nav emailalert searchbtn searchbox tablepage yinyongbenwen piczone journalimg journalInfo journalinfonormal searchdiv searchzone qikanlogo popupnotification paper paperNew
Spatiotemporal patterns and drivers of landscape ecological risk in the Weihe River Basin based on XGBoost–SHAP
Email:
DOI: 10.13637/j.issn.1009-6094.2026.0721
Published:   2026-09-11
Publication Date:   2026-09-11
Online:   2026-09-11
Mobile reading
Abstract:

This study investigates the spatiotemporal patterns and nonlinear driving mechanisms of landscape ecological risk in the Weihe River Basin from 2000 to 2020. Landscape ecological risk indices were calculated for each grid cell using land use data, and spatial autocorrelation analysis was employed to characterize spatial clustering. The Extreme Gradient Boosting Integrated with Shapley Additive Explanations (XGBoost–SHAP) framework was applied to quantify the nonlinear contributions and threshold effects of various driving factors. The landscape ecological risk values ranged from 0.01 to 0.68 across all evaluation grids. Medium-risk areas constituted the largest proportion, averaging 32.47%, with a gradually increasing trend, while high-risk areas decreased from 8.52% to 8.16% by 2020. Spatially, high-risk zones progressively retreated toward the northern Loess hilly-gully region, coinciding with areas where the Grain for Green Program was implemented. In contrast, low-risk zones in the central Guanzhong Plain exhibited fragmentation and outward displacement. Global Moran’s I values remained above 0.40 (p < 0.001) throughout the study period, indicating significant spatial autocorrelation. H–H clusters were predominantly located in the northern and southern basin, while L–L clusters were concentrated in the central plain. SHAP analysis revealed that distance to residential areas, mean annual temperature, and slope were the three most influential factors, collectively contributing nearly 60% of the total SHAP values. Nonlinear threshold effects were identified for multiple factors. Specifically, NDVI exhibited strong negative effects under low coverage conditions that diminished at higher coverage levels. Precipitation and temperature displayed inverted U-shaped patterns, with effects shifting from positive to negative across different value ranges. Slope showed a U-shaped pattern, while elevation exhibited a positive nonlinear relationship with ecological risk. GDP and population density demonstrated negative nonlinear relationships, suppressing risk only beyond certain developmental thresholds. Distance to river also showed an inverted U-shaped pattern, with risk suppression observed in both nearby and distant zones. These findings indicate that the contributions of driving factors vary dynamically across different value ranges, providing a quantitative basis for spatially targeted risk management in ecologically fragile arid and semi-arid regions.

References

[1] Perkins-Kirkpatrick S E, Stone D A, Mitchell D M, et al. On the attribution of the impacts of extreme weather events to anthropogenic climate change[J]. Environmental Research Letters, 2022, 17(2): 024009.

[2] 欧阳晓, 杨胜苏, 魏晓, 等. 人水关系视角下湖区社会-生态系统恢复力研究进展[J]. 生态学报, 2026, 46(12): 6656–6666.

[3] 彭建, 党威雄, 刘焱序, 等. 景观生态风险评价研究进展与展望[J]. 地理学报, 2015, 70(4): 664–677.

[4] 黄玺铭, 陈勇, 李森, 等. 基于"格局–服务–灾害"的生态脆弱区综合生态风险评价[J]. 水土保持研究, 2026, 33(1): 330–340, 350.

[5] 陆大道. 我国的城镇化进程与空间扩张[J]. 城市规划学刊, 2007(4): 47–52.

[6] 王文静, 逯非, 欧阳志云. 国土空间生态修复与保护空间识别: 以北京市为例[J]. 生态学报, 2022, 42(6): 2074–2085.

[7] 蓝庆新, 韩晶. 中国工业绿色转型战略研究[J]. 经济体制改革, 2012(1): 24–28.

