| 长江经济带新型城镇化与碳排放时空耦合及影响因素：基于XGBoost-SHAP模型 |
| 摘要点击 862 全文点击 44 投稿时间：2025-07-17 修订日期：2025-10-17 |
| 查看HTML全文
查看全文 查看/发表评论 下载PDF阅读器 |
| 中文关键词 长江经济带 新型城镇化 碳排放强度 耦合协调 XGBoost-SHAP模型 非线性机制 |
| 英文关键词 Yangtze River Economic Belt new-type urbanization carbon emission intensity coupling coordination XGBoost-SHAP model nonlinear mechanisms |
| DOI 10.13227/j.hjkx.202507233 |
|
| 中文摘要 |
| 随着全球气候变化加剧，碳排放问题对可持续发展的影响日益加深. 为探究长江经济带新型城镇化与碳排放的协调关系，选取2010~2022年长江经济带108个城市为研究对象，通过熵权TOPSIS法和降尺度模型等方法测度新型城镇化水平与碳排放强度，借助修正耦合协调模型分析两者的时空耦合演变. 并引入XGBoost-SHAP模型，探究耦合协调度的关键影响因素及其非线性作用机制. 结果表明：①新型城镇化与碳排放强度在空间和时间上呈现显著差异，两者表现出明显的负相关关系. 就两者水平而言，下游优于中游、中游优于上游. ②耦合协调度整体不断提升，但区域间差异与内部两极分化依然突出，空间格局由最初的“东高西低”演变为“多核心”分布形态. 下游地区核心地带的优势地位逐步扩大. ③各因素作用机制呈现非线性与区域异质性，人口密度与收入水平对协调度贡献最大，而产业结构与政府支出等变量在不同区域出现作用机制和阈值差异. 最后，根据研究发现提出基于区域差异的对策建议，以期为长江经济带实现绿色、高质量发展提供可行路径与决策参考. |
| 英文摘要 |
| As global climate change intensifies， carbon emissions increasingly constrain sustainable development. To investigate the coordinated relationship between new urbanization and carbon emissions in the Yangtze River Economic Belt， this study examines 108 cities within the region from 2010 to 2022. The entropy-weighted TOPSIS method and downscaling models were employed to measure new urbanization levels and carbon emission intensity， while an adjusted coupling coordination model analyzed their spatiotemporal coupling evolution. Furthermore， the XGBoost-SHAP model was introduced to investigate key factors influencing coupling coordination and their non-linear mechanisms. The results show： ① New-type urbanization and carbon-emission intensity exhibited marked spatial and temporal variation with a clear negative association； performance ranked as downstream > midstream > upstream. ② Coupling-coordination improved overall， yet regional disparities and internal polarization persisted. The spatial pattern shifted from an initial “high-east/low-west” distribution to a “multi-core” configuration， with downstream cores consolidating their advantage. ③ Mechanisms were non-linear and region-specific： Population density and income level contributed most to coordination， whereas industrial structure and government expenditure displayed threshold effects that differed across regions. On this basis， we propose region-differentiated policy recommendations to provide actionable pathways and decision support for green， high-quality development across the Belt. |
