中原城市群水生态空间演变时空特征及其驱动机制——基于时空立方体与可解释机器学习分析
作者:
作者单位:

作者简介:

通讯作者:

中图分类号:

基金项目:

河南省教育厅人文社科一般项目(2025-ZZJH-045);驻马店市科技创新青年专项项目计划支持(QNZX202402);中国地质大学(武汉)国家地理信息系统工程技术研究中心开放基金资助(2023KFJJ03);教育部人文社会科学研究规划基金项目(23YJA630003)


Spatiotemporal characteristics and driving mechanisms of hydro-ecological space evolution in Zhongyuan Urban Agglomeration: an analysis based on a space-time cube and interpretable machine learning
Author:
Affiliation:

Fund Project:

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 文章评论
    摘要:

    中原城市群是推动黄河流域高质量发展的核心空间载体,科学认知其水生态空间演变的时空特征及形成机制,对黄河下游地区水生态空间管制和国土空间规划具有重要支撑作用。采用时空立方体刻画2000-2023年长时序中原城市群水生态空间年际演变动态及典型模式,从规模-位置2个维度综合分析空间结构转型特征,进而利用Mantel test矩阵分析水生态空间结构转型与驱动因素的相关性关系,在此基础上运用XGBoost模型和可解释机器学习VIVI-PDP框架从驱动因素重要程度、交互作用强度和非线性依赖关系分析演变机制。结果表明:①近23年中原城市群水生态空间增加805.53km2,增幅高达33.52%,整体呈现"上升-平稳-上升-平稳"的动态变化趋势,在稳定的基础上逐步提升;②"农业空间-水生态空间"的动态转换现象尤为显著,且在城市群的六大功能分区中差异明显,其中高效生态示范区的水生态空间转型最为活跃,核心发展区与跨区域协同发展区相对活跃,而转型创新发展区和承接产业转移区转换度较低;水生态空间动态转换的高-高聚类区,即水生态空间的转入和转出均较为频繁的区域,主要集中于水资源丰富的西部、南部与东部地区,低-低聚类区则多位于北部和中部地区,其水生态空间的转入和转出均较为有限;③自然地理基础与交通区位条件是中原城市群水生态空间动态转型过程中的主导因素,然而,在水生态空间向农业空间的转出和城镇空间向水生态空间的转入过程中,社会经济因素作用逐渐凸显,自然与人文因素的交织作用使得转型过程呈现出多重因素交织、区域差异显著的驱动机制。

    Abstract:

    The Zhongyuan Urban Agglomeration functions as a strategic spatial platform driving sustainable development in the Yellow River Basin. A scientific understanding of the spatiotemporal differentiation characteristics and formation mechanisms of hydro-ecological space evolution within this region is essential to support water ecological space use regulation and integration into the national territorial spatial planning system in the lower Yellow River area. This study employs a space-time cube model to characterize the interannual dynamics and typical patterns of hydro-ecological space within Zhongyuan Urban Agglomeration from 2000 to 2023. A comprehensive analysis of spatial structure transformation is conducted from two critical dimensions-scale and location-to systematically capture the extent, spatial distribution, and dynamic shifts of hydro-ecological space within the Zhongyuan Urban Agglomeration. This analysis enables a refined understanding of spatial configuration changes, expansion-contraction dynamics, and regional structural adjustments over time. To further investigate the underlying drivers of these transformations, the study employs a Mantel test matrix to quantitatively examine the correlation between hydro-ecological spatial structure evolution and a range of potential influencing factors, including natural geographic attributes, socioeconomic conditions, and infrastructural developments. By assessing the strength and significance of these spatial associations, the study identifies key determinants shaping hydro-ecological space changes. Building upon this foundation, an XGBoost model is integrated with an interpretable machine learning framework, VIVI-PDP, to conduct an in-depth exploration of the evolution mechanisms. This approach allows for a precise evaluation of factor importance, the strength of interactions among key variables, and the complex nonlinear dependencies influencing hydro-ecological space transformation.The results reveal that: ①Over the past 23 years, hydro-ecological space expanded by 805.53 km2 (33.52% growth rate), exhibiting a phased evolutionary trajectory: expansion→stabilization→expansion→stabilization, demonstrating cumulative ecological resilience. ②The dynamic conversion between "agricultural space-hydro-ecological space" is particularly pronounced, with significant differences across the six functional zones of the urban agglomeration. Notably, hydro-ecological space transformation is most active in high-efficiency ecological demonstration zones, relatively active in core development and cross-regional coordinated development areas, and less pronounced in transformation and innovation development areas and industrial transfer zones. High-intensity transition clusters (HH-type) predominantly occur in resource-rich western/southern/eastern zones, while low-low clustering (LL-type) areas are mainly located in the northern and central areas, where both the influx and outflux of hydro-ecological space are relatively limited. ③Natural geographic conditions and transportation location are the primary driving factors in the dynamic transformation of hydro-ecological space. However, during the transition from hydro-ecological space to agricultural space and from urban space to hydro-ecological space, socioeconomic drivers emerge as dominant determinants. The interplay of natural and human factors results in a transformation process manifesting multifactorial synergies with pronounced spatial heterogeneity.

    参考文献
    相似文献
    引证文献
引用本文

张轲,尹力,赵浪,魏伟,薄立明.中原城市群水生态空间演变时空特征及其驱动机制——基于时空立方体与可解释机器学习分析.生态学报,2025,45(10):4697~4715

复制
分享
文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数: