EUVIMEDEUVIMED entdecken
Crossref

Hybrid LSTM-XGBoost framework with city embeddings for CO2 emissions forecasting and scenario comparison

Jing Li, Zhao Wang, ZhuoWen Yan, Yue Yang

Environmental Research · 2026-07 · 301 · 124476

DOI: 10.1016/j.envres.2026.124476 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Environmental Research, 2026-07. Verfasst von Jing Li und 3 weiteren Autor:innen. Beteiligte Einrichtung: Department of Resources and Environment, Hebei Minzu Normal University, Chengde, 067600, China. Hauptthemen laut MEDLINE: Carbon Dioxide, Air Pollutants, Environmental Monitoring, Air Pollution. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 67 Literaturangaben.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Developments of cities are always accompanied by environmental sacrifice, especially increasing CO2 emission is one of the significant drivers of climate warming. Accurate city-level CO2 emissions forecasting is critical for urban carbon management, where those cities are undergoing economic transitions and exhibiting divergent carbon emissions trends. Taking Hebei as an example, this study developed a hybrid deep learning framework combining LSTM networks and XGBoost models via adaptive power-law weighting, validated through a comprehensive ablation study, and further compared six scenario predictions. The model achieved an R2 of 0.958 on the test set by balancing overall accuracy with city-level robustness, capturing the non-linear dynamics and spatiotemporal heterogeneity across cities. Monte Carlo uncertainty propagation analysis confirmed model robustness under both moderate (σ = 0.10) and conservative (σ = 0.20) emission data perturbation, with scenario rankings remaining stable across 500 iterations. Feature importance analysis revealed that the relative contributions from urbanization (36.1%), economic indicators (34.4%), and spatial characteristics (25.3%) were the primary predictive drivers, with lagged and differenced features outranking their contemporaneous counterparts. Post-hoc semantic analysis of the city embeddings confirmed that the learned representations encode an interpretable industrialization-to-ecological gradient consistent with independently identified city typologies. The clustering analysis identified three urban emission clusters related to different developing stages. Under the business-as-usual scenarios, provincial emissions are expected to reach 1.01 billion tons by 2028, while the comprehensive low-carbon interventions can achieve 13.2% reduction versus the baseline projection. The results show that, for the carbon emission control in historical heavy industrial provinces, an overall low-carbon strategy centered on industrial transformation and ecological improvement is needed. Collectively, this framework offers a transferable analytical approach for evidence-based carbon management and scenario-driven policy optimization in spatially heterogeneous urban systems.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Carbon DioxideAir PollutantsEnvironmental MonitoringAir PollutionBoosting Machine Learning AlgorithmsLong Short Term MemoryCitiesForecastingPredictive Learning Models

Stichwörter der Autor:innen

CO(2) emissions forecastingCity-heterogeneity modelingMulti-source feature integrationScenario-driven policy analysisUrban carbon management

Substanzen

Carbon DioxideAir Pollutants

Autor:innen und Institutionen

  • Department of Resources and Environment, Hebei Minzu Normal University, Chengde, 067600, China.
  • Shijiazhuang Customs District Technology Center, Shijiazhuang, 050051, China.
  • Guangzhou Customs District Technology Center, Guangzhou, 510623, China.
  • Department of Resources and Environment, Hebei Minzu Normal University, Chengde, 067600, China. Electronic address: yangyue@hbun.edu.cn.

Zitieren

APA

Li, J., Wang, Z., Yan, Z., & Yang, Y. (2026). Hybrid LSTM-XGBoost framework with city embeddings for CO2 emissions forecasting and scenario comparison. Environmental Research, 301, 124476. https://doi.org/10.1016/j.envres.2026.124476

Vancouver

Li J, Wang Z, Yan Z, Yang Y. Hybrid LSTM-XGBoost framework with city embeddings for CO2 emissions forecasting and scenario comparison. Environmental Research. 2026;301:124476. doi:10.1016/j.envres.2026.124476

BibTeX
@article{Li2026,
  author = {Li, Jing and Wang, Zhao and Yan, ZhuoWen and Yang, Yue},
  title = {Hybrid LSTM-XGBoost framework with city embeddings for CO2 emissions forecasting and scenario comparison},
  journal = {Environmental Research},
  year = {2026},
  volume = {301},
  pages = {124476},
  doi = {10.1016/j.envres.2026.124476}
}
Kostenlos starten →