Solar radiation (SRAD) plays a pivotal role for crop growth and development, and is, therefore, an indispensable input for crop simulation models. Lack of weather data, particularly SRAD, is a problematic issue in many agricultural zones. To handle this dearth, there have been recent advances in the improvement and application of gridded weather data, which provide realistic alternatives to observed weather data. The main objective of the present research was to identify the most accurate gridded SRAD data (GRD) to be used in crop simulation modeling. For this purpose, six GRD products, comprising AgCFSR, AgERA5, AgMERRA, GLDAS, PGF and NASA POWER, were evaluated for irrigated and rainfed wheat simulation in six sites across Iran using the NWheat model. This is one of the newest models that have recently been added to Decision Support System for Agrotechnology Transfer (DSSAT) application. Some of the GRDs were evaluated for the first time in the current study. For a comprehensive assessment, six different rainfed and irrigated wheat cultivars were simulated for the entire range of GRDs. The simulations of grain yield and leaf area were compared with simulations using observed SRAD. The findings showed that the quality of all GRD products was appropriate for crop simulation studies; however, AgCFSR, AgERA5, and AgMERRA, with normalized root mean square error (NRMSE) ≤ 3.8% and Nash-Sutcliffe Efficiency (NSE) ≥ 0.96, had the highest accuracy for grain yield simulation, and PGF with NRMSE of 5.2% and NSE of 0.92 had the least quality. The major disadvantage of AgCFSR and AgMERRA is their limited temporal period until 2010, so that they cannot be used for the studies after 2010. Thus, all the assessed GRD, especially AgERA5 can be recommended for applying in crop simulation models in regions with unavailable SRAD data.

Keywords