Susceptibility assessment of mountain landslide based on feature indicator screening and negative sample optimization
-
摘要:
滑坡易发性评价是开展灾害监测预警工作的基础，如何科学、合理地筛选特征指标并优化评价样本仍是当前棘手且易被忽略的问题。以湖南省龙山县为例，基于斜坡单元提取高程、坡度、坡向等15项特征指标，经主成分分析(principal component analysis, PCA)、相关性分析与共线性诊断筛选优质指标，提出一种优化负样本(optimize negative samples, ONS)的方法构建评价样本，然后使用确定性系数-随机森林(certainty factor-random forest, CF-RF) 模型进行滑坡易发性制图，并根据受试者工作特征(receiver operating characteristic, ROC)曲线与合理性分析检验预测结果准确性。研究表明: ONS-CF-RF模型能显著提升模型评价精度，其受试者ROC曲线下面积(area under curve, AUC)较CF-RF模型AUC提升了10.64%; 研究区滑坡高、极高易发区较集中于研究区西北角人类聚集区，低易发区分布于受人类活动影响较小的高海拔山区。研究成果可为龙山县滑坡灾害防治提供科学指导，也可为同类型区域滑坡易发性分区提供参考依据。
Abstract:Landslide susceptibility assessment is the basis of disaster monitoring and early warning. How to scientifically and reasonably screen feature indicators and optimize assessment samples is still a difficult and easily ignored problem. The authors took Longshan County of Hunan Province as an example, and screened high-quality indicators by principal component analysis (PCA), correlation analysis and collinearity diagnosis, based on 15 feature indicators such as elevation, slope and slope direction. An optimize negative samples (ONS) method was proposed to construct assessment samples, and then certainty factor-random forest (CF-RF) model was used to map landslide susceptibility. The accuracy of prediction results was tested according to receiver operating characteristic (ROC) curve and rationality analysis. The results show that ONS-CF-RF model can significantly enhance the accuracy of model assessment. The area under curve (AUC) of this model has increased by 10.64% compared to the AUC of CF-RF model. The high landslide prone area is concentrated in human gathering area in the northwest corner of the study area, while the low landslide prone area is distributed in the high-altitude mountainous area which is less affected by human activities. The research results could provide scientific guidance for the prevention and control of landslide disaster in Longshan County, and could also provide references for the landslide prone zoning in the same type of region.
-
-
表 1 资料数据详细来源
Table 1. Detailed sources of data
序号 数据名称 来源 类型 备注 1 历史滑坡灾害点 湘西州1∶5万地质灾害详查 Excel表格 截至2017年，历史滑坡和潜在滑坡共计221处 2 地质图 湖南省自然资源事务中心 jpg图片 纸质扫描存档，比例尺为1∶200 000 3 行政区划 http://datav.aliyun.com 矢量 更新于2021年5月 4 地层年代与断层 http://dcc.ngac.org.cn 矢量 5 河网与路网 https://www.resdc.cn 矢量 6 土地利用 https://gtdc.mnr.gov.cn 栅格 湖南省土地利用图，比例尺为1∶1 200 000 7 DEM 91卫图 栅格 ALOS DEM，分辨率为12.5 m 
下载: 导出CSV
表 2 成分信息量分析
Table 2. Ingredient information analysis
主成分 特征根 特征根 方差解释率/% 累积方差解释率/% PC01 4.216 46.848 46.848 PC02 1.393 15.475 62.323 PC03 1.085 12.055 74.379 PC04 0.979 10.879 85.258 PC05 0.711 7.904 93.161 PC06 0.368 4.091 97.252 PC07 0.182 2.021 99.273 PC08 0.046 0.516 99.789 PC09 0.019 0.211 100
下载: 导出CSV
表 3 各指标VIF值
Table 3. VIF value for each indicator
