term average monthly weather data in the base period and daily weather output data were obtained for 41 years. Then by using this data, long-term daily values were calculated and compared with observed data. To compare the results, Coefficient of Determination (R2), Root Mean Square Error (RMSE) and Index of Agreement (d) was used.
The results showed that the model simulations in all stations had acceptable accuracy and the highest accuracy of model in simulation of maximum temperature (R2=0.9954) and precipitation (R2=0.3716) related to Mediterranean climate, minimum temperature (R2=0.9947) and dew point temperature (R2=0.9942) related to very wet climate, wind speed (R2=0.8094) related to arid climate and solar radiation (R2=0.9902) related to semi-arid climate. The accuracy of the model for simulation of rainfall in four stations was low (R2=0.2457, R2=0.2435, R2=0.3553 and R2=0.3716 for arid, very wet, semi-arid and Mediterranean climate, respectively). The highest accuracy of SIMETAW model in simulation of evapotranspiration was belong to mediterranean climate (R2=0.9936), semi-arid climate (R2=0.9935), arid climate (R2=0.9903) and very wet climate (R2=0.9846), respectively. In the second stage, meteorological parameters were simulated in future period (2040-2080) under two emission scenarios (A2 and B2) by HadCM3 (Hadley Centre Coupled Model, version 3) outputs and using downscaling model of SDSM) Statistical Downscaling Model(. To verify the accuracy of SDSM model in downscaling outputs of HadCM3, long-term average of daily data in the base period 2001-1961 were compared with simulated data using the SDSM in the future period. The results showed the high accuracy in simulation of all parameters. Comparing data from downscaling in the future and observed data in the base period, showed that in four stations, in total, maximum temperature, minimum temperature, dew point temperature and wind speed will increase in the future period compared to the base period while rainfall will decrease. For example, in Ahvaz station, the average increase in the future maximum temperature, minimum temperature, wind speed and dew point temperature will be equal to 3.4, 8.0, 5.0 and 18.9 percent under the A2 scenario and 2.1, 5.7, 5.0 and 14.7 percent under the B2 scenario, respectively. Also, the average decrease in the future rainfall will be equal to 17.9 percent under the A2 scenario and 20.5 percent under the B2 scenario, respectively.
In the third stage, from SDSM model outputs in the future period, long-term monthly values were prepared and used as SIMETAW model input. Then from this data, long-term monthly values were produced and then compared with SIMETAW model outputs in the base period. It was found that in four stations, in total, the potential evapotranspiration will be increased under the A2 and B2 compared to the base period. The average increase in the future potential evapotranspiration will be equal to 10.4, 9.5, 10.7 and 4.3 percent under the A2 scenario and 10.3, 8.8, 9.4 and 2.9 percent under the B2 scenario, respectively, in Mediterranean, dry, very wet and semi-arid climate. Given the high correlation between simulated and recorded values, it can be concluded that SIMETAW model can be applied effectively and efficiently with high accuracy for simulation of weather data, potential evapotranspiration and also filling gaps in four stations. Also, this model can be used as an appropriate tool for irrigation engineers and decision makers to predict irrigation needs, using monthly records when daily data is not available.