Comparative Evaluation of interpolation methods of estimation of soil erodibility factor of Universal Soil Loss Equation (Case study: Yazd- Ardakan plain)

One of the most important parameter of soil erosion is soil erodiblity. Soil erodiblity (K-USLE) plays the main role in mathematical function of erosion models as well as Universal Soil Loss Equation (USLE). In recent paper, the objectives are including determination the spatial pattern of soil erodibility of USLE in part of Yazd Ardakan plain as a typical arid plain of center of Iran and comparison and evaluation the kinds of kriging interpolation procedure consist simple, ordinary and universal. Soil was sampled in 37 points from 0-10 cm soil depth in almost uniform distribution by random classification sampling pattern. Infiltration was determined by double ring method. Soil samples were analyzed in laboratory with standard methods consist grain size distribution and organic matter. Soil erodiblity was calculated by method of Wischmeier and Smith. Normality test was carried out. ArcGIS 9.3 was applied to determine the spatial pattern of K-USLE and the effective parameters using simple, ordinary, universal and Inverse Distance Weighting (IDW) interpolation methods. MAE (Mean Absolute Error), MBE (Mean Bias Error), RMSE (Root Mean Square Error) and RI (Relative Improvement) were calculated for evaluation the interpolation procedures. K-USLE were varied at about 0.0055±0.0072, 0.005±0.0037 and 0.0173±0.0133 ton.hr/MJ.mm in fine, medium and coarse gravely plains, respectively. The result shows Gaussian model was fitted on ordinary kriging as well as spherical was applied for simple and universal kriging. In addition, the result shows the ordinary kiriging have the minimum bias in compare to the else interpolating procedures. Finally, the kriging procedures are more accurate and precision than IDW.