| 达里诺尔湖溶解性有机质荧光特征与人工神经网络非线性响应研究 |
| |
| 引用本文： | 孙伟,夏瑞,王晓,王璐,陈焰,马淑芹,张远,胡泓.达里诺尔湖溶解性有机质荧光特征与人工神经网络非线性响应研究[J].环境科学研究,2020,33(11):2458-2466. |
| |
| 作者姓名： | 孙伟 夏瑞 王晓 王璐 陈焰 马淑芹 张远 胡泓 |
| |
| 作者单位： | 1.中国环境科学研究院, 环境基准与风险评估国家重点实验室, 北京 100012 |
| |
| 基金项目： | 国家水体污染控制与治理科技重大专项（No.2017ZX07301-001）；长江生态环境保护修复联合研究项目（第一期）（No.2019-LHYJ-01-0103，2019-LHYJ-01-0102） |
| |
| 摘 要： | 研究湖泊水体DOM（溶解性有机质）特征及其与复合环境要素的响应关系，对水生态系统保护及生物地球化学循环研究具有重要意义.以内蒙古达里诺尔湖（简称“达里湖”）为研究区，利用紫外-可见光谱技术（UV-vis）和人工神经网络模型（ANN）方法，开展达里湖水体DOM特征识别及其水质响应关系的研究.结果表明:①达里湖水环境污染程度较为严重，水体呈富营养化趋势.DOM吸收系数〔α（355）〕表明，夏季湖体DOM浓度较高.②建立了α（355）与达里湖水体ρ（TN）、ρ（TP）、pH和ρ（Chla）的人工神经网络模型，该模型验证期和测试期决定系数（R2）分别为0.78和0.84，该模型具有较好的DOM模拟和预测能力.③人工神经网络模型参数敏感性分析结果表明，α（355）对ρ（Chla）变化最敏感，敏感性水平为31.95%；其次为pH和ρ（TN），α（355）对二者变化的敏感性水平分别为28.53%和25.54%；α（355）对ρ（TP）变化敏感性最低，敏感性水平为8.16%.研究显示，达里湖DOM对富营养化的发生具有较显著的表征影响，人工神经网络模型可为富营养化预测提供科学参考.
|
| 关 键 词： | 达里诺尔湖 溶解性有机质(DOM) 人工神经网络模型 响应分析 |
| 收稿时间： | 2020/7/4 0:00:00 |
| 修稿时间： | 2020/9/22 0:00:00 |
DOM Fluorescence Characteristics and Nonlinear Response of Artificial Neural Network in Dali-Nor Lake |
| |
| SUN Wei,XIA Rui,WANG Xiao,WANG Lu,CHEN Yan,MA Shuqin,ZHANG Yuan,HU Hong.DOM Fluorescence Characteristics and Nonlinear Response of Artificial Neural Network in Dali-Nor Lake[J].Research of Environmental Sciences,2020,33(11):2458-2466. |
| |
| Authors: | SUN Wei XIA Rui WANG Xiao WANG Lu CHEN Yan MA Shuqin ZHANG Yuan HU Hong |
| |
| Affiliation: | 1.State Key Laboratory of Environmental Criteria and Risk Assessment, Chinese Research Academy of Environmental Sciences, Beijing 100012, China2.College of Environmental Science and Engineering, Ocean University of China, Qingdao 266100, China |
| |
| Abstract: | Studying the relationship between the dissolved organic matter (DOM) characteristics of lakes and the complex environmental elements for DOM is of great significance for aquatic ecosystem protection and biogeochemical cycles. This article took Dali-Nor Lake (referred to as 'Dali Lake'), an important migratory bird distribution center in northern China, as a research area, and used modern acquisition methods such as artificial intelligence unmanned ships, and ultraviolet-visible spectroscopy (UV-vis) and artificial neural network model (ANN) methods to study the response relationship between DOM feature recognition and water quality in the lake water. The results showed that: (1) The water environment of Dali Lake was seriously polluted, and the water body was eutrophication. The DOM absorption coefficient showed that the DOM concentration in the lake was higher in summer. (2) The nonlinear response relationship between α(355) and ρ(TN), ρ(TP), pH and ρ(Chla) was established. The coefficient of determination (R2) during the validation period and the testing period was 0.78 and 0.84, respectively, and the model had good DOM simulation and prediction capabilities. (3) The parameter sensitivity analysis results showed that α(355) was the most sensitive to changes in ρ(Chla), with a sensitivity level of 31.95%; the sensitity levels of α(355) to pH was 28.53%; and the sensitity levels of α(355) to ρ(TN) was 25.54%; For ρ(TP), the change sensitivity was the lowest, with a sensitivity level of 8.16%. This study has shown that Dali Lake DOM has a significant effect on the occurrence of eutrophication, which can provide a scientific reference for the prediction of eutrophication. The artificial neural network model can provide a scientific reference for eutrophication prediction. |
| |
| Keywords: | Dali-Nor Lake dissolved organic matter (DOM) artificial neural network response relationship |
| 本文献已被 万方数据 等数据库收录！ |
| 点击此处可从《环境科学研究》浏览原始摘要信息 |
|
点击此处可从《环境科学研究》下载全文 |
|