地质时期 森林类型 距主道路距离 建设用地占比 距地表水源距离 距断层距离 主成分PC01 主成分PC04 地质时期 1.066 1.068 1.071 1.051 1.068 1.068 1.055 森林类型 1.070 1.075 1.070 1.072 1.073 1.027 1.072 距主道路距离 1.154 1.157 1.158 1.109 1.158 1.084 1.138 建设用地占比 1.005 1.000 1.005 1.005 1.005 1.005 1.005 距地表水源距离 1.133 1.151 1.106 1.155 1.155 1.151 1.109 距断层距离 1.014 1.016 1.018 1.018 1.018 1.014 1.011 PC01 1.163 1.114 1.091 1.166 1.162 1.162 1.129 PC04 1.134 1.148 1.132 1.152 1.107 1.145 1.115
下载: 导出CSV
表 4 研究区特征指标分级信息
Table 4. Classification information of feature indicator in the study area
类型 评价指标 分区标准 CF值 类型 评价指标 分区标准 CF值 离散型 森林类型 非林地 -0.567 1 连续型 距断层距离 ＜300 m 0.400 4 竹林地 0.408 3 [300, 600) m 0.159 2 灌木林地 -0.038 2 [600, 900) m -0.017 4 乔木林地 0.013 7 [900, 1 200) m -0.475 1 其他林地 -0.105 8 [1 200, 1 500) m -0.551 6 地层与地貌 奥陶系 0.050 2 [1 500, 1 800) m -0.136 1 志留系 0.378 7 [1 800, 2 100) m -0.440 2 泥盆系 -1.000 0 ≥2 100 m 0.075 8 二叠系 -0.792 5 PC01 ＜-26 -0.087 5 三叠系 -0.767 6 [-26, -24) -0.120 1 第四纪 0.859 5 [-24, -22) 0.374 6 白垩系 0.664 8 [-22, -20) 0.126 5 寒武系 -0.532 8 [-20, -18) 0.253 8 河流积水区 0.662 9 [-18, -16) 0.195 9 连续型 距主道路距离 ＜250 m 0.525 2 [-16, -14) 0.167 7 [250, 750) m -0.070 5 [-14, -12) -0.140 0 [750, 1 250) m 0.103 1 [-12, -10) -0.398 7 [1 250, 1 750) m -0.292 3 [-10, -8) -0.610 1 [1 750, 2 250) m -0.331 2 [-8, -6) -0.749 5 [2 250, 2 750) m -0.682 4 ≥-6 -0.716 0 ≥2 750 m -0.311 6 PC04 ＜-40 -0.655 7 建设用地占比 ＜2% -0.754 5 [-40, -20) -0.511 5 [2, 12)% 0.689 8 [-20, 0) 0.008 7 [12, 22)% 0.851 2 [0, 20) 0.123 6 [22, 32)% 0.874 7 [20, 40) 0.204 2 [32, 42)% 0.968 2 [40, 60) -0.425 6 [42, 52)% 0.743 3 [60, 80) 0.278 9 ≥52% 0.489 5 [80, 100) 0.005 5 距地表水源距离 ＜200 m 0.527 7 [100, 120) 0.214 1 [200, 450) m -0.015 2 [120, 140) -0.167 2 [450, 700) m 0.089 3 [140, 160) 0.275 7 [700, 950) m -0.086 2 [160, 180) -0.213 2 [950, 1 200) m -0.313 3 [180, 200) 0.585 0 [1 200, 1 450) m -0.601 8 ≥200 0.078 6 [1 450, 1 700) m -0.815 2 ≥1 700 m -0.877 3
下载: 导出CSV
表 5 研究区模型结果对比分析
Table 5. Comparative analysis of the model results in the study area
分区 CF-RF模型 区域面积/km2 滑坡数量/个 滑坡面积/104 m2 FR值 低易发区 751.348 3 2 1.664 0 0.037 8 中易发区 1 000.134 1 17 38.123 0 0.241 5 高易发区 837.074 2 43 22.800 5 0.730 0 极高易发区 495.120 9 155 143.907 7 4.448 7 低易发区 1 086.163 4 8 8.619 0 0.104 7 中易发区 1 060.840 2 23 37.142 0 0.308 1 高易发区 491.194 4 37 21.626 5 1.070 4 极高易发区 445.479 5 149 139.107 7 4.753 0
下载: 导出CSV
表 6 ONS-CF-RF模型统计分析
Table 6. Statistical analysis of ONS-CF-RF model
分区 分区面积/km2 分区面积占比/% 滑坡数量/个 滑坡密度/(个· 10 000 m-2) 低易发区 1 168.1 37.88 8 0.68 中易发区 1 057.3 34.29 23 2.18 高易发区 467.2 15.15 37 7.92 极高易发区 391.0 12.68 149 38.10
下载: 导出CSV
-
[1] 国家统计局. 中国统计年鉴2023[M]. 北京: 中国统计出版社, 2024: 257.
National Bureau of Statistics of China. China Statistical Yearbook 2023[M]. Beijing: China Statistics Press, 2024: 257.
[2] 许强, 郭晨, 董秀军. 地质灾害航空遥感技术应用现状及展望[J]. 测绘学报, 2022, 51(10): 2020-2033.
Xu Q, Guo C, Dong X J. Application status and prospect of aerial remote sensing technology for geohazards[J]. Acta Geodaetica et Cartographica Sinica, 2022, 51(10): 2020-2033.
[3] 李正, 冷亮, 孙永鑫, 等. 基于信息量-机器学习耦合模型的水电梯级开发流域滑坡易发性评价[J]. 测绘通报, 2024(S1): 237-241.
Li Z, Leng L, Sun Y X, et al. Landslide susceptibility assessment in the river cascade development basin based on the Ⅳ-LM coupling model[J]. Bulletin of Surveying and Mapping, 2024(S1): 237-241.
