Spatial zoning of pan evaporation and affecting variables using geostatisitc methods (Case study: Fars province)

The study of spatio-temporal variations of pan evaporation is quite important in water resource management and evapotranspiration estimation. In current research, these variations and affecting variables have been exmanied using several geostatisitc methods across Fars province, Iran. The skill of these methods were evaluated by statistical measures including R2, MRE, RMSE, NRMSE, GMER, and GSDER. Spatial variability structure of the studied climatic variables followed the Gaussian and spherical models with influence ranges of 9 to 35 km. The ordinary point kriging with spherical semivariogram model (with influence ranges of 8.9 to 35 km, nugget ratios of 0.09 to 36.44% and variability classes of moderate to strong) was the most suitable method for interpolating of pan evaporation (with R2 of 0.74) and precipitation characteristics (with R2 of 0.57 to 0.76) values. The inverse distance weighting with weighting powers of 2 to 5 was the most suitable method for prediction of temperature (with R2 of 0.62 to 0.87) and wind characteristics (with R2 of 0.73). The optimum distace between the rainguages was determined as 35 km. In general, the results showed that the northern and especially the northeastern parts of the province had lower temperature (<15 ͦ C) and pan evaporation (235 mm), and higher wind speed (>15 m s-1) in comparison to southern parts. Therfore, precise water resource management and new cropping pattern in these regions of the province may be recommended.