COMPARISON OF SNOW AVALANCHE SUSCEPTIBILITY ASSESSMENT AND POTENTIAL SNOW AVALANCHE RELEASE AREAS IDENTIFICATION ALONG YINING-AKSU RAILWAY, XINJIANG TIANSHAN MOUNTAINS
-
摘要: 准确识别雪崩潜在释放区或雪崩高易发区域对高寒山区工程建设减灾防灾意义重大,特别是在雪崩监测数据缺失地区,能够提供重要的区域性灾害风险预估参考。本文以新疆天山地区伊阿铁路沿线区域为例,将铁路沿线154个雪崩范围形成区作为评价样本,开展基于机器学习算法的雪崩易发性评价,构建新疆天山地区伊阿铁路沿线雪崩易发性评价体系;开展基于数据叠加的雪崩潜在释放区(PRA)识别,绘制伊阿铁路沿线雪崩潜在释放区分布图,并对两个结果通过Kappa系数和AUC值进行检验,并对比讨论。结果显示,支持向量机(SVM)、多层感知器(MLP)、PRA的Kappa系数分别为0.806、0.774、0.600;AUC值分别为0.993、0.961、0.802,机器学习算法在雪崩易发性评价中的表现优于传统的基于数据叠加的雪崩潜在释放区识别算法;两种机器学习算法模型均精度高,其中支持向量机(SVM)算法表现最佳,优于多层感知器(MLP),评价结果比较符合野外雪崩发育实际情况,可为高寒山区重大工程建设的雪崩防灾减灾工作提供基础的科学依据;雪崩潜在释放区的自动识别算法评价能力较弱,评价结果基本符合野外雪崩发育实际情况,对于缺乏可用数据的高寒山区具有评价意义。Abstract: Accurate identification of potential snow avalanche release areas or the snow avalanche high-proning areas are crucial in disaster mitigation and prevention of engineering construction in alpine mountains. This paper uses the Yining-Aksu railway in the Tianshan region of Xinjiang Uygur Autonomous Region as case study. It collectes 154 snow avalanche starting zones as assessment samples,and conductes snow avalanche susceptibility assessment based on machine learning algorithm,to construct the assessment system along the railway. We identify the potential snow avalanche release areas(PRA)based on data superposition,to map the distribution illustration of its PRA. Then we test and compare both results using Kappa coefficient and AUC values. The results show that the Kappa coefficients by SVM,MLP and PRA are 0.806,0.774 and 0.600,respectively,while AUC values are 0.993,0.961 and 0.802,respectively. Machine learning algorithms in snow avalanche susceptibility assessment outperform traditional PRA identification algorithm based on data superposition. Both machine learning algorithm models have high accuracy,with SVM performing the best,outperforming MLP. The assessment results are more consistent with the actual situation of snow avalanche development in the study area,and it can provide a basic scientific basis for snow avalanche prevention and mitigation in major engineering construction in alpine mountains. Albeit the automatic algorithm of PRA identification is relatively defective,the assessment results are basically consistent with the actual situation of snow avalanche development in the study area,making it meaningful for assessing in alpine mountain where available data is limited.
