Using a Multifunctional Approach for Cartographic Modeling of Organic Carbon Content in Natural and Arable Soils of the Central Caucasus

Based on the information obtained on organic carbon content in soils and remote sensing data, a mapping model reflecting the spatial variation of organic carbon content in the upper horizons (0–20 cm) of soils in Central Caucasus was created using digital soil modelling and mapping technology. For modelling we applied a multifunctional approach involving a combination of actual data on the organic carbon content (training set) with data derived from external sources of information (remote sensing data) that was processed using a stepwise discriminant analysis. The necessity of creating a model of organic carbon distribution in soils separately for artificial (agrocenoses) and natural biogeocenoses was established using statistical methods of analysis. As a result of combining two hypothetical models, a verified model reflecting the real picture of changes in the organic carbon content in soils of Central Caucasus was obtained. The reliability of the model was 68%. It contains actual data on organic carbon content in natural and agrogenic soils of Central Caucasus. This model is a necessary tool for making decisions to maintain or increase current soil carbon stocks under conditions of climate change and increasing anthropogenic impact.