Re-measurement of agricultural carbon emission rates: Drivers, regional differences and dynamic evolution

Abstract

This research paper introduces a novel approach by combining a Backpropagation (BP) neural network with a non-angular and non-radial directional distance function to construct a BPNN-DDF model. This innovative model evaluates, decomposes, and analyzes China’s agricultural sector’s carbon emission rate across nine key subregions between 2010 and 2021. The key findings of this study are that China’s agricultural carbon emission rate is decreasing, primarily due to technological advancements rather than technological efficiency. Subregions with robust economies and stable climates exhibit higher carbon emission efficiency, whereas those with underdeveloped economies, low agricultural technology, and volatile climates show relatively lower efficiency. The Dagum Gini coefficient analysis reveals a widening disparity in carbon emission rates among agricultural subregions, escalating from 0.174 in 2010 to 0.425 in 2021, indicating a growing gap between subregions that demands immediate attention. The kernel density distribution demonstrates an overall upward trend in China’s carbon emission efficiency but also highlights an increasing divergence among subregions, particularly between the South China Area, the Huang-Huai-Hai Plain, and other regions. Therefore, this paper posits that strategies focusing on technological progress, sustainable agricultural development, regional development initiatives, and addressing inter-subregional imbalances will be crucial pathways for China’s future low-carbon agricultural development.