A Comparative Study of Interpolation Methods for Mapping Soil Properties: A Case Study of Eastern Part of Madhya Pradesh, India

The selection of best interpolation technique to assess soil physico-chemical properties of any unsampled location is very important. In this study, the main objective was to compare and analyze the inverse distance weighting (IDW), ordinary kriging (OK), and spline to determine the suitable interpolation technique for mapping soil properties. Relationships between the statistical properties of the data were analyzed using soil test of pH, electric conductivity, organic carbon, nitrogen, phosphorus, potassium, sulfur, and zinc, from 2150 different locations (0–15 cm). For inverse distance weighting and spline, we used exponent value of 1, 2, and 3, and these values were decided from the RMSE value. The ranges of 5–30 closest neighboring points were selected for interpolation methods. Out of total 2150 samples, 10% points were preserved for validation using root mean square error (RMSE) test. Overall, all of the methods gave similar RMSE values. On this study, ordinary kriging (exponential) performed well for pH, K, and S, whereas IDW best for OC, N, and P while spline for EC and Zn. In all uses of IDW, the power of one was the best choice; it is possibly due to the low skewness in OC, N, and P. In all circumstances, a significance of three was established to be the best control for splines. Lognormal kriging gave better result where coefficient of skewness larger than one. We conclude that many parameters can be better identified from the RMSE value obtained from validation.