spatial interpolation

Drought risk management requires drought monitoring which is usually carried out by applying different drought indices which are, effectively, continuous functions of precipitation and other meteorological data. However, these indices are estimated at specific locations, and the spatial distribution of drought must be calculated in the form of maps. Geostatistical and deterministic techniques make it possible to interpolate spatially-referenced data. These methods are able to estimate values for arbitrary locations in regions of interest. The current study applied five spatial interpolation methods (inverse distance weighted, global polynomial interpolation, local polynomial interpolation, radial basic function, and kriging [with 4 sub-types]) to extract maps of SPI at 60 rain-gauge stations in the Salt Lake Basin of Iran. Based on the root mean square error, mean absolute error, and mean bias error values of estimations made using sampled data from 1969 to 2009, RBF and kriging techniques were the best and most suitable methods for the spatial analysis of SPI in the study area. The current study applied five spatial interpolation methods (inverse distance weighted, global polynomial interpolation, local polynomial interpolation, radial basic function, and kriging [with 4 sub-types]) to extract maps of SPI at 60 rain-gauge stations in the Salt Lake Basin of Iran. Based on the root mean square error, mean absolute error, and mean bias error values of estimations made using sampled data from 1969 to 2009, RBF and kriging techniques were the best and most suitable methods for the spatial analysis of SPI in the study area.