-
图 2 雪崩野外调查照片及遥感解译特征
a. 铁力买提达坂残留的雪崩堆积体;b. 六月残留的雪崩堆积体热融溶洞;c. 雪崩修剪形成的旗形树;d. 拉尔墩雪崩铲刮斜坡面浮土形成的堆积体
Figure 2. Snow avalanche accumulation fieldwork photos and remote sensing interpretation features:(a) Residual snow avalanche accumulation in Tielimaiti hilltop; (b) Residual snow avalanche accumulation's hot melting cave in June; (c) Snow avalanche pruning to form a flag tree; (d) Snow avalanche scraping the slope surface to form the accumulation in Laerdun
图 3 研究区雪崩分布图
a. 铁路沿线雪崩空间分布图;b. 巩乃斯、拉尔敦段雪崩频发区域;c. 拉尔敦典型雪崩流通沟槽、雪崩堆积体;d. 阿尔先段雪崩频发区域;e. 苏力间、铁力买提达坂段雪崩频发区域
Figure 3. Snow avalanches inventory illustration of the study area:(a) Snow avalanches spatial distribution illustration along the railway; (b) Frequent snow avalanches spatial distribution illustration in Gongnaisi and Laerdun; (c) Typical snow avalanche gully and accumulation in Laerdun; (d) Frequent snow avalanches spatial distribution illustration in Aerxian; (e) Frequent snow avalanches spatial distribution illustration in Sulijian and Tielimaiti hilltop
图 7 雪崩潜在释放区分布图
a. 铁路沿线雪崩潜在释放区空间分布图;b. 巩乃斯、拉尔敦段雪崩频发区域潜在释放区空间分布图;c. 阿尔先段雪崩频发区域潜在释放区空间分布图;d. 苏力间、铁力买提达坂段雪崩频发区域潜在释放区空间分布图
Figure 7. Potential snow avalanche release areas distribution illustration:(a) PRA spatial distribution illustration along the railway; (b) Frequent PRA spatial distribution illustration in Gongnaisi and Laerdun; (c) Frequent PRA spatial distribution illustration in Aerxian; (d) Frequent PRA spatial distribution illustration in Sulijian and Tielimaiti hilltop
表 1 雪崩形成区参数指征
Table 1. Snow avalanche starting zone parameter indications
参数 最小值 最大值 第一四分位数 第三四分位数 均值 坡度 10.00 75.10 26.13 37.21 31.46 粗糙度 1.00 3.89 1.11 1.26 1.20 曲率 0.03 60.64 8.37 19.90 14.97 表 2 混淆矩阵及Kappa系数
Table 2. Confusion matrix and Kappa coefficients
实际雪崩情况 预测结果 标记符号 PRA SVM MLP 1 1 N1 100 28 29 1 0 N2 8 3 5 0 1 N3 54 3 2 0 0 N4 146 28 26 Pe 0.500 0.500 0.500 Pa 0.800 0.903 0.887 Ka 0.600 0.806 0.774 -
Baum E B. 1988. On the capabilities of multilayer perceptrons[J]. Journal of Complexity,4 (3): 193-215. doi: 10.1016/0885-064X(88)90020-9 Bühler Y, Kumar S, Veitinger J, et al. 2013. Automated identification of potential snow avalanche release areas[J]. Natural Hazards and Earth System Sciences, 13 : 1321-1335. doi: 10.5194/nhess-13-1321-2013 Bühler Y, Rickenbach D V, Stoffel A, et al. 2018. Automated snow avalanche release area delineation-validation of existing algorithms and proposition of a new object-based approach for large-scale hazard indication mapping[J]. Natural Hazards and Earth System Sciences, 18 : 3235-3251. doi: 10.5194/nhess-18-3235-2018 Choubin B, Borji M, Mosavi A, et al. 2019. Snow avalanche hazard prediction using machine learning methods[J]. Journal of Hydrology, 577: 123929. doi: 10.1016/j.jhydrol.2019.123929 Cristianini N, Taylor J S. 2005. An introduction to support vector machines: and other kernel-based learning methods[M]. Cambridge: Cambridge University Press. Fan L J, Li J P, Wei Z G, et al. 2003. Annual variations of the arctic oscillation and the antarctic oscillation[J]. Chinese Journal of Atmospheric Sciences, 27 (3): 419-424. Haegeli P, Bühler Y, Sykes J. 2021. Automated snow avalanche release area delineation in data-sparse, remote, and