Analysis of spatial and temporal carbon emission efficiency in Yangtze River Delta city cluster — Based on nighttime lighting data and machine learning

Abstract

Improving carbon emission efficiency(CEE) is crucial to reducing CO2 emissions. Most studies on CO2 emission are conducted at national and industrial scales, and city-scale studies still need to be included. In order to collect more consistent city- and county-scale CO2 emission data, the sparrow search neural network is trained to fit the energy consumption CO2 emissions with nighttime light in this study. Additionally, using the SBM-DEA model and spatial econometric techniques, the CEE values of 27 cities in the Yangtze River Delta region (YRDR) from 2000 to 2020 were examined from the perspective of total factor inputs. The findings demonstrate that CEE's general trend is erratic and uneven. The CEE value of the YRDR decreases from 0.720 in 2000 to 0.628 by 2020, which means that the YRDR has redundant capital and labour inputs and insufficient economic output. The low value carbon efficiency areas are mainly concentrated in the western part of the YRDR, i.e. the Anhui Province region. Shanghai, Wuxi and Suzhou have high carbon efficiency values of 1.21, 2.08 and 1.00 respectively, and are exemplary cities in terms of carbon efficiency, while the rest of the cities have varying degrees of efficiency loss. Taking Chizhou-Jiaxing as the middle line, the CEE pattern in the YRDR presents a state of “low in the middle and high at each end,” and center of gravity for CEE generally shifts southward. Additionally, the cold-spot areas of CEE are concentrated in the southern part of Anhui Province, and develop a low-efficiency zone with Chizhou, Anqing, and Xuancheng as clusters and spreading outwards. Overall, this paper significantly narrows the spatial scale of carbon accounting studies and the findings can be applied to the formulation of customized carbon reduction policies.