欢迎访问《草业学报》官方网站,今天是

草业学报 ›› 2026, Vol. 35 ›› Issue (5): 1-19.DOI: 10.11686/cyxb2025327

• 研究论文 •    

基于GEE的渭河流域生态质量动态监测及其驱动机制

王亦波1(), 韩新宁2(), 安可3, 张梦婕4, 田慧慧1, 拓行行5, 张潇珊5, 叶发明6,7, 尹子鸣5, 马晓瑞5, 杨庆6,7, 师涛5, 李伟5()   

  1. 1.西北农林科技大学草业与草原学院,陕西 杨凌 712100
    2.宁夏师范大学资源环境与生命科学学院,宁夏 固原 756000
    3.定边县气象局,陕西 定边 718600
    4.宁夏大学农学院,宁夏 银川 750021
    5.西北农林科技大学水土保持科学与工程学院,陕西 杨凌 712100
    6.中国科学院水利部水土保持研究所,陕西 杨凌 712100
    7.中国科学院大学现代农业科学学院,北京 111049
  • 收稿日期:2025-08-14 修回日期:2025-09-15 出版日期:2026-05-20 发布日期:2026-03-11
  • 通讯作者: 韩新宁,李伟
  • 作者简介:hanxinning@163.com
    Corresponding author. E-mail: liwei2013@nwsuaf.edu.cn
    王亦波(2000-),男,浙江长兴人,在读硕士。E-mail: yiboyes@126.com
  • 基金资助:
    国家自然科学基金项目(42277464);国家重点研发计划项目(2022YFF1302800);陕西省重点研发项目(2024SF-YBXM-545);宁夏自然科学基金(2025AAC030610)

Google Earth Engine-based dynamic monitoring of ecological status in the Weihe River Basin and mechanisms driving it

Yi-bo WANG1(), Xin-ning HAN2(), Ke AN3, Meng-jie ZHANG4, Hui-hui TIAN1, Hang-hang TUO5, Xiao-shan ZHANG5, Fa-ming YE6,7, Zi-ming YIN5, Xiao-rui MA5, Qing YANG6,7, Tao SHI5, Wei LI5()   

  1. 1.College of Grassland Agriculture,Northwest A&F University,Yangling 712100,China
    2.College of Resources,Environment and Life Sciences,Ningxia Normal University,Guyuan 756000,China
    3.Dingbian County Meteorological Bureau,Dingbian 718600,China
    4.College of Agriculture,Ningxia University,Yinchuan 750021,China
    5.College of Soil and Water Conservation Science and Engineering,Northwest A&F University,Yangling 712100,China
    6.Institute of Soil and Water Conservation,Chinese Academy of Sciences & Ministry of Water Resource,Yangling 712100,China
    7.College of Advanced Agricultural Sciences,University of Chinese Academy of Science,Beijing 111049,China
  • Received:2025-08-14 Revised:2025-09-15 Online:2026-05-20 Published:2026-03-11
  • Contact: Xin-ning HAN,Wei LI

摘要:

渭河流域作为黄河流域的核心支流,其生态质量优劣直接关系到区域可持续发展进程。本研究利用Google Earth Engine(GEE)平台构建了基于遥感的生态指数(RSEI),系统分析了2000-2024年间渭河流域生态质量的时空变化及其驱动机制。结果表明:研究期内流域年均RSEI为0.37~0.53,整体以2.73×10-3·a-1的速率显著改善,空间上呈现“东南高西北低、边缘高中心低”梯度,秦岭北麓生态脊梁凸显。天水-平凉、延安-庆阳等地RSEI极显著增加,西安-咸阳都市圈显著退化,同时泾河上游矿区及都市圈西扩区存在24.4%的持续退化风险,高波动区占59.3%,生态恢复瓶颈突出。通过SHAP解释XGBoost发现,生态质量空间异质性主要由气候因素主导,其中实际蒸散发与地表温度的驱动作用较强,影响力远高于地形、人类活动等其他因素,且适宜区间内(10~25 ℃)的气温升高、降水量增加(550~600 mm)和阈值范围内人口密度形成协同作用,对RSEI提升具有显著正向效应。本研究在气候变化背景下分析和监测渭河流域RSEI的动态演变,为黄河流域生态保护修复和高质量发展提供了科学依据。

关键词: 遥感生态指数, XGBoost模型, 机器学习, 渭河流域

Abstract:

As an important tributary of the Yellow River, the ecological status of the Weihe River Basin plays a significant role in regional sustainable development. This study employed the remote sensing ecological index (RSEI), developed on the Google Earth Engine (GEE) platform, to comprehensively analyze the spatial and temporal variations in ecological status of the Weihe River Basin from 2000 to 2024, along with its underlying drivers. Methods including Theil-Sen trend analysis, Mann-Kendall test, Hurst exponent, and coefficients of variation were used to examine these changes. Additionally, the eXtreme Gradient Boosting (XGBoost) model, enhanced with SHapley Additive exPlanations (SHAP) values, was used to identify and elucidate the primary factors influencing the spatial heterogeneity of ecological status. Over the studied period, the RSEI of the Weihe River Basin is projected to increase from 0.37 to 0.53, representing an initial rise followed by stabilization. The spatial distribution shows that the ecological quality is higher in the southeast and edges, and lower in the northwest and center, with the ecological backbone formed along the northern foot of the Qinling Mountains. Meanwhile, ecological quality in the Guanzhong urban agglomeration has deteriorated. The analysis reveals a 27.4% reduction in areas of poor ecological quality, expansion of medium-quality areas, and a slow growth in high-quality areas, indicating that ecological restoration is approaching a bottleneck. Predictions suggest that 72.7% of the basin will continue to show improvements, whereas 24.4%-particularly in the mining areas of upper Jinghe River and the western expansion zones of the metropolitan area-face ongoing degradation risks. High volatility is evident in 59.3% of the basin. The spatial heterogeneity of RSEI in the Weihe River Basin is the result of the interaction of multiple factors, and there are complex synergistic and antagonistic relationships between the factors, which are driven by the dual combination of “climate mastery and anthropogenic amplification”: climatic factors [actual evapotranspiration (AET), land surface temperature (LST), potential evapotranspiration (PET), temperature (TMP), precipitation (PRE)] affect RSEI through the water-heat balance, of which precipitation is the key regulator of the negative effect of PET, and 20-25 ℃ is the optimal temperature window for water-heat synergy; AET and LST promote positive synergy within a moderate temperature range; and PET and PRE produce negative antagonism in areas with sufficient precipitation. At the same time where population density exceeds 600 persons·km-2, urbanisation triggers ecological degradation through the heat island effect (amplifying LST) and surface hardening (weakening PRE infiltration). The methodologies and findings of this study provide a detailed understanding and ongoing monitoring of the dynamic evolution of RSEI in the Weihe River Basin amidst climate and population changes. This research offers a scientific foundation for the ecological protection, restoration, and sound ecological development of the Yellow River Basin.

Key words: remote sensing ecological index, XGBoost model, machine learning, Weihe River Basin