3D+T mapping reveals soil organic matter changes between 1953 and 2022 at 25m resolution in the Netherlands

The Soil Deal for Europe is expected to yield valuable outcomes, including a better understanding of how soil health supports soil functions. Soil organic matter (SOM) is crucial for multiple soil functions, including carbon and climate regulation and nutrient cycling. Consequently, a better understanding of soil health and soil functions largely hinges on the role of SOM, which varies across space, depth, and time. As a result, high-resolution, spatio-temporally explicit assessment of SOM is required to enable management and land use decisions that are tailored to local soil conditions. We developed a modelling platform in 3D space and time (3D+T) as a new paradigm for monitoring and mapping SOM. It provides annual predictions of SOM and its uncertainty in the Netherlands at 25m resolution between 0-2m depth from 1953-2022. This innovative tool enables the identification of areas with high risk of greenhouse gas emissions and areas with potential for carbon sequestration. Furthermore, it represents a crucial first step towards establishing scientifically-backed threshold values that link SOM to soil health and carbon credits. From 1953 to 2022 we found SOM decreases of >25% in bogs and brook valleys, average decreases of 0.1-0.3% on croplands on mineral soils, and changes between 10-25% on reclaimed land due to land subsidence. Hence, SOM was highly variable over depth and time. The novel use of 3D+T spatially explicit environmental covariates has the advantage that we let machine learning detect complex relationships between SOM and peat occurrence in 3D space and time. Therefore, the implications of this research extend beyond the examined spatio-temporal domain. We can identify the key drivers of SOM changes and predict future dynamics. In addition, our findings are relevant for areas worldwide undergoing peatland conversion, land reclamation, and agricultural intensification.