forested regions[J]. Natural Hazards and Earth System Sciences, 22 (10): 3247-3270. Han L. 2004. The classification model of RS images based on artificial neural network——MLP[J]. Bulletin of Surveying and Mapping, (9): 29-30, 42. Hao J S, Huang F R, Liu Y, et al. 2018. Avalanche activity and characteristics of its triggering factors in the western Tianshan Mountains, China[J]. Journal of Mountain Science, 15 (7): 1397-1411. doi: 10.1007/s11629-018-4941-2 Hong H, Liu J, Zhu A. 2020. Modeling landslide susceptibility using LogitBoost alternating decision trees and forest by penalizing attributes with the bagging ensemble[J]. Science of The Total Environment, 718: 137231. doi: 10.1016/j.scitotenv.2020.137231 Huang J P, Ling S X, Wu X Y, et al. 2022a. GIS-based comparative study of the Bayesian Network, Decision Table, Radial Basis Function Network and Stochastic Gradient Descent for the spatial prediction of landslide susceptibility[J]. Land, 11(436): 1-25. Huang J P, Wu X Y, Ling S X, et al. 2022b. A bibliometric and content analysis of research trends on GIS-based landslide susceptibility from 2001 to 2020[J]. Environmental Science and Pollution Research, 29 : 86954-86993. doi: 10.1007/s11356-022-23732-z Kumar S, Srivastava P K, Snehmani, et al. 2019. Geospatial probabilistic modelling for release area mapping of snow avalanches[J]. Cold Regions Science and Technology, 165: 102813. doi: 10.1016/j.coldregions.2019.102813 Lima P, Steger S, Glade T, et al. 2022. Literature review and bibliometric analysis on data-driven assessment of landslide susceptibility[J]. Journal of Mountain Science, 19 (6): 1670-1698. doi: 10.1007/s11629-021-7254-9 Liu F Z, Wang L, Xiao D S. 2021. Application of machine learning model in landslide susceptibility evaluation[J]. The Chinese Journal of Geological Hazard and Control, 32 (6): 98-106. Liu Y H, Fang R K, Su Y C, et al. 2021. Machine learning based model for warning of regional landslide disasters[J]. Journal of Engineering Geology, 29 (1): 116-124. Lora D, Contador I, Perez-regadera J F, et al. 2016. Features of the area under the receiver operating characteristic(ROC)curve. A good practice[J]. Stata Journal, 16 (1): 185-196. doi: 10.1177/1536867X1601600115 Lorenza S. 1995. Support-vector networks[J]. Meachine Learing, 20 : 273-297. Luo L G, Pei X J, Huang R Q, et al. 2021. Landslide susceptibility assessment in Jiuzhaigou scenic area with GIS based on certainty factor and Logistic regression model[J]. Journal of Engineering Geology, 29 (2): 526-535. Mandrekar, Jayawant N. 2010. Receiver Operating Characteristic Curve in diagnostic test assessment[J]. Journal of Thoracic Oncology, 5 (9): 1315-1316. doi: 10.1097/JTO.0b013e3181ec173d Margherita M, Gruber U, Stoffel A. 2002. Definition and characterisation of potential avalanche release areas[J]. Cold Regions Science and Technology, 37 : 407-419. Mcclung D M, Peter S. 2006. The avalanche handbook[M]. The Mountaineers Books. Moore I D, Grayson R B, Landson A R. 1991. Digital terrain modelling: a review of hydrological, geomorphological, and biological applications[J]. Hydrological Process, 5 : 3-30. doi: 10.1002/hyp.3360050103 Omid R, Omid G, Teimur T, et al. 