[8] Ye Jinfen, Wang Yubo, Yang Jiayan. Green finance and environmental governance in river basin economies: mechanism analysis based on green innovation and industrial upgrading in the Yangtze River region[J]. Finance Research Letters, 2025, 86: 108872.

[9] Xu Wanxia, Wang Jinman, Zhang Min, et al. Construction of landscape ecological network based on landscape ecological risk assessment in a large-scale opencast coal mine area[J]. Journal of Cleaner Production, 2021, 286: 125523.

[10] Hunsaker C T, Graham R L, Suter G W, et al. Assessing ecological risk on a regional scale[J]. Environmental Management, 1990, 14, 325–332.

[11] 肖杨, 毛显强. 区域景观生态风险空间分析[J]. 中国环境科学, 2006(5): 623–626.

[12] 刘凤莲, 罗芹芹, 杨博文, 等. 泾河流域景观生态风险时空分异特征[J]. 贵州师范大学学报(自然科学版), 2026, 44(1): 108–118.

[13] Chang Sen, Wei Yaqi, Dai Zhenzhong, et al. Landscape ecological risk assessment and its driving factors in the Weihe River Basin, China[J]. Journal of Arid Land, 2024, 16(5): 603–614.

[14] 王舒, 刘凤莲, 杜汶胶, 等. 滇西北高原湖泊流域景观生态风险时空演变及驱动力识别[J]. 环境科学, 2025, 46(5): 3114–3126.

[15] Xie Hualin, Wang Peng, Huang Hongsheng. Ecological risk assessment of land use change in the Poyang Lake eco-economic zone, China[J]. International Journal of Environmental Research and Public Health, 2013, 10(1): 328–346.

[16] 高庆彦, 潘玉君, 吴仕海, 等. 多情景下曲靖市土地利用与景观生态风险调控[J]. 安全与环境学报, 2025, 25(8): 3273–3283.

[17] Zhou Di, Shi Ping, Wu Xiaoqing, et al. Effects of urbanization expansion on landscape pattern and region ecological risk in Chinese coastal city: a case study of Yantai City[J]. The Scientific World Journal, 2014, 2014: 821781.

[18] 李建华, 向明顺. 大熊猫国家公园(四川片区)景观生态风险分析及预测[J]. 内江师范学院学报, 2026, 41(6): 37–45.

[19] Chen Jian, Dong Bin, Li Haoran, et al. Study on landscape ecological risk assessment of hooded crane breeding and overwintering habitat[J]. Environmental Research, 2020, 187: 109649.

[20] 罗芹芹, 刘凤莲, 朱柯权, 等. 赤水河流域(贵州段)景观生态风险时空分异及空间自相关分析[J]. 水土保持研究, 2025, 32(3): 282–290.

[21] 杨华, 周辉, 汤静, 等. 洞庭湖区城市景观生态风险评价及驱动力分析: 以常德市和益阳市为例[J]. 城市学刊, 2024, 45(1): 23–31.

[22] 袁荣燕, 张宁, 邵丽文, 等. 1990—2020年间内蒙古自然保护区景观生态风险评估及驱动因素分析[J]. 内蒙古大学学报(自然科学版), 2023, 54(5): 519–528.

[23] Li Yaqian, Chen Youliang, Karimian H, et al. Spatiotemporal analysis of air quality and its relationship with meteorological factors in the Yangtze River Delta[J]. Journal of Elementology, 2020, 25(3): 1059–1075.

[24] 刘敏, 蔡海生, 张学玲, 等. 基于Geodetector–GTWR的江西省景观生态风险及驱动因素[J]. 环境科学, 2025, 46(10): 6455–6464.

[25] Yu Tao, Bao Anming, Xu Wenqiang, et al. Exploring variability in landscape ecological risk and quantifying its driving factors in the Amu Darya Delta[J]. International Journal of Environmental Research and Public Health, 2019, 17(1): 79.

[26] 何明春, 董慧, 陈罗敏, 等. 典型喀斯特流域景观生态风险驱动因素及预测: 以打邦河流域为例[J]. 水土保持通报, 2025, 45(5): 218–229.

