基于BP神经网络的粤北某地区地下水脆弱性评价及其风险管控

张涛; 王夏晖; 毕二平; 黄国鑫; 杨瑞杰

doi:10.13205/j.hjgc.202312034

基于BP神经网络的粤北某地区地下水脆弱性评价及其风险管控

doi: 10.13205/j.hjgc.202312034

张涛^1,2,
王夏晖²,
毕二平³,
黄国鑫^2, ,,
杨瑞杰⁴

1. 山东省环科院环境检测有限公司, 济南 250013;
2. 生态环境部环境规划院, 北京 100041;
3. 中国地质大学(北京) 水资源与环境学院, 北京 100083;
4. 韶关市环境污染控制中心, 广东韶关 512026

基金项目:

国家重点研发计划项目(2018YFC1800205)

详细信息

作者简介:
张涛(1997-),男,硕士,主要研究方向为地下水污染防治。847403810@qq.com

通讯作者:
黄国鑫(1980-),男,研究员,主要研究方向为土壤和地下水污染防治。huanggx@caep.org.cn

计量
- 文章访问数: 69
- HTML全文浏览量: 11
- PDF下载量: 2
- 被引次数: 0
出版历程
- 收稿日期: 2022-03-01
- 网络出版日期: 2024-03-08

GROUNDWATER VULNERABILITY EVALUATION AND RISK CONTROL IN A CERTAIN AREA IN NORTHERN GUANGDONG PROVINCE BASED ON BP NEURAL NETWORK

1. Shandong Huankeyuan Environmental Testing Co., Ltd., Jinan 250013, China;
2. Chinese Academy of Environmental Planning, Beijing 100041, China;
3. School of Water Resources and Environment, China University of Geosciences (Beijing), Beijing 100083, China;
4. Shaoguan Environmental Pollution Control Center, Shaoguan 512026, China

摘要

摘要: 针对传统DRASTIC模型在参数权重确定过程中主观性强问题,以粤北某地区浅层地下水为研究对象,利用采集的地下水相关数据和新增土地利用类型参数,在优化BP神经网络算法和构建DRASTICL模型基础上,借助地下水NO₃^-浓度进行模型验证,建立耦合BP神经网络算法的BP-DRASTICL模型,进而根据地下水脆弱性空间分布特点提出了地下水污染风险管控对策。结果表明:训练函数为trainlm、学习率为0.1、隐含层神经元节点数为6时,BP神经网络算法效果最好,相应地获得的最优BP-DRASTICL模型参数权重依次为0.1420(地下水埋深,D)、0.1151(净补给量,R)、0.0791(含水层介质,A)、0.1833(土壤介质,S)、0.0908(地形,T)、0.1574(包气带介质影响,I)、0.0891(渗透系数,C)和0.1433(土地利用类型,L)。D、S、T和L对评价结果的影响最大。与DRASTIC模型、DRASTICL模型相比,BP-DRASTICL模型的Pearson和Spearman相关系数最高,分别达到0.615和0.656,表明硝酸盐浓度与脆弱性指数之间具有很高的线性关系。研究区地下水脆弱性总体属于极低脆弱性和低脆弱性,高脆弱性和极高脆弱性地区主要分布在研究区中部。针对研究区脆弱性的空间分布特征,差异化提出了地下水污染风险管控对策。利用BP神经网络算法确定DRASTICL模型参数权重,因减少了人为主观性的影响,比传统专家打分法更准确。
- DRASTIC模型 /
- 地下水脆弱性 /
- 机器学习 /
- 参数权重
Abstract: Aiming at the high subjectivity of the traditional DRASTIC model in the process of the parameter weight determination, taking shallow groundwater in a certain area in northern Guangdong Province as a study target, the BP neural network was optimized and the DRASTICL model was constructed by using the collected shallow groundwater-related data and adding land use type parameter. On this basis, groundwater NO₃^- concentration was used to verify the models, and further the BP neural network and the constructed DRASTICL model were coupled to establish a BP-DRASTICL model. Finally, risk control strategies were suggested according to the spatial distribution characteristics of groundwater vulnerability. The results showed that when the training function was trainlm, the learning rate was 0.1, and the number of hidden layer neurons was 6, BP neural network performed best, and accordingly the optimal BP-DRASTICL parameter weights were 0.1420 (groundwater depth, D), 0.1151 (recharge, R), 0.0791 (aquifer media, A), 0.1833 (soil medium, S), 0.0908 (topography, T), 0.1574 (influence of vadose zone media, I), 0.0891 (hydraulic conductivity, C) and 0.1433 (land use type, L). D, S, T and L had the greatest influence on the evaluation results. Compared with the DRASTIC models and the DRASTICL model, the BP-DRASTICL model had the highest Pearson (0.615) and Spearman (0.656) correlation coefficients, indicating a high linear correlation between the actual nitrate concentration and the vulnerability index. The groundwater vulnerability was generally in the extremely low, and low level, across the study area, and the areas with the high and extremely high vulnerability level were mainly distributed in the middle of the study area. According to the spatial distribution characteristics of the vulnerability, differentiated strategies were proposed for groundwater pollution risk control. Using the BP neural network to determine the parameter weights of the DRASTICL model is more accurate than using the traditional expert scoring method, because it reduces the influence of human subjectivity.
- DRASTIC model /
- groundwater vulnerability /
- machine learning /
- parameters weighting

HTML全文

参考文献(28)

