顾及InSAR形变的CNN滑坡易发性动态评估——以刘家峡水库区域为例

doi:10.13722/j.cnki.jrme.2022.0266

摘要
图/表
参考文献
相关文章 (15)

全文: PDF (23215 KB) HTML (1 KB)
输出: BibTeX | EndNote (RIS)

摘要已有的滑坡易发性评估方法所采用的滑坡因子多为静态数据(如地形，地质)，缺乏动态数据(如地表形变)，无法充分挖掘正在变形的滑坡特征，导致滑坡易发性评估可靠性较差。合成孔径雷达干涉测量(synthetic aperture radar interferometry，InSAR)二维InSAR数据可反映滑坡在垂直和水平方向的形变特征。引入二维InSAR形变数据作为动态因子，结合地形、地质、水文以及人文共计16种滑坡影响因子，构建卷积神经网络模型(convolutional neural network，CNN)进行滑坡易发性动态评估；采用多种评价指标来衡量模型精度，同时对比有无顾及二维InSAR因子的滑坡易发性评估结果，并与支持向量机(support vector machines，SVM)方法进行对比，此外，在不同场景进行应用。实验结果表明，顾及InSAR形变动态因子的多种评价指标整体精度有所提高，模型对正处于缓慢变形的滑坡易发性区域的识别具有良好的效果，在训练数据中，识别率由0.78提高至0.93；在验证数据中，识别率提升0.21，揭示出顾及二维InSAR形变因子可以提升滑坡易发性动态评估准确度。构建的CNN模型相比SVM方法评估结果更加可靠，在不同场景下更具有普适性和泛化能力。刘家峡水库区域高易发性主要分布于杨塔乡东北部山区、红泉镇以及三塬镇等区域，应加强实时监测。提出的顾及二维InSAR形变因子CNN滑坡易发性动态评估模型可为滑坡灾害预防提供新思路。

	服务

	把本文推荐给朋友
	加入我的书架
	加入引用管理器
	E-mail Alert
	RSS
	作者相关文章

关键词 ：边坡工程, 滑坡易发性评估, 二维时序InSAR, 深度学习, CNN

Abstract：The landslide factors used in the existing landslide susceptibility evaluation methods are mostly static data(eg. topography，geology) and lack of dynamic data(eg. surface deformation)，which cannot fully extract the characteristics of deformation landslides，resulting in poor reliability of landslide susceptibility evaluation. Synthetic aperture radar interferometry(InSAR) data can reflect the deformation characteristics of landslides in both vertical and horizontal directions. In this study，two-dimensional InSAR deformation data are introduced as dynamic factors，topography，geology，hydrology and humanities are combined，and a convolutional neural network(CNN) model is constructed for dynamic evaluation of landslide susceptibility. Multiple evaluation metrics are used to evaluate the model accuracy，while landslide susceptibility evaluation results with and without taking into two-dimensional InSAR factors are compared. The proposed method is compared with support vector machines(SVM) method，in addition，the proposed method is applied in different cases. The results show that the overall accuracy of multiple evaluation indexes considering InSAR deformation dynamic factors has been improved，and the model has shown good results in identifying landslide susceptibility areas that are in slow deformation，and the recognition rate has been improved from 0.78 to 0.93 in the training data. In the validation data，the recognition rate is improved by 0.21，revealing that considering the 2D InSAR deformation factors can improve the accuracy of dynamic evaluation of landslide susceptibility. The CNN model constructed in this study has more reliable evaluation results compared with SVM method，and is more applicable and generalizable in different cases. The high susceptibility areas in the Liujiaxia Reservoir are mainly located in the northeast of Yangta Township，Hongquan Township and Sanluo Township of the mountainous areas，real-time monitoring should be strengthened. The dynamic evaluation model of landslide susceptibility using CNN considering 2D InSAR deformation factors proposed can provide the new ideas for landslide disaster prevention.

Key words： slope engineering landslide susceptibility evaluation 2D time series InSAR deep learning CNN

引用本文:

高秉海1，2，3，何毅1，2，3，4，张立峰1，2，3，姚圣1，2，3，杨旺1，2，3，陈毅1，2，3，何旭1，2，3，赵占骜1，2，3，. 顾及InSAR形变的CNN滑坡易发性动态评估——以刘家峡水库区域为例[J]. 岩石力学与工程学报, 2023, 42(2): 450-465.
GAO Binghai1，2，3，HE Yi1，2，3，4，ZHANG Lifeng1，2，3，YAO Sheng1，2，3，YANG Wang1，2，3， CHEN Yi1，2，3，HE Xu1，2，3，ZHAO Zhan?ao1，2，3，CHEN Hesheng1，2，3. Dynamic evaluation of landslide susceptibility by CNN considering InSAR deformation：A case study of Liujiaxia reservoir . , 2023, 42(2): 450-465.

