A regional geospatial model of the variability of forest litter carbon stocks and subsequent assessment of environmental factors influencing soil organic carbon accumulation were performed in Republic of Karelia and the Karelian Isthmus, Leningrad Oblast. Modeling was based on 137 field samples collected between 2007 and 2010 as part of the International Co-operative Programme on Assessment and Monitoring of Air Pollution Effects on Forests (ICP Forests). Spatial predictors characterizing soil formation factors were used for modeling soil organic carbon stocks according to the SCORPAN model. Based on correlation analysis and the exclusion of multicollinear variables, the set of independent predictors was finalized. Random Forest machine learning algorithm was applied for regression modeling. The resulting carbon stock model explains 46% of the variability in carbon stocks across the study area (R = 0.46; MAE = 1.84; RMSE = 2.59). The average carbon stock in the forest litter is 3.9 kg/m, with a minimum of 1.4 kg/m and a maximum of 7.4 kg/m. The study revealed that the most significant factors influencing the variability of forest litter carbon stocks in Karelia and the Karelian Isthmus are climate (37.2%), spatial position (22.6%), and vegetation (17.9%).
A.N. Narykova (Wed,) studied this question.