Study of Landslide Susceptibility Mapping Based on Fractal Dimension Integrating Support Vector Machine with Index of Entropy Model
-
摘要:
陕西省宝鸡市北部黄土高原滑坡灾害频发，严重威胁当地人民的经济发展和生产生活。本研究基于分形维数，分别利用熵权模型（IOE）、支持向量机模型（SVM）和两种混合模型即F-IOE和F-SVM对滑坡可能发生的范围进行定量预测。首先，利用179个滑坡样本制作滑坡编录图，将70%（125个）的滑坡样本用于训练，其余30%（54个）用于测试。随后，提取12种滑坡影响因子，分别计算每个因子的信息增益率和分形维数，并使用训练数据建立4种滑坡易发性分区模型。最后，利用受试者工作特征曲线（ROC）和统计学指标包括阳性预测率（PPR）、阴性预测率（NPR）和准确率（ACC）测试模型的性能，并比较模型的泛化性。结果表明，F-SVM模型在训练和测试数据集上分别得到最高的PPR、NPR、ACC和AUC值，其次是F-IOE模型。最终，F-SVM模型在所有模型中表现最优，因此，基于分形维数构建的混合模型比原始模型更具优势，可为当地滑坡防治决策提供参考。
Abstract:Landslides occur frequently on the Loess Plateau in the north of Baoji City, Shaanxi Province, which seriously threaten the economic development, production and life of the local people. Based on fractal dimension, entropy weight model (IOE), support vector machine model (SVM) and two hybrid models, namely F-IOE and F-SVM, are used to quantitatively predict the possible occurrence range of landslide. First of all, 179 landslide samples were used to make landslide cataloguing maps, 70% (125) of the landslide samples were used for training, and the remaining 30% (54) were used for testing. Then, 12 kinds of landslide influence factors are extracted, information gain rate and fractal dimension of each factor are calculated respectively, and four landslide vulnerability zoning models are established using training data. Finally, the performance of the model was tested using the receiver operating characteristic curve (ROC) and statistical indicators including positive predictive rate (PPR), negative predictive rate (NPR) and accuracy rate (ACC), and the generalization of the model was compared. The results show that F-SVM model has the highest PPR, NPR, ACC and AUC values in training and test data sets respectively, followed by F-IOE model. Finally, F-SVM model is the best among all models. Therefore, the hybrid model based on fractal dimension has more advantages than the original model, which can provide reference for local landslide control decisions.
-
Key words:
- GIS /
- research on landslide susceptibility /
- mixed model /
- fractal dimension
-
-
表 1 研究区地层岩性单元表
Table 1. Lithological units of study area
类别 地质年代 编码 主要地层岩性 A 新进纪 Q4 砂，砾石，黄土 更新纪 Q3 黄土，砾石 B 上新统 N2 砂土 中新统 N1 石英砂，黏土 C 晚白垩纪 K1 泥岩，砂质泥岩，泥质砂岩 D 早侏罗纪 J3 块状聚集物，粘胶岩，粉砂质泥岩 中侏罗纪 J2 石质，泥岩，粉质泥岩 晚侏罗纪 J1 粉砂岩，煤层 E 早三叠纪 T3 泥岩，页岩，煤层 中三叠纪 T2 中细砂岩，粉砂岩，泥岩 晚三叠纪 T1 石质，细砂岩，粉砂岩，砂质泥岩 F 二叠纪 P 砂质泥岩，细砂岩，粉砂岩 
下载: 导出CSV
表 2 影响因子的VIF和TOL表
