Machine Learning Models of Groundwater Arsenic Spatial Distribution in Bangladesh: Influence of Holocene Sediment Depositional History

Tan, Zhen1,2,3; Yang, Qiang4; Zheng, Yan1,2

2020-08-04

10.1021/acs.est.0c03617

ENVIRONMENTAL SCIENCE & TECHNOLOGY   影响因子和分区

0013-936X

1520-5851

Recent advances in machine learning methods offer the opportunity to improve risk assessment and to decipher factors influencing the spatial variability of groundwater arsenic ([As](gw)). A systematic comparison reveals that boosted regression trees (BRT) and random forest (RF) outperform logistic regression. The probability of [As](gw), exceeding 5 mu g/L (approximate median value of Bangladesh [As](gw)), 10 mu g/L (WHO provisional guideline value), and 50 mu g/L (Bangladesh drinking water standard) is modeled by BRT and RF methods for Bangladesh and its four subregions demarcated by major rivers. Of the 109 geo-environmental and hydrochemical predictor variables, phosphorus and iron emerge as the most important across spatial scales, consistent with known As mobilization mechanisms. Well depth is significant only when hydrochemical parameters are not considered, consistent with prior studies. A peak of probability of [As](gw). exceedance at similar to 30 m depth is evident in the partial dependence plots (PDPs) for spatial-parameter-only models but not in the equivalent all-parameter models, suggesting that sediment depositional history explains interdependent spatial patterns of groundwater As-P-Fe in Holocene aquifers. The South region exhibits a decrease of probability of [As](gw), exceedance below 150 m depth in PDPs for spatial-parameter-only and all-parameter models, supporting that the deeper Pleistocene aquifer is a low-As water resource.

SCI ; EI

英语

NI论文

第一 ; 通讯

National Natural Science Foundation of China[41831279][41772265] ; Strategic Priority Research Program of the Chinese Academy of Sciences[XDA20060402] ; Shenzhen Science and Technology Innovation Commission[KQJSCX20170728163124680]

Engineering ; Environmental Sciences & Ecology

Engineering, Environmental ; Environmental Sciences

WOS:000558753900031

AMER CHEMICAL SOC

20203809188839

Iron compounds ; Potable water ; Decision trees ; Risk assessment ; Aquifers ; Groundwater resources ; Hydrogeology ; Machine learning

Water Resources:444 ; Groundwater:444.2 ; Geology:481.1 ; Artificial Intelligence:723.4 ; Chemical Products Generally:804 ; Accidents and Accident Prevention:914.1 ; Combinatorial Mathematics, Includes Graph Theory, Set Theory:921.4 ; Systems Science:961

ENVIRONMENT/ECOLOGY

Web of Science

1.Southern Univ Sci & Technol, Guangdong Prov Key Lab Soil & Groundwater Pollut, Sch Environm Sci & Engn, Shenzhen 518055, Peoples R China
2.Southern Univ Sci & Technol, State Environm Protect Key Lab Integrated Suiface, Sch Environm Sci & Engn, Shenzhen 518055, Peoples R China
3.Peking Univ, Coll Engn, Beijing 100871, Peoples R China
4.Columbia Univ, Lamont Doherty Earth Observ, Palisades, NY 10964 USA

Tan, Zhen,Yang, Qiang,Zheng, Yan. Machine Learning Models of Groundwater Arsenic Spatial Distribution in Bangladesh: Influence of Holocene Sediment Depositional History[J]. ENVIRONMENTAL SCIENCE & TECHNOLOGY,2020,54(15):9454-9463.

Tan, Zhen,Yang, Qiang,&Zheng, Yan.(2020).Machine Learning Models of Groundwater Arsenic Spatial Distribution in Bangladesh: Influence of Holocene Sediment Depositional History.ENVIRONMENTAL SCIENCE & TECHNOLOGY,54(15),9454-9463.

Tan, Zhen,et al."Machine Learning Models of Groundwater Arsenic Spatial Distribution in Bangladesh: Influence of Holocene Sediment Depositional History".ENVIRONMENTAL SCIENCE & TECHNOLOGY 54.15(2020):9454-9463.