City-level carbon emissions accounting and mitigation strategies considering urban characteristics: a case study of the Yangtze River Delta region, China.

Abstract

Exploring the disparities in CO₂ emissions (CE) among different city types is essential for formulating effective decarbonization policies. However, due to the limitations of urban energy data, it is challenging to conduct CE analysis at the city level. This study proposes a city-level CE accounting model based on multi-source data to estimate emissions for 41 cities in the Yangtze River Delta (YRD) from 2013 to 2022. Using the K-means algorithm, cities are categorized into five distinct types, while a Geographically and Temporally Weighted Regression (GTWR) model is employed to analyze their spatiotemporal drivers. The results demonstrate that the proposed model in this study significantly enhances the accuracy of city-level CE estimation compared to traditional Nighttime Light (NTL)-based fitting methods, with the R² value increasing from 0.63 to 0.76. Based on this improved model, the result shows that the total CE in the YRD rose from 1,700 Mt in 2013 to approximately 1,931 Mt in 2022. The region exhibits a high degree of emission concentration, where a mere 14.6% of cities with annual emissions exceeding 80 Mt contribute to 40.3% of the total regional emissions. GTWR analysis reveals that population, per capita GDP, industry structure, and energy intensity consistently drive emissions upward, while urbanization exhibits dual effects. Optimizing industrial structure and reducing energy intensity are effective ways to curb the growth of urban CE. This study improves the estimation method for urban CE based on NTL data and provides scientific support for various city types to formulate tailored mitigation pathways.