Application of acceptance probability approach for determination of optimal rain gauge network density (Case study: South Khorasan province)

For accurate estimation of rainfall, as a key element in agricultural and water balance studies, an optimum density of raingauges is required. Although many approaches based on geostatistic are developed to optimize raingauges network, but majority of them suffer from drawbacks. This study aimed to assess a newly developed method in geostatistic based on acceptance probability, for designing the raingauge network with least error in South Khorasan province. The linear moment method was used for testing the homogeneity of the study stations. Then, by choosing a suitable semi-variogram, the acceptance probability in the region was calculated. Based on the spatial pattern of annual rainfall, the acceptance probability was worked out for various parts of the province and the acceptance accuracy (AP) values were analyzed at different levels of probability. The results showed that 20 stations of existing network had no significant effect on estimating the rainfall and it can be recommended to shift their location in order to obtain an optimal network. Also, similar to the existing network of 63 stations, the remaining 43 stations could cover 36% of the province at the probability acceptance level of 80%. Besides, the results indicated that by adding 27 rain gauges to the locations specified in the optimal density, the performance of the optimized network will be approximately doubled comparing to previously existing one, which means 65% coverage of province.