Applications of geographically weighted machine learning models for predicting soil heavy metal concentrations across mining sites.

IF 8.2 1区 环境科学与生态学 Q1 ENVIRONMENTAL SCIENCES
Science of the Total Environment Pub Date : 2024-12-20 Epub Date: 2024-11-27 DOI:10.1016/j.scitotenv.2024.177667
Hyemin Jeong, Younghun Lee, Byeongwon Lee, Euisoo Jung, Jai-Young Lee, Sangchul Lee
{"title":"Applications of geographically weighted machine learning models for predicting soil heavy metal concentrations across mining sites.","authors":"Hyemin Jeong, Younghun Lee, Byeongwon Lee, Euisoo Jung, Jai-Young Lee, Sangchul Lee","doi":"10.1016/j.scitotenv.2024.177667","DOIUrl":null,"url":null,"abstract":"<p><p>The accurate prediction of soil heavy metal contamination is crucial for the effective environmental management of abandoned mining areas. However, conventional machine learning models (CMLMs) often fail to account for the spatial heterogeneity of soil contamination, which limits their predictive accuracy. This study evaluated the performance of geographically weighted machine learning models (GWMLMs) in predicting soil Cd and Pb concentrations in abandoned mines in the Republic of Korea. We compared two GWMLMs (Geographically Weighted Random Forest and Geographically Weighted Extreme Gradient Boosting) with four CMLMs (Random Forest, Gradient Boosting, Light Gradient Boosting, and extreme Gradient Boosting). The data used in this study included soil samples from six abandoned mining sites with various geographical and soil input variables. The results showed that the GWMLMs consistently outperformed the CMLMs in predicting heavy metal contamination. For Cd predictions, GWMLMs exhibited on average 0.02 lower root mean square error and mean absolute error values, with a 0.26 increase in R<sup>2</sup> values compared to CMLMs. Similarly, for Pb predictions, the GWMLMs showed 0.18 and 0.13 lower root mean square error and mean absolute error values, respectively, and a 0.17 increase in R<sup>2</sup> relative to the CMLMs. The findings demonstrate the usefulness of GWMLMs for predicting the spatial distribution of soil heavy metals. SHapley Additive exPlanations analysis exhibited elevation and distance from abandoned mining sites as the most influential factors in predicting both Cd and Pb concentrations. This study highlights the value of GWMLMs that incorporate spatial heterogeneity into CMLMs for enhancing prediction accuracy and providing crucial insights for environmental management in mining-impacted regions.</p>","PeriodicalId":422,"journal":{"name":"Science of the Total Environment","volume":" ","pages":"177667"},"PeriodicalIF":8.2000,"publicationDate":"2024-12-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Science of the Total Environment","FirstCategoryId":"93","ListUrlMain":"https://doi.org/10.1016/j.scitotenv.2024.177667","RegionNum":1,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"2024/11/27 0:00:00","PubModel":"Epub","JCR":"Q1","JCRName":"ENVIRONMENTAL SCIENCES","Score":null,"Total":0}
引用次数: 0

Abstract

The accurate prediction of soil heavy metal contamination is crucial for the effective environmental management of abandoned mining areas. However, conventional machine learning models (CMLMs) often fail to account for the spatial heterogeneity of soil contamination, which limits their predictive accuracy. This study evaluated the performance of geographically weighted machine learning models (GWMLMs) in predicting soil Cd and Pb concentrations in abandoned mines in the Republic of Korea. We compared two GWMLMs (Geographically Weighted Random Forest and Geographically Weighted Extreme Gradient Boosting) with four CMLMs (Random Forest, Gradient Boosting, Light Gradient Boosting, and extreme Gradient Boosting). The data used in this study included soil samples from six abandoned mining sites with various geographical and soil input variables. The results showed that the GWMLMs consistently outperformed the CMLMs in predicting heavy metal contamination. For Cd predictions, GWMLMs exhibited on average 0.02 lower root mean square error and mean absolute error values, with a 0.26 increase in R2 values compared to CMLMs. Similarly, for Pb predictions, the GWMLMs showed 0.18 and 0.13 lower root mean square error and mean absolute error values, respectively, and a 0.17 increase in R2 relative to the CMLMs. The findings demonstrate the usefulness of GWMLMs for predicting the spatial distribution of soil heavy metals. SHapley Additive exPlanations analysis exhibited elevation and distance from abandoned mining sites as the most influential factors in predicting both Cd and Pb concentrations. This study highlights the value of GWMLMs that incorporate spatial heterogeneity into CMLMs for enhancing prediction accuracy and providing crucial insights for environmental management in mining-impacted regions.

应用地理加权机器学习模型预测矿区土壤重金属浓度。
准确预测土壤重金属污染对废弃矿区的有效环境管理至关重要。然而,传统的机器学习模型(CMLMs)往往无法考虑土壤污染的空间异质性,从而限制了其预测精度。本研究评估了地理加权机器学习模型(GWMLMs)在预测大韩民国废弃矿区土壤镉和铅浓度方面的性能。我们比较了两种 GWMLM(地理加权随机森林和地理加权极端梯度提升)和四种 CMLM(随机森林、梯度提升、轻度梯度提升和极端梯度提升)。本研究使用的数据包括来自六个废弃矿区的土壤样本,这些样本具有不同的地理和土壤输入变量。结果表明,GWMLM 在预测重金属污染方面的表现始终优于 CMLM。在镉预测方面,与 CMLMs 相比,GWMLMs 的均方根误差和平均绝对误差平均降低了 0.02,R2 值增加了 0.26。同样,在铅预测方面,GWMLMs 的均方根误差和平均绝对误差值分别比 CMLMs 低 0.18 和 0.13,R2 值增加了 0.17。研究结果表明,GWMLMs 可用于预测土壤重金属的空间分布。SHapley Additive exPlanations 分析表明,海拔高度和与废弃矿址的距离是预测镉和铅浓度的最大影响因素。本研究强调了将空间异质性纳入 CMLMs 的 GWMLMs 在提高预测准确性方面的价值,并为受采矿影响地区的环境管理提供了重要见解。
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
Science of the Total Environment
Science of the Total Environment 环境科学-环境科学
CiteScore
17.60
自引率
10.20%
发文量
8726
审稿时长
2.4 months
期刊介绍: The Science of the Total Environment is an international journal dedicated to scientific research on the environment and its interaction with humanity. It covers a wide range of disciplines and seeks to publish innovative, hypothesis-driven, and impactful research that explores the entire environment, including the atmosphere, lithosphere, hydrosphere, biosphere, and anthroposphere. The journal's updated Aims & Scope emphasizes the importance of interdisciplinary environmental research with broad impact. Priority is given to studies that advance fundamental understanding and explore the interconnectedness of multiple environmental spheres. Field studies are preferred, while laboratory experiments must demonstrate significant methodological advancements or mechanistic insights with direct relevance to the environment.
×
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