Evaluation of Different Interpolation Methods in Spatial Estimation of Soil Properties in Sistan Plain

Soil is a heterogeneous environment and its properties changes with space and time. Therefore, the investigation of spatial variation of soil properties is essential. This study aimed at investigating the spatial variation of soil parameters such as bulk density, salinity, lime percentage, and soil pH through application of different interpolation methods and choosing the most appropriate method for estimating the above parameters all over the plain. In order to apply geostatistical methods such as Kriging, co-Kriging and inverse distance weighting to determine the spatial variability of soil parameters, 301 soil samples were collected from the plain systematically and regularly, with the distance of 1500m between each two samples and then, soil samples were transferred to the laboratory. After that, the above-mentioned characteristics were measured. Using ArcGIS and GS softwares the variograms of these parameters were determined and the zoning maps for each parameter were drawn. Then, the data analysis was done in SPSS software. The criteria for evaluation of different geostatistical methods in this study was the mean and root mean square error by using cross validation technique. The results showed that squared inverse distance weighting method for bulk density and co-Kriging method for salinity, lime and soil pH had the lowest root mean square error and were the most accurate methods. Zoning maps showed a high percentage of clay in the soils of the northern part of the plain, which affect these soils salinity. In addition, the soil pH and amount of lime in Sistan’s soils were high so this region had calcareous and alkaline soils. bulk density has sharp, sudden changes in the region.