Comparison of Spatial Interpolation Methods to Map Soil Chemical Properties: A Case Study in Chrey Bak Catchment, Cambodia
    1. ITC

Received: January 22,2024 / Revised: Accepted: January 22,2024 / Published: June 01,2020

Download PDF
Browse Figures
×

 Assessing the characteristic of soil properties is important for various purposes, especially for agriculture. The spatial distribution mapping for soil properties in Chrey Bak Catchment can be generated by geospatial analyst (Ordinary Kriging) methods with different semivariogram models: Gaussian model, Exponential model, Pentaspherical model and Hole effect model. The objectives of this study are to: (i) compare and evaluate the most accurate interpolation method for soil properties, and (ii) generate the spatial distribution mapping for soil properties such as pH, Electrical Conductivity (EC), Total Nutrient (TN), potassium (K), Total Phosphorus (TP), Soil Organic Matter (SOM), Cation Exchange Capacity (CEC). Four indicator methods, including (Residual Sum of Squares (RSS), Root Mean Squared Error (RMSE), Mean deviation or Mean error (ME) and Average Kriging Standard Error (AKSE)) were used to evaluate the most accurate interpolation method. The log transformation chosen to estimate the direction or distribution of all parameters with the directional experimental semivariograms is 55o for all soil parameters. The nugget/sill ratio, RSS, and RMSE of Hole Effect model show the best statistical result than other models of OK interpolation method. And then, the soil properties (pH, EC, TN, K, TP, SOM, CEC) distribution mapping was plotted. These maps revealed the understanding of the soil quality in Chrey Bak Catchment.