Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
To provide a more comprehensive reconstruction of China’s energy consumption, this paper built a machine-learning-based geospatial model that shows great accuracy in recovering historical biomass consumption data using the household survey dataset for China, combined with province-level characteristics and spatiotemporal information. Our study suggested that 6.9 ± 2.6 giga-tons of coal equivalent of biomass were uncounted in China’s statistics, representing 15.9 ± 6.0 percent for China and 2.5 ± 0.9 percent for global final energy consumption. This new estimate significantly reshaped our understanding of China’s energy composition, sectoral mix, indoor air pollutants, and the factors driving energy consumption. These findings provide a replicable template for developing countries hoping to uncover the biomass demand to better design public policy to achieve Sustainable Development Goals. JEL Classification: Q41, Q53, R12, C81, O13