Use of Gridded Weather Datasets in Simulation of Wheat Yield and Water Requirement (Case Study: Iran’s Qazvin Plain)

Use of Gridded Weather Datasets in Simulation of Wheat Yield and Water Requirement (Case Study: Iran’s Qazvin Plain)

Temperature and rainfall affect the quantity and quality of agricultural products. Therefore, it is important to estimate its spatial-temporal changes. In many region of the country, due to the low density of meteorological stations or the small statistical period of new stations, limited time and space information is available. Therefore, this study aims to use the data of CRU, AgMERRA, AgCFSR and GPCC gridded weather datasets in estimation of yield and water requirement of wheat and compare them with the estimated values ​​with the information of Qazvin Synoptic Station. For this purpose, monthly weather time series of Qazvin synoptic station were extracted from 1980 to 2010 along with the data from the selected gridded datasets extracted from the closest grid cell to the synoptic station (K1), the average of four closest grid cells to the synoptic station (K4), and the average of eight closest grid cells to the synoptic station (K8). The quality of the gridded datasets was assessed with four statistical indices (R2, RMSE, NRMSE, ME) in a direct and indirect way (the latter using the outputs of the AquaCrop model). In estimating wheat water requirement, GPCC database with four points (K4) and one point (K1) showed the best performance. Wheat yield simulated with AgMERRA data with one (K1) and four (K4) closest grid cells had the highest correlation with the simulated values with data from the synoptic station. Results showed that all selected gridded datasets can be used to simulate grain yield with satisfactory performance, but only data from GPCC-CUR dataset would result in reliable estimation of wheat water requirement.