Table 2. Variance inflation factors (VIF) and tolerances of each conditioning factor
影响因子 TOL VIF 坡度 0.934 1.071 坡向 0.926 1.080 高程 0.656 1.525 距河流的距离 0.908 1.101 距道路的距离 0.877 1.141 距断层的距离 0.916 1.092 NDVI 0.597 1.675 土地利用类型 0.650 1.538 地层岩性 0.814 1.228 降雨量 0.814 1.229 平面曲率 0.912 1.096 剖面曲率 0.925 1.082
下载: 导出CSV
表 3 影响因子与滑坡的空间关系表
Table 3. Spatial relationship between influencing factors and landslides
影响因子 等级 fj FRij Pij Hj Hjmax Ij Wj F-Wj 坡度（°） ＜5 0.0772 0.3035 0.0367 2.2857 2.5850 0.1158 0.1595 0.1201 5~10 0.1906 0.8709 0.1053 10~15 0.2285 0.9124 0.1104 15~20 0.2372 1.1518 0.1393 20~25 0.1682 1.9427 0.2350 ＞25 0.2284 3.0864 0.3733 坡向 水平 0.0000 0.0000 0.0000 2.9004 3.1699 0.0850 0.0732 0.0946 北 0.0459 0.6097 0.0787 东北 0.0317 0.3564 0.0460 东 0.2239 0.9802 0.1266 东南 0.1567 1.0364 0.1338 南 0.1429 1.1387 0.1470 西南 0.1990 1.0215 0.1319 西 0.2260 1.6248 0.2098 西北 0.1251 0.9771 0.1262 高程（m） ＜850 0.5803 2.9085 0.4024 2.4666 2.8074 0.1214 0.1253 0.1279 850~950 0.3346 1.1898 0.1646 950~ 1050 0.2131 0.7202 0.0996 1050 ~1150 0.1009 0.5549 0.0768 1150 ~1250 0.0942 0.5738 0.0794 1250 ~1350 0.1606 0.8630 0.1194 ＞ 1350 0.2092 0.4174 0.0578 距河流的距离（m） ＜200 0.3998 2.2738 0.4703 2.0154 2.3219 0.1320 0.1277 0.1009 200~400 0.2220 0.8336 0.1724 400~600 0.0539 0.4464 0.0923 600~800 0.0706 0.8641 0.1787 ＞800 0.1942 0.4171 0.0863 距断层的距离（m） ＜2000 0.4358 1.2216 0.2275 2.2883 2.3219 0.0145 0.0155 0.1263 2000 ~4000 0.2725 1.2343 0.2299 4000 ~6000 0.3286 1.1496 0.2141 6000 ~8000 0.3992 1.1191 0.2084 ＞ 8000 0.2951 0.6445 0.1200 距道路的距离（m） ＜100 0.4866 1.1476 0.5302 1.3723 2.0000 0.3139 0.1698 0.1030 100~200 0.1407 0.7709 0.3562 200~300 0.1334 0.2458 0.1130 ＞300 0.0000 0.0000 0.0000
下载: 导出CSV
表 4 模型性能评价统计指标计算结果表
Table 4. Calculation result of statistical indicators for model performance evaluation
指标 模型 IOE SVM F-IOE F-SVM 真阳性 108 113 110 118 真阴性 104 110 109 110 假阳性 21 15 16 15 假阴性 17 12 15 7 PPR（%） 83.72 88.28 87.30 88.72 NPR（%） 85.95 90.16 87.90 94.02 ACC（%） 84.80 89.00 87.60 91.20
下载: 导出CSV
表 5 模型测试评价统计指标计算结果表
Table 5. Calculation result of statistical indicators for model validation evaluation
指标 模型 IOE SVM F-IOE F-SVM 真阳性 48 52 110 118 真阴性 47 48 109 110 假阳性 7 6 16 15 假阴性 6 2 15 7 PPR（%） 87.27 89.66 86.21 92.73 NPR（%） 88.68 96.00 92.00 94.34 ACC（%） 87.96 92.59 88.89 93.52
下载: 导出CSV
-
[1] 韩玲, 张庭瑜, 张恒. 基于IOE和SVM模型的府谷镇滑坡易发性分区[J]. 水土保持研究, 2019, 26: 373-378
HAN Ling, ZHANG Tingyu, ZHANG Heng. Landslide susceptibility mapping based on IOE and SVM model in Fugu town[J]. Research of Soil and Water Conservation, 2019, 26: 373-378.