[27] 陈慧敏, 周冬梅, 张军, 等. 基于景观格局的渭河流域甘肃段生态风险评价[J]. 农业工程, 2022, 12(11): 72–79.

[28] 刘凤莲, 罗芹芹, 杨博文, 等. 基于地理探测器的泾河流域景观生态风险及驱动因素[J]. 干旱区研究, 2025, 42(8): 1488–1500.

[29] 甘子琪, 石淞, 傅婷, 等. 基于XGBoost–SHAP的松花江流域三生空间生态风险耦合协调及驱动机制分析[J]. 环境科学, 2026, 47(7): 4785–4797.

[30] 何刚, 符艺文, 张世玉. 基于生态系统服务价值的淮河流域生态补偿分区[J/OL]. 安全与环境学报: 1–16[2026–08–08]. https://doi.org/10.13637/j.issn.1009-6094.2026.0221.

[31] 甘子琪, 董宏艳, 石淞, 等. 基于可解释性机器学习下哈尔滨市景观生态风险时空演变及驱动力[J/OL]. 环境科学: 1–18[2026–08–07]. https://doi.org/10.13227/j.hjkx.202509060.

[32] 毕文焌, 闫庆武, 赵甫心, 等. 基于XGBoost–SHAP方法的天山中段景观生态风险时空演变及驱动因素[J/OL]. 干旱区地理: 1–16[2026–08–07]. https://link.cnki.net/urlid/65.1103.X.20260128.1853.001. Bi Wenjun, Yan Qingwu, Zhao Fuxin, et al. Spatiotemporal evolution and driving factors of landscape ecological risk in the middle section of the Tianshan Mountains based on XGBoost–SHAP method[J/OL]. Arid Land Geography: 1–16[2026–08–07]. https://link.cnki.net/urlid/65.1103.X.20260128.1853.001.

[33] 李新周, 刘晓东, 马柱国. 近百年来全球主要干旱区的干旱化特征分析[J]. 干旱区研究, 2004(2): 97–103.

[34] Wang Yirui, Song Jinxi, Zhang Xuexian, et al. Sustainable development evaluation and its obstacle factors of the Weihe River Basin in Shaanxi Province, China[J]. Frontiers in Earth Science, 2021, 9: 744224.

[35] 马永欢, 樊胜岳, 姜德娟, 等. 我国北方土地荒漠化成因与草业发展研究[J]. 干旱区研究, 2003(3): 217–220.

[36] 王亦波, 韩新宁, 安可, 等. 基于GEE的渭河流域生态质量动态监测及其驱动机制[J]. 草业学报, 2026, 35(5): 1–19.

[37] 姜彬, 尹涵, 李春雨, 等. 黄河支流陕西段渭河流域沿岸农业土壤肥力综合评价[J]. 西北大学学报(自然科学版), 2024, 54(3): 387–397.

[38] 邹佳莹, 王亦波, 韩新宁, 等. 渭河流域植被覆盖时空变化及其驱动因素[J]. 草地学报, 2025, 33(10): 3381–3395.

[39] 刘珊珊, 吴文婕, 菊春燕, 等. 绿洲城市"三生空间"的景观生态风险时空演变格局及其影响因子[J]. 环境污染与防治, 2023, 45(11): 1566–1573.

[40] 张锦明, 丁建丽, 王瑾杰, 等. 基于土地利用的干旱区典型绿洲景观生态风险评价[J]. 环境污染与防治, 2025, 47(6): 98–102.

[41] 李保杰, 杜欣颖, 褚帅. 淮海经济区景观生态风险多情景模拟[J]. 中国环境科学, 2025, 45(4): 2112–2121.

[42] 王敏, 胡守庚, 张绪冰, 等. 干旱区绿洲城镇景观生态风险时空变化分析: 以张掖绿洲乡镇为例[J]. 生态学报, 2022, 42(14): 5812–5824.

