>> 自然科学版 >> 网络首发 >> 正文
基于优化 MaxEnt 模型的怒江州滑坡易发性评价
时间: 2024-01-02 次数:

李益敏,向倩英,邓选伦,等. 基于优化 MaxEnt 模型的怒江州滑坡易发性评价[J. 河南理工大学学报(自然科学版),doi:10.16186/j.cnki.1673-9787.2023070010.

LI Y M,XIANG Q Y,DENG X L,et al. Evaluation of landslide susceptibility in Nujiang prefecture based on optimized MaxEnt modelJ.Journal of Henan Polytechnic University( Natural Science) doi:10.16186/j.cnki.1673-9787.2023070010.

基于优化MaxEnt模型的怒江州滑坡易发性评价(网络首发)

李益敏12,向倩英3,邓选伦1,冯显杰3

1.云南大学 地球科学学院,云南 昆明 6505002.云南大学 云南省高校国产高分卫星遥感地质工程研究中心,云南 昆明 6505003.云南大学 国际河流与生态安全研究院,云南 昆明 650500

摘要: 目的 怒江州是典型的高山峡谷地区,地质灾害(滑坡)频发,严重制约着当地的发展,为解决这一问题,方法 综合考虑怒江州实际情况,从气象水文、地形地貌、地层岩性、植被生态和人类活动5个方面选取坡向、高程等14 个影响因子判断相关性,构建评价指标体系,对最大熵(MaxEnt)模型的特征类(feature combinationFC)和正则化乘数(regularization multiplierRM)参数进行优化,对比优化前后小样本赤池信息量准则(akaike information criterionAICc)、遗漏率(omission rateOR)和AUCarea under curve)值,然后基于优化的最大熵(MaxEnt)模型预测滑坡灾害的发生,实现怒江州滑坡易发性评价。结果 结果表明:优化后的 MaxEnt 模型在研究区滑坡易发性预测中适用性优秀(AUC=0.913);运用刀切法(Jackknife)计算各影响因子对易发性的影响程度,高程(S323.2%)、坡度(S922.4%)、居民点密度(S514.2%)、距河流距离(S1313.7%)、距道路距离(S49.6%)和岩性(S78.7%)是位列前六的因子,累计贡献度达91.8%;极高、高、中、低滑坡易发性等级的空间占比分别为 4.88%8.96%18.40%67.76%,县域中极高和高易发区占比最大的是泸水市,整体上,极高、高易发区主要沿河流和道路分布于峡谷中,低易发区主要分布于人类活动少、河谷不发育的区域。结论 优化后的 MaxEnt模型更适合怒江州滑坡易发性预测,研究结果可为怒江州防灾减灾与土地利用规划提供参考。

关键词: 怒江州;最大熵(MaxEnt)模型;滑坡;易发性

中图分类号:P694

doi:10.16186/j.cnki.1673-9787. 2023070010

基金项目: 国家自然科学基金资助项目(41161070);云南省科技厅-云南大学联合基金重点资助项目(2019FY003017);云南大学大湄公河次区域气候变化研究省创新团队项目(2019HC027);中国地质调查局项目(DD20221824

收稿日期:2023/07/06

修回日期:2023/09/14

网络首发日期:2024/01/02

 

 

Evaluation of landslide susceptibility in Nujiang prefecture based on optimized MaxEnt model

LI Yimin12XIANG Qianying3DENG Xuanlun1FENG Xianjie3

 1.School of Earth SciencesYunnan UniversityKunming 650500YunnanChina2.Yunnan University Engineering Research Center of High-resolution Satellite Remote SensingYunnan UniversityKunming 650500YunnanChina3.Institute of International Rivers and Ecological SecurityYunnan UniversityKunming 650500YunnanChina

Abstract: Objective Nujiang Prefecture is a typical alpine canyon area.Landscape geological disasters are frequentwhich seriously restricts local development.It is urgent to carry out sensitive evaluation.Methods Comprehensively considering the actual situation of Nujiang Prefecture14 influencing factors such as slope direction and elevation were selected from five aspects of meteorology and hydrologytopography and geomorphologystratigraphic lithologyvegetation ecology and human activities to judge the correlationbuild an evaluation index systemand analyze the MaxEnt model feature combinationFC and regularization multiplier RM parameters were optimizedand the akaike information criterionAICc),Omission RateOR and AUCarea under curve valueand then based on the optimized MaxEnt model to predict the occurrence of landslide hazardsto realize the sensitivity evaluation of landslide in Nujiang Prefecture.Results The optimized MaxEnt model was suitable for landslide sensitivity predictionAUC=0.913.Jackknife method was used to calculate the influence degree of each influencing factor on sensitivitysuch as elevationS323.2%),slopeS922.4%),settlement densityS514.2%),distance from riverS1313.7%),distance from roadS49.6% and lithologyS78.7% is the top six factorswith a cumulative contribution of 91.8%The spatial proportions of extremely highhighmedium and low landslide susceptibility levels were 4.88%8.96%18.40% and 67.76%respectively.The highest proportion of extremely high and high susceptibility areas was found in Lushui City.On the wholeextremely high and high susceptibility areas were mainly distributed in valleys along rivers and roadswhile low susceptibility areas were mainly distributed in areas with little human activities and undeveloped river valleys.Conclusion The optimized MaxEnt model is more suitable for landslide sensitivity prediction in Nujiang Prefectureand the research results can provide reference for disaster prevention and reduction and land use planning in Nujiang Prefecture.

Key words: Nujiang Statemaximum entropyMaxEntmodellandslidesusceptibility

CLC:P694

最近更新