链接本文:

http://rockmech.whrsm.ac.cn/CN/10.13722/j.cnki.jrme.2022.0266 或 http://rockmech.whrsm.ac.cn/CN/Y2023/V42/I2/450

[1]	ZHOU X，WEN H，ZHANG Y，et al. Landslide susceptibility mapping using hybrid random forest with GeoDetector and RFE for factor optimization[J]. Geoscience Frontiers，2021，12(5)：355–373.
[3]	李文彬，范宣梅，黄发明，等. 不同环境因子联接和预测模型的滑坡易发性建模不确定性[J]. 地球科学，2021，46(10)：3 777–3 795. (LI Wenbin，FAN Xuanmei，HUANG Faming，et al. Uncertainties of landslide susceptibility modeling under different environmental factor connections and prediction models[J]. Earth Science，2021，46(10)：3 777–3 795.(in Chinese))
[4]	CANTARINO I，CARRION M A，GOERLICH F，et al. A ROC analysis-based classification method for landslide susceptibility maps[J]. Landslides，2018，10(2)：265–282.
[9]	王文辉. 联合SBAS-InSAR与机器学习的滑坡隐患识别[硕士学位论文][D]. 兰州：兰州交通大学，2021.(WANG WenHui. Landslide hazard identification based on SBAS-InSAR and machine learning—Taking Lanzhou city as an example[M. S. Thesis][D]. Lanzhou： Lanzhou Jiaotong University，2021.(in Chinese))
[23]	HE A，PAN P，DAI L，et al. Application of kernel-based Fisher discriminant analysis to map landslide susceptibility in the Qinggan River delta，Three Gorges，China[J]. Geomorphology，2012，20(171–172)：30–40.
[25]	黎志恒，张永军，梁收运. 兰州城市地质灾害与人类工程活动[J]. 兰州大学学报：自然科学版，2014，50(5)：588–593.(LI Zhiheng，Zhang Yongjun，LIANG Shouyun，et al. Urban geological hazards and human engineering activities in Lanzhou[J]. Journal of Lanzhou University：Natural Science Edition，2014，50(5)：588–593.(in Chinese))
[27]	李小林，郭小花，李万花. 黄河上游龙羊峡—刘家峡河段巨型滑坡形成机理分析[J]. 工程地质学报，2011，19(4)：516–529.(LI Xiaolin，GUO Xiaohua，LI Wanhua，et al. Analysis of the formation mechanism of giant landslides in the Long Yangxia-Liujiaxia section of the upper Yellow River [J]. Journal of Engineering Geology，2011，19(4)：516–529.(in Chinese))
[28]	SAMEEN M，PRADHAN B，LEE S. Application of convolutional neural networks featuring Bayesian optimization for landslide susceptibility assessment[J]. Catena，2020，186(3)：1–46.
[5]	HE Y，ZHAO Z，YANG W，et al. A unified network of information considering superimposed landslide factors sequence and pixel spatial neighbourhood for landslide susceptibility mapping[J]. International Journal of Applied Earth Observation and Geoinformation，2021，104(18)：1–15.
[13]	陈贵辉. 刘家峡区域经济开发史及特点[J]. 开发研究，2003，(1)：52–56.(CHEN Guihui. History and characteristics of economic development in Liujiaxia region[J]. Development Research，2003，(1)：52–56.(in Chinese))
[21]	CHEN W，POURGHASEMI H R，PANAHI M，et al. Spatial prediction of landslide susceptibility using an adaptive neuro-fuzzy inference system combined with frequency ratio，generalized additive model，and support vector machine techniques[J]. Geomorphology，2017，9(297)：69–85.
[29]	LECUN Y，BOSER B，DENKER J，et al. Backpropagation applied to handwritten zip code recognition[J]. Neural Computation，1989，1(4)：541–551.
[7]	GUZZETTI F，MANUNTA M，ARDIZZONE F，et al. Analysis of ground deformation detected using the SBAS-DInSAR technique in Umbria，Central Italy[J]. Pure and Applied Geophysics，2009，166(8–9)：1 425–1 459.
[11]	ACOSTA G，A RODRíGUEZ，EUILLADES P，et al. Detection of active landslides by dinsar in Andean Precordillera of San Juan，Argentina[J]. Journal of South American Earth Sciences，2021，108(4)：1–15.
[12]	杨旺，何毅，张立峰，等. 甘肃金川矿区地表三维形变InSAR监测[J]. 自然资源遥感，2022，34(1)：177–188.(YANG Wang，HE Yi，ZHANG Lifeng，et al. InSAR monitoring of 3D surface deformation in Jinchuan mining area，Gansu Province[J]. Remote Sensing for Natural Resources，2022，34(1)：177–188.(in Chinese))
