Identification and interaction analysis of environmental and microbial drivers of CO2-equivalent emissions in green waste composting using explainable machine learning

Composting plays a crucial role in the resource recovery of green waste; however, the emission of greenhouse gases (CO 2 , CH 4 , N 2 O) during the process may compromise its environmental benefits. Traditional linear models show clear limitations in identifying key drivers and capturing variable interactions, making them insufficient for revealing the complex dynamics of CO 2 -equivalent emissions. In this study, four machine learning models were developed and compared using five combinations of input datasets to systematically assess the joint influence of environmental factors and microbial communities on greenhouse gas emissions. Results showed that the XGBoost model performed best when integrating both environmental factors and microbial taxa (bacteria and fungi), achieving high predictive accuracy ( R 2 = 0.9863, RMSE = 319.17). SHapley Additive exPlanations analysis identified Time, Temperature, Rhizophagus , and Steroidobacter as major contributors, while partial dependence plots revealed notable non-linear interactions between microbial and temperature variables. Mantel tests further confirmed synergistic regulation between environmental and microbial drivers. The findings highlight the essential role of integrating environmental and microbial features for accurate prediction and mechanistic understanding. The explainable modelling framework proposed offers theoretical and methodological support for emission mitigation strategies in green waste composting, facilitating ecological optimisation and low-carbon transformation under the ‘dual carbon’ targets. • ML models are used to predict and optimise greenhouse gas emissions. • The best input feature set combines environmental factors and microbial communities. • XGBoost yields the highest prediction accuracy with an R² of 0.9863. • Rhizophagus and Steroidobacter are key genera suppressing GHG emissions. • Time and temperature are the main environmental drivers of GHG emissions.