[4] 安雪莲, 密长林, 孙德亮, 等. 基于不同评价单元的三峡库区滑坡易发性对比--以重庆市云阳县为例[J]. 吉林大学学报: 地球科学版, 2024, 54(5): 1629-1644.
An X L, Mi C L, Sun D L, et al. Comparison of landslide susceptibility in three gorges reservoir area based on different evaluation units: Take Yunyang County in Chongqing as an example[J]. Journal of Jilin University (Earth Science Edition), 2024, 54(5): 1629-1644.
[5] 唐亚明, 薛强, 毕俊擘, 等. 陕北黄土崩塌灾害风险评价指标体系构建[J]. 地质通报, 2012, 31(6): 979-988.
Tang Y M, Xue Q, Bi J B, et al. The construction of factors for assessing the risk of collapse at loess slopes in northern Shaanxi Province[J]. Geological Bulletin of China, 2012, 31(6): 979-988.
[6] Jaafari A. Landslide susceptibility assessment using novel hybridized methods based on the support vector regression[J]. Ecological Engineering, 2024, 208: 107372. doi: 10.1016/j.ecoleng.2024.107372
[7] Ciurleo M, Cascini L, Calvello M. A comparison of statistical and deterministic methods for shallow landslide susceptibility zoning in clayey soils[J]. Engineering Geology, 2017, 223: 71-81. doi: 10.1016/j.enggeo.2017.04.023
[8] Hussin H Y, Zumpano V, Reichenbach P, et al. Different landslide sampling strategies in a grid-based bi-variate statistical susceptibility model[J]. Geomorphology, 2016, 253: 508-523. doi: 10.1016/j.geomorph.2015.10.030
[9] S ˇ ilhán K. A new tree-ring-based index for the expression of spatial landslide activity and the assessment of landslide hazards[J]. Geomatics, Natural Hazards and Risk, 2021, 12(1): 3409-3428. doi: 10.1080/19475705.2021.2011790
[10] 洪浩源, 王德生, 朱阿兴. 面向机器学习型区域滑坡易发性评价的训练样本采样方法[J]. 地理学报, 2024, 79(7): 1718-1736.
Hong H Y, Wang D S, Zhu A X. A new training data sampling method for machine learning-based landslide susceptibility mapping[J]. Acta Geographica Sinica, 2024, 79(7): 1718-1736.
[11] 郭飞, 赖鹏, 黄发明, 等. 基于知识图谱的滑坡易发性评价文献综述及研究进展[J]. 地球科学, 2024, 49(5): 1584-1606.
Guo F, Lai P, Huang F M, et al. Literature review and research progress of landslide susceptibility mapping based on knowledge graph[J]. Earth Science, 2024, 49(5): 1584-1606.
[12] Aleotti P, Chowdhury R. Landslide hazard assessment: Summary review and new perspectives[J]. Bulletin of Engineering Geology and the Environment, 1999, 58(1): 21-44. doi: 10.1007/s100640050066
[13] 罗路广, 裴向军, 崔圣华, 等. 九寨沟地震滑坡易发性评价因子组合选取研究[J]. 岩石力学与工程学报, 2021, 40(11): 2306-2319.
Luo L G, Pei X J, Cui S H, et al. Combined selection of susceptibility assessment factors for Jiuzhaigou earthquake-induced landslides[J]. Chinese Journal of Rock Mechanics and Engineering, 2021, 40(11): 2306-2319.
[14] Xia D, Tang H M, Glade T, et al. KNN-GCN: a deep learning approach for slope-unit-based landslide susceptibility mapping incorporating spatial correlations[J]. Mathematical Geosciences, 2024, 56(5): 1011-1039. doi: 10.1007/s11004-023-10132-3
[15] 李霞, 宿星, 张满银, 等. 基于证据权法与多源数据的陇中生态脆弱区滑坡敏感性评价--以天水市为例[J]. 冰川冻土, 2023, 45(1): 67-79.
Li X, Su X, Zhang M Y, et al. Landslide susceptibility using weights-of-evidence approach and multi-source data in Longzhong ecologically vulnerable area: A case study of Tianshui City[J]. Journal of Glaciology and Geocryology, 2023, 45(1): 67-79.
[16] 姚翔曦, 张英, 张国治, 等. 基于数据扩充和故障特征优化的SCNGO-SVM-AdaBoost变压器故障诊断技术[J]. 南方电网技术, 2025, 19(6): 14-25.
Yao X X, Zhang Y, Zhang G Z, et al. SCNGO-SVM-AdaBoost transformer fault diagnosis technology based on data augmentation and fault feature optimization[J]. Southern Power System Technology, 2025, 19(6): 14-25.
