Landslide susceptibility mapping using physics-guided machine learning: a case study of a debris flow event in Colorado Front Range

IF 5.6 1区 工程技术 Q1 ENGINEERING, GEOLOGICAL
Te Pei, Tong Qiu
{"title":"Landslide susceptibility mapping using physics-guided machine learning: a case study of a debris flow event in Colorado Front Range","authors":"Te Pei,&nbsp;Tong Qiu","doi":"10.1007/s11440-024-02384-y","DOIUrl":null,"url":null,"abstract":"<div><p>Landslides are common geohazards worldwide, resulting in significant losses to economies and human lives. Data-driven approaches, especially machine learning (ML) models, have been widely used recently for landslide susceptibility mapping (LSM) by extracting features from geospatial variables based on their contribution to landslide occurrences using known distributions of landslides as the training dataset. However, challenges remain in applying ML models for LSM models due to the scarcity and uneven spatial distribution of landslide data coupled with the spatial heterogeneity of hillslope conditions. Moreover, ML models developed with limited data often exhibit unexpected behaviors, resulting in poor interpretability and predictions that deviate from intuitive expectations and established domain knowledge. To overcome these challenges, this study proposes a physics-guided machine learning (PGML) framework that integrates landslide domain knowledge into ML models for LSM. The PGML framework was developed and assessed using a detailed debris flow inventory from a storm event in the Colorado Front Range. Based on the infinite slope model, the factor of safety for the study area was first determined and was subsequently used to constrain the prediction of ML models through a modified loss function and measure the physics consistency of model predictions. To evaluate the robustness and generalizability of the models, this study uses geographical sample selections for model performance evaluation, where ML models are trained and tested across heterogeneous ecoregions. The results of this study demonstrated the efficacy of both physics-based and data-driven methods in determining landslide susceptibility in the study area; however, pure data-driven ML models produced physically unrealistic results and poor generalization performance in new ecoregions. With the incorporation of physical constraints, the PGML model demonstrated notable enhancements in physics consistency and generalization capability, along with reduced model uncertainties across various ecoregions, surpassing the performance of benchmark ML models.</p></div>","PeriodicalId":49308,"journal":{"name":"Acta Geotechnica","volume":"19 10","pages":"6617 - 6641"},"PeriodicalIF":5.6000,"publicationDate":"2024-08-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Acta Geotechnica","FirstCategoryId":"5","ListUrlMain":"https://link.springer.com/article/10.1007/s11440-024-02384-y","RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"ENGINEERING, GEOLOGICAL","Score":null,"Total":0}
引用次数: 0

Abstract

Landslides are common geohazards worldwide, resulting in significant losses to economies and human lives. Data-driven approaches, especially machine learning (ML) models, have been widely used recently for landslide susceptibility mapping (LSM) by extracting features from geospatial variables based on their contribution to landslide occurrences using known distributions of landslides as the training dataset. However, challenges remain in applying ML models for LSM models due to the scarcity and uneven spatial distribution of landslide data coupled with the spatial heterogeneity of hillslope conditions. Moreover, ML models developed with limited data often exhibit unexpected behaviors, resulting in poor interpretability and predictions that deviate from intuitive expectations and established domain knowledge. To overcome these challenges, this study proposes a physics-guided machine learning (PGML) framework that integrates landslide domain knowledge into ML models for LSM. The PGML framework was developed and assessed using a detailed debris flow inventory from a storm event in the Colorado Front Range. Based on the infinite slope model, the factor of safety for the study area was first determined and was subsequently used to constrain the prediction of ML models through a modified loss function and measure the physics consistency of model predictions. To evaluate the robustness and generalizability of the models, this study uses geographical sample selections for model performance evaluation, where ML models are trained and tested across heterogeneous ecoregions. The results of this study demonstrated the efficacy of both physics-based and data-driven methods in determining landslide susceptibility in the study area; however, pure data-driven ML models produced physically unrealistic results and poor generalization performance in new ecoregions. With the incorporation of physical constraints, the PGML model demonstrated notable enhancements in physics consistency and generalization capability, along with reduced model uncertainties across various ecoregions, surpassing the performance of benchmark ML models.

Abstract Image

利用物理引导的机器学习绘制滑坡易发性地图:科罗拉多前沿山脉泥石流事件案例研究
山体滑坡是全球常见的地质灾害,给经济和人类生命造成了重大损失。最近,数据驱动方法,特别是机器学习(ML)模型,已被广泛应用于滑坡易发性绘图(LSM),方法是从地理空间变量中提取特征,这些特征基于已知的滑坡分布作为训练数据集对滑坡发生的贡献。然而,由于滑坡数据稀缺且空间分布不均,再加上山坡条件的空间异质性,将 ML 模型应用于 LSM 模型仍面临挑战。此外,利用有限数据开发的 ML 模型经常表现出意想不到的行为,导致可解释性差,预测结果偏离直观预期和已有的领域知识。为了克服这些挑战,本研究提出了一个物理引导机器学习(PGML)框架,将滑坡领域的知识整合到用于 LSM 的 ML 模型中。该 PGML 框架是利用科罗拉多前沿山脉风暴事件中的详细泥石流清单开发和评估的。在无限坡度模型的基础上,首先确定了研究区域的安全系数,随后通过修正的损失函数对 ML 模型的预测进行约束,并测量模型预测的物理一致性。为了评估模型的稳健性和普适性,本研究采用地理样本选择进行模型性能评估,在不同生态区域对 ML 模型进行训练和测试。研究结果表明,基于物理的方法和数据驱动的方法在确定研究区域的滑坡易发性方面都很有效;但是,纯数据驱动的 ML 模型在新的生态区中产生了不切实际的物理结果和较差的泛化性能。PGML 模型加入了物理约束条件,在物理一致性和泛化能力方面都有显著提高,同时降低了各生态区模型的不确定性,其性能超过了基准 ML 模型。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
求助全文
约1分钟内获得全文 求助全文
来源期刊
Acta Geotechnica
Acta Geotechnica ENGINEERING, GEOLOGICAL-
CiteScore
9.90
自引率
17.50%
发文量
297
审稿时长
4 months
期刊介绍: Acta Geotechnica is an international journal devoted to the publication and dissemination of basic and applied research in geoengineering – an interdisciplinary field dealing with geomaterials such as soils and rocks. Coverage emphasizes the interplay between geomechanical models and their engineering applications. The journal presents original research papers on fundamental concepts in geomechanics and their novel applications in geoengineering based on experimental, analytical and/or numerical approaches. The main purpose of the journal is to foster understanding of the fundamental mechanisms behind the phenomena and processes in geomaterials, from kilometer-scale problems as they occur in geoscience, and down to the nano-scale, with their potential impact on geoengineering. The journal strives to report and archive progress in the field in a timely manner, presenting research papers, review articles, short notes and letters to the editors.
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术官方微信