Comparison of geostatistic methods for rain gauge network optimization of Khuzestan province

Design of the rain gauge network to increase accuracy and cost decrease in precipitation prediction value needs to find optimum location of the rain gauge station. Most precipitation prediction errors is due to location of the rain gage station that optimum design of the rain gauge stations can reduce precipitation prediction error. In this study used geostatistic methods named Kriging and cokriging to optimization of the Khuzestan rain gauge station network. 49 rain gauges used for 6 daily rainfall. After determination of the best model of semivariograms and the best geostatistic method, in the first step, the rain gauge which precipitation prediction in them has appropriate accuracy determined with the help of other rain gauge network. Result showed that kriging had more accuracy and most daily rainfalls was fitted with gaussian model. In the next step, using assessment of network prediction error value which remain from first step, the points that needs new rain gage determined. Finally7 unnecessary rain gage stations determined and 5 new rain gages added to the network.