[17] 夏辉, 殷坤龙, 梁鑫, 等. 基于SVM-ANN模型的滑坡易发性评价--以三峡库区巫山县为例[J]. 中国地质灾害与防治学报, 2018, 29(5): 13-19.
Xia H, Yin K L, Liang X, et al. Landslide susceptibility assessment based on SVM-ANN models: A case stualy for Wushan County in the Three Gorges Reservoir[J]. The Chinese Journal of Geological Hazard and Control, 2018, 29(5): 13-19.
[18] 田尤, 张佳佳, 殷红, 等. 基于多种数理模型及其融合的察雅县城滑坡易发性评价[J]. 科学技术与工程, 2024, 24(29): 12452-12460.
Tian Y, Zhang J J, Yin H, et al. Landslide susceptibility evaluation in Chaya County based on multiple mathematical models and their fusion[J]. Science Technology and Engineering, 2024, 24(29): 12452-12460.
[19] 王守华, 王睿菘, 孙希延, 等. 基于CF-CNN-LSTM模型的滑坡易发性评价[J]. 自然灾害学报, 2024, 33(5): 84-95.
Wang S H, Wang R S, Sun X Y, et al. Landslide susceptibility evaluation based on CF-CNN-LSTM model[J]. Journal of Natural Disasters, 2024, 33(5): 84-95.
[20] 葛睿雅, 李晓晖, 袁峰, 等. 基于大数据和熵权-随机森林的城市地下空间需求评价[J]. 合肥工业大学学报: 自然科学版, 2025, 48(3): 360-368.
Ge R Y, Li X H, Yuan F, et al. Urban underground space demand evaluation based on big data and entropy-random forest[J]. Journal of Hefei University of Technology (Natural Science), 2025, 48(3): 360-368.
[21] Yuan X Y, Liu C, Nie R H, et al. A comparative analysis of certainty factor-based machine learning methods for collapse and landslide susceptibility mapping in Wenchuan County, China[J]. Remote Sensing, 2022, 14(14): 3259.
[22] 王伟, 宋月, 黄莉, 等. 粤港澳大湾区复合灾害系统脆弱性评估[J/OL]. 西南交通大学学报, (2024-09-03).
http://kns.cnki.net/kcms/detail/51.1277.u.20240903.0915.002.html .Wang W, Song Y, Huang L, et al. Vulnerability assessment of composite disaster systems in the Greater Bay Area[J/OL]. Journal of Southwest Jiaotong University, (2024-09-03).
http://kns.cnki.net/kcms/detail/51.1277.u.20240903.0915.002.html .[23] 自然资源部中国地质调查局. DD 2019-08地质灾害调查技术要求(1: 50 000)[S]. 北京: 中国地质调查局, 2019.
China Geological Survey. DD 2019-08 Technical Requirement for Geo-hazard Survey (1: 50 000)[S]. Beijing: China Geological Survey, 2019.
[24] 何哲, 石玉玲, 李富春, 等. 基于LightGBM模型的甘肃省临夏县滑坡易发性评价[J]. 水资源与水工程学报, 2024, 35(1): 197-205, 216.
He Z, Shi Y L, Li F C, et al. Landslide susceptibility assessment based on LightGBM model in Linxia County, Gansu Province[J]. Journal of Water Resources&Water Engineering, 2024, 35(1): 197-205, 216.
[25] 夏洋德龙, 罗伟奇, 高俊华, 等. 基于逻辑回归的择优耦合模型滑坡易发性评估[J]. 地理空间信息, 2025, 23(6): 61-64, 78.
Xia Y D L, Luo W Q, Gao J H, et al. Preferred coupled model landslide susceptibility evaluation based on logistic regression[J]. Geospatial Information, 2025, 23(6): 61-64, 78.
[26] 陈建平, 辛亚波, 王泽鹏, 等. 样本选取对地质灾害易发性评价的影响--以山西柳林县为例[J]. 中国地质灾害与防治学报, 2024, 35(3): 152-162.
Chen J P, Xin Y B, Wang Z P, et al. Effect of sample selection on the susceptibility assessment of geological hazards: A case study in Liulin County, Shanxi Province[J]. The Chinese Journal of Geological Hazard and Control, 2024, 35(3): 152-162.
[27] 黄永芳, 郭永刚, 黄艳婷. 基于加权信息量和加权确定性系数的藏东南滑坡易发性评价[J]. 中国地质调查, 2024, 11(3): 108-116. doi: 10.19388/j.zgdzdc.2024.03.14
Huang Y F, Guo Y G, Huang Y T. Landslide susceptibility assessment in the southeastern Tibet based on weighted informativeness and weighted certainty factor[J]. Geological Survey of China, 2024, 11(3): 108-116. doi: 10.19388/j.zgdzdc.2024.03.14
-
图(9)
表(6)
计量
- 文章访问数: 35
- PDF下载数: 5
- 施引文献: 0