[15]	WANG Y，FANG Z，HONG H. Comparison of convolutional neural networks for landslide susceptibility mapping in Yanshan County，China[J]. Science of the Total Environment，2019，666(MAY 20)：975–993.
[17]	BUI D T，PRADHAN B，LOFMAN O，et al. Spatial prediction of landslide hazards in Hoa Binh province(Vietnam)：A comparative assessment of the efficacy of evidential belief functions and fuzzy logic models[J]. Catena，2012，96(none)：28–40.
[19]	MARJANOVI M，KOVA EVI M，BAJAT B，et al. Landslide susceptibility assessment using SVM machine learning algorithm[J]. Engineering Geology，2011，123(3)：225–234.
[20]	HONG H，ILIA I，TSANGARATOS P，et al. A hybrid fuzzy weight of evidence method in landslide susceptibility analysis on the Wuyuan area，China[J]. Geomorphology，2017，290(1)：1–16.
[2]	ZHAO Z A，HE Y，YAO S，et al. A comparative study of different neural network models for landslide susceptibility mapping[J]. Advances in Space Research，2022，70(2)：383–401.
[6]	张玘恺，凌斯祥，李晓宁，等. 九寨沟县滑坡灾害易发性快速评估模型对比研究[J]. 岩石力学与工程学报，2020，39(8)：1 595–1 610. (ZHANG Qikai，LING Sixiang，LI Xiaoning，et al. Comparison of landslide susceptibility mapping rapid assessment models in Jiuzhaigou County，Sichuan province，China[J]. Chinese Journal of Rock Mechanics and Engineering，2020，39(8)：1 595–1 610.(in Chinese))
[8]	FERRETTI A，PRATI C，ROCCA F. Permanent scatterers in SAR interferometry[J]. IEEE Transactions on Geoscience and Remote Sensing，2002，39(1)：8–20.
[10]	YAO S，YANG K，DONG X，et al. Large-gradient deformation monitoring and parameter inversion in a mining area using a method combining a dynamic prediction model and InSAR[J]. Remote Sensing Letters，2021，12(9)：838–847.
[14]	蔡广珍，杨子和，王东平，等. 甘肃临夏州地质灾害成因分析及防御对策[J]. 农业灾害研究，2015，5(4)：32–35，55.(CAI Guangzhen，YANG Zihe，WANG Dongping，et al. Analysis of geological disaster causes and defense measures in Linxia Prefecture，Gansu[J]. Agricultural Disaster Research，2015，5(4)：32–35，55.(in Chinese))
[16]	REICHENBACH P，ROSSI M，MALAMUD B，et al. A review of statistically-based landslide susceptibility models[J]. Earth-Science Reviews，2018，1(180)：60–91.
[18]	陈学兄. 基于遥感与GIS的中国水土流失定量评价[博士学位论文][D]. 杨凌：西北农林科技大学，2013.(CHEN Xuexiong. Quantitative evaluation of soil erosion in China based on remote sensing and GIS [Ph. D. Thesis][D]. Yangling：Northwest Agriculture and Forestry University of Science and Technology，2013.(in Chinese))
[26]	MODESTE G，DOUBRE C，MASSON F. Time evolution of mining-related residual subsidence monitored over a 24-year period using InSAR in southern Alsace，France[J]. International Journal of Applied Earth Observation and Geoinformation，2021，3(102)：1–11.
[22]	HUANG F，YAO C，LIU W，et al. Landslide susceptibility assessment in the Nantian area of China：a comparison of frequency ratio model and support vector machine[J]. Geomatics，2018，1(9)：919–938.
[24]	朱赛楠，殷跃平，王猛，等. 金沙江结合带高位远程滑坡失稳机理及减灾对策研究——以金沙江色拉滑坡为例[J]. 岩土工程学报，2021，43(4)：688–697.(ZHU Sainan，YIN Yueping，WANG Meng，et al. Research on the destabilization mechanism and disaster mitigation measures of high remote landslides in the Jinsha River combined zone—taking the Jinsha River Sera landslide as an example[J]. Chinese Journal of Geotechnical Engineering，2021，43(4)：688–697.(in Chinese))
[30]	王毅，方志策，牛瑞卿，等. 基于深度学习的滑坡灾害易发性分析[J]. 地球信息科学学报，2021，23(12)：2 244–2 260.(WANG Yi，FANG Zhice，NIU Ruiqing，et al. Landslide hazard susceptibility analysis based on deep learning[J]. Journal of Geoinformation Science，2021，23(12)：2 244–2 260.(in Chinese))