2019. Spatial modeling of snow avalanche using machine learning models and geo-environmental factors: comparison of effectiveness in two mountain regions[J]. Remote Sensing, 11: 2995. doi: 10.3390/rs11242995 Pan J P, Liu C G. 2008. Interpreting the 3-13 Gozigou snow avalanche[J]. Meteorological Knowledge, (3): 43-46. Qiu J Q. 2005 Snow Avalanches[M]. Urumqi: Xinjiang Science and Technology Press. Rumelhart D E, Widrow B, Lehr M. 1994. The basic ideas in neural networks[J]. Communications of the ACM, 37 (3): 87-92. doi: 10.1145/175247.175256 Schweizer J, Bartelt P, Herwijnen A V. 2015. Snow avalanches in snow and ice-related hazards, risks and disasters[M]. Waltham: Academic Press. Schweizer J, Jamieson J B, Schneebeli M. 2003. Snow avalanche formation[J]. Reviews of Geophysics, 41 (4): 1-25. Schweizer J, Kronholm K, Jamieson J B, et al. 2008. Review of spatial variability of snowpack properties and itsimportance for avalanche formation[J]. Cold Regions Science and Technology, 51(2-3): 253-272. doi: 10.1016/j.coldregions.2007.04.009 Tamura R, Kobayashi K, Takano Y, et al. 2019. Mixed integer quadratic optimization formulations for eliminating multicollinearity based on variance inflation factor[J]. Journal of Global Optimization, 73 (2): 431-446. doi: 10.1007/s10898-018-0713-3 Veitinger J, Purves R S, Sovilla B. 2016. Potential slab avalanche release area identification from estimated winter terrain: a multi-scale, fuzzy logic approach[J]. Natural Hazards and Earth System Sciences, 16 (10): 2211-2225. doi: 10.5194/nhess-16-2211-2016 Viglietti D, Letey S, Motta R, et al. 2010. Snow avalanche release in forest ecosystems: A case study in the Aosta Valley Region(NW-Italy)[J]. Cold Regions Science and Technology, 64 : 167-173. doi: 10.1016/j.coldregions.2010.08.007 Wang S, Ling S X, Wu X Y, et al. 2023. Key predisposing factors and susceptibility assessment of landslides along the Yunnan-Tibet traffic corridor, Tibetan Plateau: Comparison with the LR, RF, NB, and MLP techniques[J]. Frontiers in Earth Science, 10: 1100363. doi: 10.3389/feart.2022.1100363 Wang Y L. 1992. China Snow Avalanche Study[M]. Beijing: Ocean Press. Wen H, Wang D, Wang S R, et al. 2021. Key predisposing factors and susceptibility mapping of snow avalanche in Parlung-Tsangpo catchment, southeast Tibetan Plateau[J]. Journal of Engineering Geology, 29 (2): 404-415. Wen H, Wu X Y, Liao X, et al. 2022a. Application of machine learning methods for snow avalanche susceptibility mapping in the Parlung Tsangpo catchment, southeastern Qinghai-Tibet Plateau[J]. Cold Regions Science and Technology, 198: 103535. doi: 10.1016/j.coldregions.2022.103535 Wen H, Wu X Y, Ling S Y, et al. 2022b. Characteristics and susceptibility assessment of the earthquake-triggered landslides in moderate-minor earthquake prone areas at southern margin of Sichuan Basin, China[J]. Bulletin of Engineering Geology and the Environment, 81: 346. doi: 10.1007/s10064-022-02821-w Wen H, Wu X Y, Zhao S Y, et al. 2022. Snow avalanche susceptibility evaluation in the central Shaluli Mountains of Tibetan Plateau based on machine learning method[J]. Journal of Glaciology and Geocryology, 44 (6): 1694-1706. Yang J, He Q, Liu Y. 