Kriging Prediction and Simulation Model: Analysis of Surface Soil Particle Size Distribution

Abstract

Kriging is a statistical approach that takes into account spatial autocorrelation data. Accordingly, it allows better prediction of soil particle sizes than with simple interpolation methods such as linear and spline interpolation. In this paper, we analyze the soil texture in the Kalikonto Watershed, Batu City, using a Kriging simulation, and 150 points obtained with simultaneous field investigation and digital DEM generation. The Silt variable was used for interpolation to map where soil particles are distributed in space. Simulation results show that the Spherical variogram Kriging model has a strong spatial relationship, reaching significant levels of significance. Thus, its predicted values exhibit little divergence from real-world data quality. The Mean Square Error (MSE) is 0.002084. The predicted distribution of soil particles matches closely with field observations and thus provides a more accurate analysis space for land management. The innovativeness of this paper lies in optimizing a model for the Spherical variogram to act as a predictor and using more forecast points than previously done studies. This approach enables representation of more accurate spatial relations in land management for land use and soil conservancy practices.