Abstract
Soil Organic Carbon (SOC) constitutes the largest terrestrial carbon pool. Understanding its dynamics and the environmental factors that influence its behaviour as a sink or source of atmospheric CO2 is crucial to quantifying the carbon budget at a global scale. At the European level, most existing studies accounting for SOC stocks focus on fitting predictive models to ascertain the distribution of SOC. However, developing methodologies for monitoring and identifying the environmental factors that control SOC storage in Europe remains a key research challenge. Here, we present a modelling procedure for mapping and monitoring SOC contents that uses Visible-Near Infrared (VNIR) spectroscopic measurements and a series of environmental covariates to ascertain the key environmental processes that significantly contribute to SOC sequestration processes. Our results show that it follows a geographically non-stationary process in which the influencing environmental factors have different weights depending on the spatial location. This implies that SOC stock modelling should not rely on a single model but on a combination of different statistical models depending on the environmental characteristics of each area. A cluster classification of European soils in relation to those factors resulted in the determination of four groups for which specific models have been obtained. Differences in climate, soil pH, content of coarse fragments, or land cover type are the main factors explaining the differences in SOC in topsoil across Europe. We found that climatic conditions are the main driver of SOC storage at the continental scale, but we also found that parameters like land cover type influence SOC content found at local scales in certain areas. Our methodology, developed at a continental scale, could be used in future research aimed at improving the predictive performance of SOC assessments at the European scale.
