Applicability of GIS-based spatial interpolation and simulation for estimating the soil organic carbon storage in karst regions

The applicability of the ordinary kriging method for estimating the soil organic carbon (SOC) stored in karst regions was investigated. A total of 23,536 soil samples were analysed from 2755 soil profiles collected using a grid-based sampling method in a typical small karst basin of western Guizhou in Southwest China. Corrections for the gravel content and rock exposure rate were applied to the GIS-based spatial interpolation and simulation and were compared with the same approach with the addition of soil profiles. With the addition of the soil profile data, the SOC stored in the karst catchment was accurately calculated as follows: 1.48 × 108 kg at a depth of 10 cm, 2.65 × 108 kg at 20 cm, 3.43 × 108 kg at 30 cm, and 5.39 × 108 kg at 100 cm. With the interpolation that was corrected for the rock exposure rate and soil depth, the resulting carbon storage estimation was 1.14–1.19 times higher than the most accurate estimate (that with the soil profiles), with an error rate of 114%–119%. Since the conventional geostatistical method failed to accurately fit the data, including the spatial distribution, micro-geomorphic features, rock exposure rate, and depth of the soil patches in the highly sloped exposed bedrock, must be used to correct the estimation of the SOC storage and organic carbon density in karst areas.