Machine learning enhancing biochar abatement predictions: Advancing China climate goals for food production and promoting application

Under 1.5 °C target, agricultural food production faces severe challenges. While biochar is recognized as a negative emission technology, data limitations and complex factors uncertainties hinder accurate quantification of its potential and development of adoption strategies. Here, we developed a machine learning model to generate high–resolution maps of N 2 O reduction and soil organic carbon (SOC) sequestration from biochar application to China staple crops based on meta–dataset of field observations, and assessed the potential economic benefits of incorporating machine learning–guided biochar carbon credits into voluntary carbon market. Results showed that 476.02 CO 2 –eq from N 2 O reduction, SOC sequestration and fossil fuel offsets through biochar and 122.34 Tg CO 2 –eq of allowable GHG emission from the climate target, could keep GHG emission from crop production within the carbon budget limit of 1.5 °C. Meanwhile, biochar projects included in the voluntary carbon market could reduce biochar price by 40.2 %, increase project profits by 2.4 % and farmers income by 22.3 %, overcoming cost–related promotion challenges. Our results provided feasible options for controlling GHG emission from staple food production to meet climate goals, while contributing to increase the potential for biochar application in carbon markets.