نوع مقاله : مقاله پژوهشی
چکیده
کلیدواژهها
عنوان مقاله [English]
Drought is the most complicated natural hazards. This phenomenon has different characteristics which are calculated by different indices, where as these indices are depondant on climatic variables and meteorological stations, therefore their ability for recording is not continuse. On this basis , for estimating the amount of meteorological variables in non-sampled locations and preparing continuse maps, spatial interpolation methods are used . In this paper, fristly by using Standard Precipitation Index(SPI ),the drought severity is calculated, and then for assessment of different interpolation methods, different types of spatial interpolation methods including Kriging, Cokriging, Inverse Distance Weighting (IDW), Local Polynomial, Completely Regularzide Spline, Spline with Tension and Thin Plate Spline were used. Also for comparing interpolation methods, the cross-validation technique and error criterions including MSE,MBA , RMSE and %G were used.
The results indicate that Kriging Ordinery method is the best method for interpolation of drought severity and SPI index, because it has the least errors in most of months and in 9 months of the year, G% value is calculated as positive. In fact, this method has the least deviation. The most errors of interpolation methods is relevant to Thin Plate Spline method, indeed each of the three evaluation criteria shows greater value of error than the other methods.
کلیدواژهها [English]
)