Analysis of carbon emission characteristics and driving factors at sub-municipal level

Abstract Addressing climate change calls for granular carbon management at the sub-municipal level to enable targeted local actions. This study examines Tongzhou District, a rapidly developing Beijing sub-center, to demonstrate a transferable framework for such analysis. Following the IPCC guideline, this study quantified energy-related carbon equivalent emissions for 2018–2020 by overcoming key limitations of traditional macro-scale carbon accounting through multi-source data integration, reallocation of transport sector energy use, and testing of localized emission factors for key fuels. Results show that CO 2 constitutes >99% of total emissions. Natural gas, gasoline, and kerosene rank as the top three emitting fuels. Transport, residential, industrial, and service sectors are identified as key sources. Driving factors were analyzed via Kaya-LMDI decomposition. During 2018–2019, population growth exhibited a positive driving effect of +6.1% on emissions while energy efficiency contributed a −2.2% reduction effect. In the subsequent 2019–2020 period, however, population growth drove a +5.6% increase, which was outweighed by significant reduction effects from energy efficiency (−11.8%), energy structure (−8.2%), and economic activity (−1.2%). Quantitative uncertainty analysis via error propagation revealed low carbon accounting uncertainties of 7.63% (2018), 7.55% (2019), and 7.53% (2020). This study not only informs Tongzhou’s low-carbon policy but also provides a validated, scalable methodology for micro-scale carbon accounting, enhancing the precision of urban low-carbon development measures.