Comparing Applicability of Some Interpolation Methods for Estimating Soil PH and EC, A Case Study of Doroudzan of Fars Province

Evaluation and selection of most appropriate interpolation method to estimate soil characteristics, play an important role in water strategies and agriculture management. Also due to changes in soil properties in space and time, analysis of the spatial variations of soil properties is important to show the specific ecosystem. The aim of this study, determination of the accuracy of the five interpolation method Inverse Distance Weighting, Global polynomial interpolation, Local polynomial interpolation, Radial Basis Functions and kriging with determine the EC and PH parameters. For this purpose, we attempted to dig to 48 soil profiles in 15 districts of the Doroudzan, Fars Province. The samples of EC and PH were collected from the depths of 0 to 30 cm on the crossing lines. For increasing accuracy of this research, all powers and models of interpolation methods were compared. Semivariograms analysis in GS software indicated that PH and EC were best fitted to Guassian and spherical models. The best interpolation method for mapping the spatial distribution of the parameters were chosen based on values obtained from the error evaluation criteria include the RMSE, MSE, MAE, R2 and MBE. The results showed that LPI in power 2 is the best method to calculate PH and IDW in power 1 has the next place. Also Kriging in Guassian model and IDW in power 1 have the highest accuracy for EC parameter in the study area. Moreover, results showed that the accuracy of RBF and GPI method were less than other methods.