[1]	李怀恩, 贾斌凯, 成波, 等. 海绵城市雨水径流集中入渗对土壤和地下水影响研究进展[J]. 水科学进展, 2019, 30(4):589-600.
[2]	刘乃静, 赵银鑫, 吴文忠, 等. 基于AHP的DRASTIC模型在银川市潜水防污性能评价中的应用[J]. 水土保持通报, 2021, 41(1):212-218.
[3]	王昕翌, 王敏, 冯建国, 等. 地下水水源地保护区划分方法研究:以大汶口为例[J]. 人民长江, 2021, 52(4):61-67.
[4]	AKINLALU A A, MOGAJI K A, ADEBODUM T S. Assessment of aquifer vulnerability using a developed "GODL" method (modifed GOD model) in a schist belt environ, Southwestern Nigeria[J]. Environmental Monitoring and Assessment, 2021, 193(4):199-226.
[5]	ZHANG L Z, ZHANG Y B, ZHOU X Y, et al. Groundwater vulnerability assessment system based on GIS[J]. Advanced Materials Research, 2011, 301(303):724-730.
[6]	吕文凯, 周金星, 万龙, 等. 滇东岩溶断陷盆地水资源脆弱性评价[J]. 地球学报, 2021, 42(3):341-351.
[7]	张钧帅, 汪丙国, 刘天奇. 江汉平原浅层地下水防污性能模糊综合评价与验证[J]. 地质科技通报, 2020, 39(6):154-164.
[8]	MFONKA Z, NGOUPAYOU J R N, NDJIGUI P D, et al. A GIS-based DRASTIC and GOD models for assessing alterites aquifer of three experimental watersheds in Foumban (Western-Cameroon)[J]. Groundwater for Sustainable Development, 2018, 7:250-264.
[9]	于林弘, 陶志斌, 扈胜涛, 等. DRASTIC模型在地下水脆弱性评价中的应用[J]. 人民黄河, 2020, 42(增刊1):45-46, 50.
[10]	杨宁, 陶志斌, 高松, 等. 基于AHP的DRASTIC模型对莱州地区地下水脆弱性研究[J]. 地质学报, 2019, 93(增刊1):133-137.
[11]	朱飞, 熊丽君, 吴建强, 等. 基于改进DRASTIC模型的平原河网地区地下水脆弱性评价[J]. 环境科学与技术, 2020, 43(2):187-193.
[12]	王建红, 余启明, 李平平. 基于GIS技术与DRASTIC模型的民勤盆地地下水脆弱性评价[J]. 兰州大学学报(自然科学版), 2015, 51(6):882-887.
[13]	ASLAM B, ISMAIL S, ALI I. A GIS-based DRASTIC model for assessing aquifer susceptibility of Safdarabad Tehsil, Sheikhupura District, Punjab Province, Pakistan[J]. Modeling Earth Systems and Environment, 2020, 6:995-1005.
[14]	李亮, 王敏, 邢怀学, 等. 基于AHP-DRASTIC评价模型的无锡市潜水污染风险评价[J]. 合肥工业大学学报(自然科学版), 2019, 42(3):310-314.
[15]	周书葵, 江海浩, 陈朝猛, 等. 改进AHP-DRASTIC模型用于地下水U(Ⅵ)污染风险评价及回归分析[J]. 环境工程, 2016, 34(1):130-134.
[16]	YANG J, TANG Z H, JIAO T, et al. Combining AHP and genetic algorithms approaches to modify DRASTIC model to assess groundwater vulnerability: a case study from Jianghan Plain, China[J]. Environmental Earth Science, 2017, 76(12):426-442.
[17]	贺新春, 邵东国, 陈南祥, 等. 几种评价地下水环境脆弱性方法之比较[J]. 长江科学院院报, 2005, 22(3):17-20 , 24.
[18]	徐超, 周嘉月, 何旭佳, 等. 基于改进DRASTIC模型的陕西省地下水脆弱性评价[J]. 中国农村水利水电, 2020(3):44-51.
[19]	任加国, 龚克, 马福俊, 等. 基于BP神经网络的污染场地土壤重金属和PAHs含量预测[J]. 环境科学研究, 2021, 34(9):2237-2247.
[20]	KUANG Y T, SINGH R, SINGH S, et al. A novel macroeconomic forecasting model based on revised multimedia assisted BP neural network model and ant Colony algorithm[J]. Multimedia Tools and Applications, 2017, 76(18):18749-18770.
[21]	张天云, 陈奎, 魏伟, 等. BP神经网络法确定工程材料评价指标的权重[J]. 材料导报, 2012, 26(1):159-163.
[22]	赵军, 张祯宇, 谢哲宇, 等. 基于BP人工神经网络的闽江口水厂水质模拟[J]. 环境科学与技术, 2020, 43(增刊1):198-203.
[23]	黄元生, 张利君. 基于遗传算法的BP-LSSVM组合变权模型权重优化的短期电价预测研究[J]. 煤炭工程, 2019, 51(5):172-176.
[24]	孙会君, 王新华. 应用人工神经网络确定评价指标的权重[J]. 山东科技大学学报(自然科学版), 2001, 20(3):84-86.
[25]	钟佐燊. 地下水防污性能评价方法探讨[J]. 地学前缘, 2005, 12(增刊1):3-11.
[26]	IIAMURUGAN O, JOTHIBASU A, ANBAZHAGAN S. Geospatial technology and modified DRASTIC model to assess the groundwater pollution vulnerability along a stretch of Cauvery River, South India[J]. Environmental Earth Science, 2022, 81:85.
[27]	YU H, WU Q, ZENG Y F, et al. Integrated variable weight model and improved DRASTIC model for groundwater vulnerability assessment in a shallow porous aquifer[J]. Journal of Hydrology, 2022, 608:127538.
[28]	陈红兵, 孙俊辉, 王聪聪, 等. 应用遗传算法优化BP神经网络预测太阳能PV/T系统热电产出[J]. 热科学与技术, 2021, 20(5):480-487.