Integrating Nighttime Lights and Urban Socioeconomic Attributes for City-Scale CO2 Accounting: A Comparative Study of the Beijing–Tianjin–Hebei Urban Agglomeration and the Tokyo Metropolitan Area

Incomplete city-level energy statistics constrain long-term territorial CO2 accounting. This study examines a top-down allocation framework combining harmonized nighttime lights, employment-based urban typologies, and socioeconomic attributes in the Beijing–Tianjin–Hebei (BTH) region and the Tokyo Metropolitan Area (TMA). DMSP/OLS and NPP/VIIRS observations for 1992–2019 were harmonized, and type-specific regressions reconstructed nighttime-light shares from socioeconomic shares. The resulting weights allocated provincial or prefectural direct energy-related CO2 totals to 13 BTH cities and 312 TMA municipalities for 1995–2019. Against ODIAC, the reported root mean square error (RMSE) decreased from 64.20 to 61.44 Mt CO2 in BTH and from 2.46 to 2.35 Mt CO2 in the TMA, corresponding to descriptive reductions of approximately 4.30% and 4.47%. Adjustment generally increased allocations to resource-extraction, energy-supply, and heavy-industrial groups. These differences characterize allocation behavior and external agreement; shared proxies and differences in accounting boundaries prevent these reductions from being interpreted as independently established accuracy gains. Regional trajectories are inherited from higher-level accounts, while differences between the applications also reflect administrative scale and predictor availability. The framework provides an exploratory approach to territorial spatial accounting. Independent validation and quantified uncertainty are necessary before using the estimates for binding policy decisions.