Ensemble intelligence prediction algorithms and land use scenarios to measure carbon emissions of the Yangtze River Delta: A machine learning model based on Long Short-Term Memory

PLoS One. 2024 Dec 9;19(12):e0311441. doi: 10.1371/journal.pone.0311441. eCollection 2024.

Abstract

Land use in urban agglomerations is the main source of carbon emissions, and reducing them and improving land use efficiency are the keys to achieving sustainable development goals (SDGs). To advance the literature on densely populated cities and highly commercialized regions, this research evaluates the total-factor carbon emission efficiency index (TCEI) of 27 cities in China's Yangtze River Delta (YRD) urban agglomeration at different stages from 2011 to 2020 using two-stage dynamic data envelopment analysis (DEA). The study carries out regression analysis and a long-short-term memory model (LSTM) to respectively filter out the factors and predict TCEI. The results indicate the following. (1) The total efficiency of 27 cities has significantly improved from 2011 to 2020, and there are obvious spatial heterogeneity characteristics. (2) In terms of stages, most cities' efficiency values in the initial stage (energy consumption) exceed those in the second stage (sustainable land utilization). (3) In terms of influencing factors, urban green space's ability to capture carbon has a notably positive correlation with carbon emission efficiency. In contrast, the substantial carbon emissions resulting from human respiration are a negative factor affecting carbon emission efficiency. (4) Over the forthcoming six years, the efficiency value of land use TCEI in the YRD urban cluster is forecasted to range between 0.65 and 0.75. Those cities with the highest performance are projected to achieve an efficiency value of 0.9480. Lastly, this research investigates the interaction between actors and land resources on TCEI, resulting in a beneficial understanding for the former to make strategic adjustments during the urbanization process.

MeSH terms

  • Algorithms
  • Carbon* / analysis
  • China
  • Cities
  • Environmental Monitoring / methods
  • Humans
  • Machine Learning*
  • Memory, Short-Term
  • Rivers* / chemistry
  • Sustainable Development
  • Urbanization

Substances

  • Carbon

Grants and funding

This study was supported by Jiangsu Province Social Science Foundation Project (22GLD019), Major Project of Philosophy and Social Science Research in Universities of Jiangsu Province(2022SJZD053).