2022. Winter-spring prediction of snow avalanche susceptibility using optimisation multi-source heterogeneous factors in the western Tianshan Mountains, China[J]. Remote Sensing, 14 (6): 1-34. Yang Q. 2021. SVM Algorithm for N1+N2 structure syntax relation determination[J]. Computer Engineering and Applications, 57 (20): 104-108. Yariyan P, Avand M, Abbaspour R A, et al. 2020. GIS-based spatial modeling of snow avalanches using four novel ensemble models[J]. Science of The Total Environment, 745: 141008. doi: 10.1016/j.scitotenv.2020.141008 Yu Z Y, Cheng Y X, Lü Y, et al. 2022. Rockfall hazard assessment in canyon areas incorporating regional-scale identification of potential rockfall source areas[J]. Journal of Engineering Geology, 30 (5): 1583-1596. Yuanhai S, Naiyang D, Lingwei H, et al. 2020. Key issues of support vector machines and future prospects[J]. Scientia Sinica Mathematica, 50 (9): 1233-1248. doi: 10.1360/SSM-2020-0015 Zhang H Y, Huang H B, Yan Z X. 2004. Forming conditions and provention treatment of loess landlide in the Xinyuan Mountain area, xinjiang[J]. Xinjiang Geology, 3 (22): 233-237. Zhang L X, Wei W S. 2002. Variation trends of snowcover in the middle mountains of Western Tianshan Mts. and their relations to temperature and precipitation[J]. Scientia Geographica Sinica, 22 (1): 67-71. Zhang Q K, Ling S X, Li X N, et al. 2020. Comparison of landslide susceptibility mapping rapid assessment models in Jiuzhaigou County, Sichuan province, China[J]. Chinese Journal of Rock Mechanics and Engineering, 39 (8): 1595-1610. 仇家琪. 2005. 雪崩学[M]. 乌鲁木齐: 新疆科学技术出版社. 范丽军, 李建平, 韦志刚, 等. 2003. 北极涛动和南极涛动的年变化特征[J]. 大气科学, 27 (3): 419-424. https://www.cnki.com.cn/Article/CJFDTOTAL-DQXK200303010.htm 韩玲. 2004. 基于人工神经网络——多层感知器(MLP)的遥感影像分类模型[J]. 测绘通报, (9): 29-30, 42. https://www.cnki.com.cn/Article/CJFDTOTAL-CHTB200409010.htm 刘福臻, 王灵, 肖东升. 2021. 机器学习模型在滑坡易发性评价中的应用[J]. 中国地质灾害与防治学报, 32 (6): 98-106. https://www.cnki.com.cn/Article/CJFDTOTAL-ZGDH202106012.htm 刘艳辉, 方然可, 苏永超, 等. 2021. 基于机器学习的区域滑坡灾害预警模型研究[J]. 工程地质学报, 29 (1): 116-124. doi: 10.13544/j.cnki.jeg.2020-533 罗路广, 裴向军, 黄润秋, 等. 2021. GIS支持下CF与Logistic回归模型耦合的九寨沟景区滑坡易发性评价[J]. 工程地质学报, 29 (2): 526-535. doi: 10.13544/j.cnki.jeg.2019-202 潘继鹏, 刘成刚. 2008. 解读3·13果子沟雪崩[J]. 气象知识, (3): 43-46. 王彦龙. 1992. 中国雪崩研究[M]. 北京: 海洋出版社. 文洪, 王栋, 王生仁, 等. 2021. 藏东南帕隆藏布流域雪崩关键影响因素与易发性区划研究[J]. 工程地质学报, 29 (2): 404-415. doi: 10.13544/j.cnki.jeg.2021-0121 文洪, 巫锡勇, 赵思远, 等. 2022. 基于机器学习法的青藏高原沙鲁里山系中段雪崩易发性评价研究[J]. 冰川冻土, 44 (6): 1694-1706. https://www.cnki.com.cn/Article/CJFDTOTAL-BCDT202206002.htm 杨泉. 2021. N1+N2结构语法关系判定的SVM算法[J]. 计算机工程与应用, 57 (20): 104-108. https://www.cnki.com.cn/Article/CJFDTOTAL-JSGG202120012.htm 俞朝悦, 成玉祥, 吕艳, 等. 2022. 融合区域潜在落石源区识别的峡谷区落石危险性评价[J]. 工程地质学报, 30 (5): 1583-1596. doi: 10.13544/j.cnki.jeg.2022-0477 张鸿义, 黄洪标, 闫中学. 2004. 新疆新源山区黄土滑坡形成条件与防治措施[J]. 新疆地质, 22 (3): 233-237. https://www.cnki.com.cn/Article/CJFDTOTAL-XJDI200403001.htm 张丽旭, 魏文寿. 2002. 天山西部中山带积雪变化趋势与气温和降水的关系——以巩乃斯河谷为例[J]. 地理科学, 22 (1): 67-71. https://www.cnki.com.cn/Article/CJFDTOTAL-DLKX200201012.htm 张玘恺, 凌斯祥, 李晓宁, 等. 2020. 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, 39 (8): 1595-1610. https://www.cnki.com.cn/Article/CJFDTOTAL-YSLX202008009.htm -