Improvement in precision of forest soil organic C (SOC) change estimates are limited by the high spatial variability of SOC, while accuracy problems are associated with potential systematic errors in repeated measurements. This affects C estimates based on soil sampling schemes, and may result in inadequate process representation in models. However, combining traditional sample plot based inventories with spatially explicit data from other sources would enable a more accurate and precise uptake/emission estimates in GHG reporting with proper follow up mitigation actions on higher spatial resolution. Mainly we are working on improving the accuracy in SOC change estimates by combining the SOC change information with high-resolution auxiliary environmental and landform data to improve current model applications. These auxiliary environmental and landform data include high-resolution (2 m) digital elevation model data, forest parameters, spectral and climate data (temperature, precipitation). Geospatial analyses as well as raster and vector processing will be carried out using SAGA GIS, Orfeo toolbox and R software. Mainly machine learning techniques will be evaluated for modelling.