A new study in JGR: Biogeosciences finds that three widely used soil biogeochemistry models perform poorly when predicting soil organic carbon (SOC) stocks in sub-Saharan Africa. The models were developed using data from North Temperate regions and overemphasize net primary productivity while missing important controls like organo-mineral interactions and exchangeable calcium. Even when fitted with local observations, the models still fail, suggesting the underlying structure is wrong. Soil organic carbon is one of the largest carbon pools in the biosphere. Inaccurate models mean we cannot reliably estimate current carbon storage or predict future changes in understudied regions. This has direct implications for carbon credit projects that rely on soil carbon sequestration, especially in tropical and subtropical areas where data is scarce.
