LIU Yabin,LIU Yunteng,ZHU Zhenmei,et al.A Comparative Study on Landslide Hazard Risk Assessment Models in Sichuan Province[J].Journal of Chengdu University of Information Technology,2024,39(01):101-107.[doi:10.16836/j.cnki.jcuit.2024.01.015]
四川省滑坡灾害危险性评价模型比较研究
- Title:
- A Comparative Study on Landslide Hazard Risk Assessment Models in Sichuan Province
- 文章编号:
- 2096-1618(2024)01-0101-07
- Keywords:
- landslide disasters; informative approach; random forest approach; logistic regression methods; model evaluation
- 分类号:
- P49
- 文献标志码:
- A
- 摘要:
- 为探讨四川省滑坡灾害分布状况及对滑坡灾害危险性评价建模效果进行比较,以四川省为研究区,分析地质灾害尤其是滑坡灾害的时空分布,运用信息量法、基于信息量的逻辑回归法、随机森林法3种方法,取坡度、植被覆盖率、地震烈度、断层密度及年累计降雨量5个因子建立危险性评价模型,结合ROC曲线和AUC值对模型性能进行精度比较。研究结果表明:(1)坡度因子对滑坡灾害的贡献最大,3种评价模型中坡度、地震烈度及降雨3个因子均有较大贡献。(2)滑坡灾害主要分布于地震带附近,在成都平原处灾害发生几率最小。(3)3种评价方法均适用于滑坡危险性区划工作,具有良好的建模效果。相对于其他两种模型,逻辑回归模型在实际情况下更加适合四川省滑坡灾害的危险性评价工作。
- Abstract:
- To explore the distribution of landslide disasters in Sichuan Province and compare the modeling effect of landslide hazard risk assessment, this paper takes Sichuan Province as the research area to analyze the temporal and spatial distribution of geological disasters, especially landslide disasters, and uses three methods: information quantity method, information-based logistic regression method, and random forest method, to establish a risk evaluation model based on five factors: slope, vegetation coverage, seismic intensity, fault density and annual cumulative rainfall. This paper combines the ROC curve and AUC value to accurately compare the model performance. The results show that:(1)The slope factor has the greatest contribution to landslide disasters, and the three factors of the slope, seismic intensity and rainfall in the three evaluation models have a greater impact.(2)Landslide disasters are mainly distributed near seismic zones, and the probability of disasters is the smallest in the Chengdu Plain.(3)The three evaluation methods are suitable for landslide hazard zoning and have good modeling effects. Among the three models, the logistic regression model is more suitable for the risk assessment of landslide disasters in Sichuan Province under actual circumstances.
参考文献/References:
[1] 方匡南,吴见彬,朱建平,等.随机森林方法研究综述[J].统计与信息论坛,2011,26(3):32-38.
[2] 赵艳南,牛瑞卿.基于证据权法的滑坡危险性区划探索[J].地理与地理信息科学,2010,26(6):19-23.
[3] 李亭,田原,邬伦,等.基于随机森林方法的滑坡灾害危险性区划[J].地理与地理信息科学,2014,30(6):25-30.
[4] Xiao C C,Tian Y,Shi W Z,et al.A new method of pseudo absence data generation in landslide susceptibility mapping with a case study of Shenzhen[J].Science China(Technological Sciences),2010,53(S1):75-84.
[5] 司康平,田原,汪大明,等.滑坡灾害危险性评价的3种统计方法比较——以深圳市为例[J].北京大学学报(自然科学版),2009,45(4):639-646.
[6] 王丽婷,丁晓青,方驰.基于随机森林的人脸关键点精确定位方法[J].清华大学学报(自然科学版),2009,49(4):543-546.
[7] 陈朝亮.基于GIS的内江市地质灾害易发性评价研究[D].绵阳:西南科技大学,2019.
[8] 张晓东.基于遥感和GIS的宁夏盐池县地质灾害风险评价研究[D].北京:中国地质大学,2018.
[9] 叶帮苹,冯汉中,刘志红,等.基于Logistic模型的四川山洪流域危险性评价[J].成都信息工程大学学报,2020,35(5):573-578.
[10] 胡德勇,李京,陈云浩,等.GIS支持下滑坡灾害空间预测方法研究[J].遥感学报,2007(6):852-859.
[11] Tian Y,Xiao C C,Liu Y,et al.Effects of raster resolution on landslide susceptibility mapping:A case study of Shenzhen[J].Science in China(Series E:Technological Sciences),2008(S2):188-198.
[12] 阮沈勇,黄润秋.基于GIS的信息量法模型在地质灾害危险性区划中的应用[J].成都理工学院学报,2001(1):89-92.
[13] 柏延臣,王劲峰.结合多分类器的遥感数据专题分类方法研究[J].遥感学报,2005(5):555-563.
[14] 唐伟,周志华.基于Bagging的选择性聚类集成[J].软件学报,2005(4):496-502.
[15] 许向宁. 高地震烈度区山体变形破裂机制地质分析与地质力学模拟研究[D].成都:成都理工大学,2006.
[16] 唐晓玲,钱江澎.四川省丹巴县建设街滑坡特征及形成机理[J].四川地质学报,2013,33(3):340-343.
相似文献/References:
[1]何 强,王增武,卢德全,等.南充市滑坡灾害易发性区划与评价[J].成都信息工程大学学报,2021,36(01):118.[doi:10.16836/j.cnki.jcuit.2021.01.018]
HE Qiang,WANG Zengwu,LU Dequan,et al.Risk Zonation and Evaluation of Landslide Geohazards in Nanchong[J].Journal of Chengdu University of Information Technology,2021,36(01):118.[doi:10.16836/j.cnki.jcuit.2021.01.018]
备注/Memo
收稿日期:2023-02-28
基金项目:四川省科技厅重点研发资助项目(2022YFS0542)
通信作者:刘云腾.E-mail:592206414@qq.com