| 基于机器学习算法和受体模型联用的土壤重金属溯源解析 |
| 摘要点击 3215 全文点击 666 投稿时间：2024-08-18 修订日期：2024-09-24 |
| 查看HTML全文
查看全文 查看/发表评论 下载PDF阅读器 |
| 中文关键词 土壤 重金属 APCS-MLR受体模型 机器学习 源解析 |
| 英文关键词 soil heavy metals APCS-MLR receptor model machine learning source apportionment |
| DOI 10.13227/j.hjkx.202408149 |
|
| 中文摘要 |
| 以重庆市煤矸山周边土壤为研究对象,运用决策树（DT）、随机森林（RF）和支持向量机（SVM）等3种机器学习算法开展土壤重金属影响因素分析,并将机器学习算法和绝对因子得分-多元线性回归（APCS-MLR）受体模型联用,开展土壤重金属溯源解析. 土壤表层ω（Cd）、 ω（Hg）、 ω（As）、 ω（Pb）、 ω（Cr）、 ω（Cu）、 ω（Ni）和ω（Zn）均值分别为0.44、0.18、9.92、32.3、129、100、72.8和148 mg·kg-1,结合剖面数据分析,研究区Cd、Hg、As、Pb、Cr、Cu、Ni和Zn含量受不同程度人为活动影响. 机器学习算法表明,随机森林（RF）算法优于决策树（DT）和支持向量机（SVM）算法,Cd、Hg、As、Pb、Cr、Cu、Ni和Zn拟合度（R2）分别为0.783、0.728、0.528、0.753、0.753、0.853、0.822和0.756. 煤矸山堆存量（X1）、土壤点位与煤矸山相对高差（X2）和与煤矸山距离（X3）是影响土壤重金属含量的首要人为因素. 结合受体模型源解析表明,研究区受自然源、矿业源和混合源（大气沉降、农业生产、生活、交通排放等）影响,贡献率分别为42.5%、37.1%和20.4%. 机器学习和受体模型联用,可以使源解析结果更加全面、准确和可靠. |
| 英文摘要 |
| To analyze the source apportionment and influence factors of heavy metals in soils surrounding a coal gangue heap in Chongqing, three machine learning algorithms （decision tree （DT）, random forest （RF）, and support vector machine （SVM）） and the absolute principal component scores-multiple linear regression （APCS-MLR） receptor model were used. The surface soil results showed that the average values of Cd, Hg, As, Pb, Cr, Cu, Ni, and Zn were 0.44, 0.18, 9.92, 32.3, 129, 100, 72.8, and 148 mg·kg-1. Combined profile soil data showed that Cd, Hg, As, Pb, Cr, Cu, Ni, and Zn were affected by human activities to varying degrees. Using machine learning algorithms analysis, RF was better than DT and SVM, and R2 values of Cd, Hg, As, Pb, Cr, Cu, Ni, and Zn were 0.783, 0.728, 0.528, 0.753, 0.753, 0.853, 0.822, and 0.756. “The number of coal gangue units” （X1）, “the vertical height difference between the sampling point and coal gangue heap” （X2）, and “the distance between the sampling point and the coal gangue heap” （X3） were the key driving factors by human activities. Combined with APCS-MLR model analysis, the soil in the study area was affected by natural sources, mining sources, and mixed sources （including atmospheric deposition, agricultural production, life and traffic emissions, etc.）, with contribution rates of 42.5%, 37.1%, and 20.4%, respectively. The combined application of the machine learning algorithms and receptor model can make the results of source apportionment more comprehensive, accurate, and reliable. |
|
|
|
