Assessing SPI-3 months spatial variation using Kriging, Case study, Fars province

Due to the lack of facilities and commitment in climate stations establishment in large scale interpolation has become an important technique in spatial mapping studies. In this study, the Kriging interpolation technique as a prevalent technique has been explained and 17 Kriging models were evaluated, in order to select the most suitable one, using a case study. Because of increasing trend of drought in Iran and management necessity of water resources, meteorological drought of Fars province has been investigated in this study. For this purpose the calculated SPI-3 months of the Fars province synoptic stations were calculated. SPI-3 months is widely applied in agricultural drought studies. The empirical semi-variogram of the SPI-3 of the stations was calculated and graphed, fitting 17 Kriging models to select the best one using the cross-validation technique. Results showed that the Linear, Power, Square root, Logarithmic and Gaussian models were the best interpolation techniques for SPI-3 spatial mapping. Geostatistic methods could be used to determine the adequacy of the rainfall gauge stations, especially in the western mountainous region of the province. A complete-length of data could modify the results. The Gaussian model as one of the best-fitted models defines good spatial continuous of precipitation in the region. The semivariograms were fitted to data with the hypothesis of no precipitation trend. Considering the linear and non-linear trends of data, and analyzing error maps may result in better models.