[2] 刘光旭, 席建超, 戴尔阜, 等. 中国滑坡灾害承灾体损失风险定量评估[J]. 自然灾害学报, 2014: 8
LIU Guangxu, XI Jianchao, DAI Erfu, et al. Loss risk assessment of the hazard-affectted body of landslides in China[J]. Journal of Natural Disasters, 2014: 8.
[3] 孙萍萍, 张茂省, 贾俊, 等. 中国西部黄土区地质灾害调查研究进展[J]. 西北地质, 2022, 55(03): 96-107
SUN Pingping, ZHANG Maoxing, JIA Jun, et al. Progress in Geological Hazard Investigation and Research in Loess Regions of Western China [J]. Northwestern Geology, 2022, 55(03): 96-107.
[4] 陕西省地质矿产局. 陕西省区域地质志 [M]. 西安: 地质出版社, 1989. Bureau of Geology and Mineral Resources of Shaanxi Province, 1989. Regional geology of Shaanxi Province[M]. Xi'an: Geological Publishing House,1989.
[5] 张茂省, 薛强, 贾俊, 等. 山区城镇地质灾害调查与风险评价方法及实践[J]. 西北地质, 2019, 52(02): 125-135
ZHANG Maoxing, XUE Qiang, JIA Jun, et al. Method And Practice of Geological Hazard Investigation and Risk Assessment in Mountainous Towns[J]. Northwestern Geology, 2019, 52(02): 125-135.
[6] 贾俊, 李志忠, 郭小鹏, 等. 多源遥感技术在降雨诱发勉县地质灾害调查中的应用[J]. 西北地质, 2023, 56(03): 268-280
JIA Jun, LI Zhizhong, GUO Xiaopeng, et al. Application of Multi-Source Remote Sensing Technology in the Investigation of Geological Disasters Induced by Rainfall in Mianxian County [J]. Northwestern Geology, 2023, 56(03): 268-280.
[7] 武雪玲, 沈少青, 牛瑞卿. GIS支持下应用PSO-SVM模型预测滑坡易发性[J]. 武汉大学学报(信息科学版), 2016, 41: 665-671
WU Xueling, SHEN Shaoqing, NIU Ruiqin. PSO-SVM model is used to predict landslide susceptibility with GIS support [J]. Geomatics and Information Science of Wuhan University, 2016, 41: 665-671.
[8] 谢振华, 窦培谦. 基于BP神经网络的矿山排土场滑坡预警模型[J]. 金属矿山, 2017, （6）: 166-169 doi: 10.3969/j.issn.1001-1250.2017.09.034
XIE Zhenhua, DOU Peiqian. Landslide early warning model for mine dumping site based on BP neural network[J]. Metal Mine, 2017,（6）: 166-169. doi: 10.3969/j.issn.1001-1250.2017.09.034
[9] 杨光, 徐佩华, 曹琛等. 基于确定性系数组合模型的区域滑坡敏感性评价[J]. 工程地质学报, 2019, 27: 1153-1163
YANG Guang, XU Peihua, CAO Chen et. al, Assessment of regional landslide susceptibility based on combined model of certainty factor method [J]. Journal of Engineering Geology, 2019, 27: 1153-1163.
[10] 张文龙, 张振凯, 杨帅. 勉略宁地区地质灾害危险性智能评价和区划研究[J]. 西北地质, 2023, 56(01): 276-282
ZHANG Wenlong, ZHANG Zhenkai, YANG Shuai. Research on Intelligent Evaluation and Zoning of Geological Hazard Risk in Mian-lv-ning Area[J]. Northwestern Geology, 2023, 56(01): 276-282.