[43] 杨馗, 信桂新, 蒋好雨, 等. 基于最佳尺度的景观生态风险时空变化研究: 以重庆市江津区为例[J]. 生态与农村环境学报, 2021, 37(5): 576–586.

[44] 张利平, 曲建光. 松花江流域(哈尔滨段)景观格局粒度效应[J]. 测绘工程, 2021, 30(1): 51–59.

[45] 黄秀雨, 严志强, 胡宝清, 等. 喀斯特地区土地利用及生态系统服务价值空间自相关分析: 以广西百色市为例[J]. 湖北农业科学, 2023, 62(8): 44–53.

[46] 张金懿, 郝晋珉, 王楠, 等. 山东省耕地质量空间自相关分析与耕地保护分区研究[J]. 土壤通报, 2023, 54(4): 757–767.

[47] Chen Tianqi, Guestrin C. XGBoost: a scalable tree boosting system[C]//Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, August 13–17, 2016, San Francisco, USA. New York: ACM, 2016: 785–794.

[48] Yang Chao, Chen Mingyang, Yuan Quan. The application of XGBoost and SHAP to examining the factors in freight truck-related crashes: an exploratory analysis[J]. Accident Analysis & Prevention, 2021, 158: 106219.

[49] Lundberg S M, Lee S I. A unified approach to interpreting model predictions[C]//Proceedings of the 31st International Conference on Neural Information Processing Systems, December 4–9, 2017, Long Beach, USA. Red Hook: Curran Associates Inc., 2017: 4768–4777.

[50] Sun Linlin, Xu Delan, Mu Zhiqiang, et al. Spatiotemporal evolution and driving factors of landscape ecological risk in the Weihe River Basin from 1990 to 2020[J]. Journal of Resources and Ecology, 2024, 15(4): 1004–1014.

[51] Li Qing, Li Wenjing, Wang Sheng, et al. Assessing heterogeneity of trade-offs/synergies and values among ecosystem services in Beijing-Tianjin-Hebei urban agglomeration[J]. Ecological Indicators, 2022, 140: 109026.

[52] Tian Yulu, Xu Dingxue, Song Jiahua, et al. Impacts of land use changes on ecosystem services at different elevations in an ecological function area, northern China[J]. Ecological Indicators, 2022, 140: 109035.

[53] 陈谢扬, 杨雨萌, 朱炳臣, 等. 生态–风险网络耦合下渭河流域生态安全格局构建[J]. 地理学报, 2026, 81(3): 897–914.

[54] 王成, 夏宇, 苏强, 等. 基于土地利用变化的皖江流域景观生态风险评价与模拟预测研究[J]. 长江流域资源与环境, 2026, 35(2): 455–468.

[55] 佘欢, 李鹏, 王孝康, 等. 2000—2020年渭河流域生态环境质量时空变化及其驱动因素[J]. 水土保持通报, 2025, 45(1): 317–326.

[56] 何佳瑛, 蒋晓辉, 雷宇昕. 黄土高原生态工程对关键生态系统服务时空变化的影响: 以延河流域为例[J]. 生态学报, 2023, 43(12): 4823–4834.

[57] 王敬儒, 蔺宝钢, 岳邦瑞, 等. 基于生态交错带景观特征分析的秦岭北麓生态修复分区方法研究[J]. 中国园林, 2026, 42(5): 115–122.

Basic Information:

DOI:10.13637/j.issn.1009-6094.2026.0721

China Classification Code:X826;P901

Citation Information:

[1]Meng Xiaolan,Xu Fujun,Zhai Meng ,et al.Spatiotemporal patterns and drivers of landscape ecological risk in the Weihe River Basin based on XGBoost–SHAP[J].Journal of Safety and Environment().DOI:10.13637/j.issn.1009-6094.2026.0721.

Published:  

2026-09-11

Publication Date:  

2026-09-11

Online:  

2026-09-11

Search Advanced Search