What is digital soil mapping?

As shown in the video above, digital soil mapping uses soil samples and biogeochemically-informed predictors — derived from remote sensing and environmental data — to estimate soil organic carbon stocks and changes across space.

Perennial’s proprietary model, ATLAS-SOC, has integrated more than 350,000 soil samples and is peer-reviewed, validated by third parties, and designed to quantify SOC across diverse regions, land uses, and crop types.

For each project, we train ATLAS-SOC on local samples to build a model tailored to that landscape. It then fills in the gaps between samples, providing hundreds to thousands of times more SOC estimates than sampling alone. Uncertainty is modeled at the pixel level.

By combining dense data inputs with pixel-level uncertainty modeling, we achieve low uncertainty while reducing the need for extensive sampling.