[11] 张庭瑜, 韩玲, 张恒等. 混合分类模型在滑坡易发性分区中的适用性研究——以延安市宝塔区为例[J]. 干旱区资源与环境, 2020, 22: 192-201
ZHANG Tingyu, HAN Ling, ZHANG heng, et. al, Feasibility of hybrid classification models for landslide susceptibility mapping-A case of Baota district, Yan’an City[J]. Journal of Arid Land Resources and Environment, 2020, 22: 192-201.
[12] 张向营, 张春山, 孟华君等. 基于Random Forest和AHP的贵德县北部山区滑坡危险性评价[J]. 水文地质工程地质, 2018, 45: 142-149
ZHANG Xiangying, ZHANG Chunshan, MENG Huajun, et. al. Landslide hazard evaluation in the northern mountainous area of Guide County based on Random Forest and AHP [J]. Hydrogeology & Engineering Geology, 2018, 45: 142-149.
[13] Abedini M, Tulabi S. Assessing LNRF, FR, and AHP models in landslide susceptibility mapping index: a comparative study of Nojian watershed in Lorestan province, Iran [J]. Environmental Earth Sciences, 2018, 77: 405-405. doi: 10.1007/s12665-018-7524-1
[14] Arabameri A, Pradhan B, Rezaei K, et al. GIS-based landslide susceptibility mapping using numerical risk factor bivariate model and its ensemble with linear multivariate regression and boosted regression tree algorithms [J]. Journal of Mountain Science, 2018, 16: 595-618.
[15] Balogun A-L, Rezaie F, Pham Q B, et al. Spatial prediction of landslide susceptibility in western Serbia using hybrid support vector regression (SVR) with with GWO, BAT and COA algorithms[J]. Geoscience Frontiers , 2021, 12: 101-104.
[16] Bazi Y, Bashmal L, Rahhal M, et al. Vision Transformers for Remote Sensing Image Classification [J]. Remote Sensing, 2021, 13: 516. doi: 10.3390/rs13030516
[17] Behnia P, Blais-stevens A. Landslide susceptibility modelling using the quantitative random forest method along the northern portion of the Yukon Alaska Highway Corridor [J], Canada. Natural Hazards, 2018, 90: 1407–1426. doi: 10.1007/s11069-017-3104-z
[18] Bouboulis P, Dalla L, Drakopoulos V. Construction of recurrent bivariate fractal interpolation surfaces and computation of their box-counting dimension[J]. Journal of Approximation Theory , 2006, 141: 99-117. doi: 10.1016/j.jat.2006.01.006
[19] Bui D T, Tuan T A, Hoang N D, et al. Spatial prediction of rainfall-induced landslides for the Lao Cai area (Vietnam) using a hybrid intelligent approach of least squares support vector machines inference model and artificial bee colony optimization[J]. Landslides , 2017, 14: 447-458.
[20] Chen W, Li Y, Tsangaratos P, et al. Groundwater Spring Potential Mapping Using Artificial Intelligence Approach Based on Kernel Logistic Regression, Random Forest, and Alternating Decision Tree Models [J]. Applied Sciences, 2020, 10: 425-441. doi: 10.3390/app10020425
[21] Chen W, Xie X, Wang J, et al. A comparative study of logistic model tree, random forest, and classification and regression tree models for spatial prediction of landslide susceptibility [J]. Catena, 2017, 151: 147-160. doi: 10.1016/j.catena.2016.11.032
[22] Hong H, Liu J, Zhu A-X, et al. A novel hybrid integration model using support vector machines and random subspace for weather-triggered landslide susceptibility assessment in the Wuning area (China) [J]. Environmental Earth Sciences, 2017, 76: 652-667. doi: 10.1007/s12665-017-6981-2
[23] Islam M Z, Giggins H . Knowledge Discovery through SysFor - a Systematically Developed Forest of Multiple Decision Trees [C], Proceedings of the 9-th Australasian Data Mining Conference. December 15-16; Ballarat, Australia, 2011.
