Evaluating inverse distance weighting and kriging methods in estimation of some physical and chemical properties of soil in Qazvin Plain

Abstract

Today, the presence of accurate information about variability of soil properties been considered more than ever to apply this information in economic modeling, environmental predictions, accurate farming and natural resources management. The present research was conducted in some lands of Qazvin plain to study variability of some chemical and physical properties of soil by sampling 62 observational points in depth of 20 cm above soil surface. Initial statistical study of data indicated that the studied properties follow normal distribution in the region. Spatial variations of the studied properties showed that spherical model was the best fitted model to semivariogramin other properties than silt percent and bulk densityand total porosity. The highest radius for the studied properties was 21100 m related to bulk density, total porosity and electric conductivity and pH. Spatial dependence class was observed medium to strong in all physcial and chemcial properties. To validate intrapolation methods, three indices of evaluation, R 2 , MBE, MAE which indicate accuracy of each of the intrapolation methods were used and results showed that the studied properties had spatial structure, their impact range had good variability and kriging estimator better can show variability of the studied properties in the region in comparison to IDW method. At the end, considering the best interpolation method, spatial variability map of each of the properties was prepared in ArcGIS software .