Application of Ordinary Kriging in Mapping Soil Organic Carbon in Chad using SoilGrids data

Abstract

Quantifying the pattern and spatial distribution of soil organic carbon (SOC) is fundamental to understanding many ecosystem processes. This study aimed to apply ordinary kriging (OK) to model the spatial distribution of SOC in Chad. Nine hundred ninety-five sampling locations from the region were used to extract Soil organic carbon from three raster layers. Those raster layers represented the SOC of 0-5cm, 5-15cm, and 15-30cm of soil horizon and were downloaded from the Soil Grids website. The mean value of the soil carbon derived from the three horizons was used as 0-30cm horizon data and analyzed using R-4.1.3 version software and ArcGIS 10.5. Different variogram models were firstly examined on the variogram cloud, and the best fit was selected based on RMSE, MSE, and MAE criteria. The results indicated that the Gaussian model best fit for data, with 27.84, -3.35, and 20.95 obtained for RMSE, MAE, and ME. The short-range spatial dependence of SOC was strong, with a nugget close to zero. The spatial dependency of the data was medium with a chunk to sill ratio of 0.36. The southern part of the country has a higher concentration of SOC than the northern part. It can be concluded that the generated map could serve as a proxy for SOC in the region where evidence of spatial structure and quantitative estimates of uncertainty are reported. Therefore, the maps produced can be used for many applications, including soil sampling optimization.

Key Words: Gaussian semivariogram, spatial distribution, Soil organic carbon, spatial autocorrelation