[24] Jaafari A, Najafi A, Pourghasemi H R, et al. GIS-based frequency ratio and index of entropy models for landslide susceptibility assessment in the Caspian forest, northern Iran[J]. International Journal of Environmental Science & Technology , 2014, 11: 909-926.
[25] Kim J, Lee S, Jung H. Landslide susceptibility mapping using random forest and boosted tree models in Pyeong-Chang, Korea[J]. Geocarto International , 2018, 45: 1000-1015.
[26] Kumar D, Thakur M, S. Dubey C, et al. Landslide susceptibility mapping & prediction using Support Vector Machine for Mandakini River Basin, Garhwal Himalaya, India [J]. Geomorphology, 2017, 295: 115-125. doi: 10.1016/j.geomorph.2017.06.013
[27] Li J, Zhang Y. GIS-supported certainty factor (CF) models for assessment of geothermal potential: A case study of Tengchong County, southwest China [J]. Energy, 2017, 140: 552-565. doi: 10.1016/j.energy.2017.09.012
[28] Li R, Wang N. Landslide Susceptibility Mapping for the Muchuan County (China): A Comparison Between Bivariate Statistical Models (WoE, EBF, and IoE) and Their Ensembles with Logistic Regression [J]. Symmetry, 2019, 11: 762-781. doi: 10.3390/sym11060762
[29] Lombardo L, Tanyas H, Nicu I C. Spatial modeling of multi-hazard threat to cultural heritage sites [J]. Engineering Geology, 2020, 277: 105776. doi: 10.1016/j.enggeo.2020.105776
[30] Mahdadi F, Boumezbeur A, Hadji R, et al. GIS-based landslide susceptibility assessment using statistical models: a case study from Souk Ahras province, N-E Algeria [J]. Arabian Journal of Geosciences, 2018, 11: 476-488. doi: 10.1007/s12517-018-3770-5
[31] Pandey V K, Pourghasemi H R. Landslide susceptibility mapping using maximum entropy and support vector machine models along the highway corridor, Garhwal Himalaya [J]. Geocarto International, 2020, 35: 168-187. doi: 10.1080/10106049.2018.1510038
[32] Pham Q B, Yacine A, Ali S A, et al. A comparison among fuzzy multi-criteria decision making, bivariate, multivariate and machine learning models in landslide susceptibility mapping[J]. Geomatics, Natural Hazards and Risk , 2021, 12: 1741-1777. doi: 10.1080/19475705.2021.1944330
[33] Sameen M I, Pradhan B, Lee S. Application of convolutional neural networks featuring Bayesian optimization for landslide susceptibility assessment [J]. Catena, 2020, 186: 104249. doi: 10.1016/j.catena.2019.104249
[34] Shuren W, Han J, Shi J, et al. A Study of the Relationship between Landslide Stability and Fractal Dimension of the Major Landslide Trace Pattern near Badong in the Reservoir Region of the Yangtze Gorge Project [J]. Acta Geosicientia Sinica, 2005, 26: 455-460.
[35] Thai P B, Indra P. Machine Learning Methods of Kernel Logistic Regression and Classification and Regression Trees for Landslide Susceptibility Assessment at Part of Himalayan Area, India [J]. Indian Journal of Science & Technology, 2018, 11: 1-10.
[36] Wang Z, Brenning A. Active-Learning Approaches for Landslide Mapping Using Support Vector Machines [J]. Remote Sensing, 2021, 13: 2588-2607. doi: 10.3390/rs13132588
[37] Youssef A M, Pradhan B, Dikshit A, et al. Landslide susceptibility mapping using CNN-1D and 2D deep learning algorithms: comparison of their performance at Asir Region, KSA[J]. Bulletin of engineering geology and the environment , 2022, 81: 110-135. doi: 10.1007/s10064-022-02609-y
-
图(7)
表(5)
计量
- 文章访问数: 218
- PDF下载数: 15
- 施引文献: 0