Unraveling the Distribution of Black Carbon in Chinese Forest Soils Using Machine Learning Approaches

Black carbon (BC) is a highly persistent yet poorly understood component of forest soil carbon reservoirs, while its inventory, distribution, and determining factors in forest soils on a large geographic scale remain unclear. Here, we characterized soil BC across 68 Chinese forest sites using benzene polycarboxylic acid method and developed machine learning (ML) models to predict and interpret potential impacts of soil organic matter (SOM) properties, soil physiochemical properties, meteorological conditions, wildfire history, and microbial diversity on BC. Results revealed that SOM properties were the most critical in predicting BC, complemented by the negative impact of mean annual temperature and alkaline mineral composition. The superior prediction accuracy for BC with higher condensed aromaticity (more benzene hexa- and penta-carboxylic acid monomers) likely results from its simpler sources and greater resistance to transformation. This study introduces an effective ML model for predicting and interpreting soil BC inventory to better understand BC cycling. The black carbon (BC) derived from incomplete biomass combustion serves as a potential reservoir to store carbon in land ecosystems. Despite BC's significance in carbon sequestration, its inventory and distribution on a large geographic scale remains elusive. As such, we characterized both the quantity and quality of soil BC for forest sites across China. We find that soil BC content averaged 1.99 +/- 1.94 mg C g-1 and constituted 8.8% +/- 4.9% of soil organic carbon, without showing a clear geographic distribution pattern. Besides the well-acknowledged predictive power of machine learning (ML) methodologies, they were introduced to determine the impacts of soil properties and environmental parameters in controlling the BC distribution. We discovered that soil organic matter properties were the most important parameters in predicting BC content, accounting for over 50% of the contribution in model construction based on mean absolute Shapley value. Alongside meteorological conditions, they were further extracted as the key parameters for predicting BC content to simplify the ML model. As a promising complement to traditional geochemical approaches, this research highlights the potential of leveraging ML to predict and interpret BC inventory on a global scale. The black carbon (BC) inventory and distribution in Chinese forest soils on a large geographic scale were first determined Machine learning (ML) models exhibited superior prediction accuracy for BC with higher condensed aromaticity Soil organic matter properties were the most critical in predicting BC content