研究报告
王小平,张飞,Abduwasit Ghulam,于海洋,任岩,王娟,张月.艾比湖流域地表水水质指标与水体指数关系研究[J].环境科学学报,2017,37(3):900-909
艾比湖流域地表水水质指标与水体指数关系研究
- The relationship between the surface water quality indices and hydrology of Ebinur Lake watershed
- 基金项目:国家自然科学基金(No.41361045,41130531);新疆联合基金本地优秀青年培养专项(No.U1503302);自治区青年科技创新人才培养工程项目(No.2013731002);新疆绿洲生态(教育部省部共建)重点实验室开放课题(No.XJDX0201-2012-01)
- 王小平
- 1. 新疆大学资源与环境科学学院, 乌鲁木齐 830046;2. 新疆大学绿洲生态教育部重点实验室, 乌鲁木齐 830046
- 张飞
- 1. 新疆大学资源与环境科学学院, 乌鲁木齐 830046;2. 新疆大学绿洲生态教育部重点实验室, 乌鲁木齐 830046;3. 新疆智慧城市与环境建模普通高校重点实验室, 乌鲁木齐 830046
- Abduwasit Ghulam
- 1. 新疆大学资源与环境科学学院, 乌鲁木齐 830046;4. 美国圣路易斯大学可持续发展中心, 圣路易斯 63108
- 于海洋
- 1. 新疆大学资源与环境科学学院, 乌鲁木齐 830046;2. 新疆大学绿洲生态教育部重点实验室, 乌鲁木齐 830046
- 任岩
- 1. 新疆大学资源与环境科学学院, 乌鲁木齐 830046;2. 新疆大学绿洲生态教育部重点实验室, 乌鲁木齐 830046
- 王娟
- 1. 新疆大学资源与环境科学学院, 乌鲁木齐 830046;2. 新疆大学绿洲生态教育部重点实验室, 乌鲁木齐 830046
- 张月
- 1. 新疆大学资源与环境科学学院, 乌鲁木齐 830046;2. 新疆大学绿洲生态教育部重点实验室, 乌鲁木齐 830046
- 摘要:传统的水域水质监测手段不仅成本高,而且空间信息有限,难以对相关水域进行全面的监测与评价.利用遥感技术进行水域水质监测可克服这些局限.本文以新疆艾比湖流域为研究对象,结合2015年5月实测水质数据和从准同步Landsat OLI数据上提取的水体指数值,利用空间分析和多元统计方法进行分析.结果发现,水体指数EWI、AWEIsh、Vegetation index(VI)、NDWI、NWI、NEW与水质指标之间的相关性显著(0.55 ≤ r ≤ 0.88).因此,选用以上6种水体指数与水质指标进行回归分析并建立数学关系估算模型,发现模型均方根误差均较低.利用同期实测数据对估算模型进行精度验证,发现验证判定系数高,验证点的相对平均误差偏低,均方根误差偏小.与此同时,利用2015年10月的40个采样点对模型进行二次验证,发现验证的判定系数满足0.22 < R2 < 0.81,模型的均方根误差较低.在估算模型中,总磷(TP)、五日生化需氧量(BOD5)、悬浮物(SS)、pH、色度和浊度两期数据的验证判定系数R2在0.5以上,均方根误差较小.因此,利用艾比湖流域水体指数建立的TP、BOD5、SS、pH、色度和浊度估算模型具有较好的普适性.该研究不仅可以为干旱区湖泊的遥感识别奠定基础,而且可为遥感技术应用于地表水水质指标值的提取提供一定的科学依据.
- Abstract:The traditional method of monitoring water quality is not only labor intensive, but also costly, and it is difficult to conduct a comprehensive, accurate monitoring and evaluation over large areas. Remote sensing provides effective means to monitor water quality. Combined with the measured water quality data (COD,TDS,EC,TP, BOD5, SS, pH,Chroma and NTU) and Landsat OLI derived water index, this work presents spatial analysis and multivariate statistical analysis method taking Ebinur Lake watershed in Xinjiang, China as the study site. The results show that correlation between the water quality parameters and the spectral indices (EWI,AWEIsh, NDWI,Vegetation index, NEW, and NWI) is significant, and therefore we chose the six spectral indices as water indices and the indicators of water quality for regression analysis. The model accuracy was verified by field measured data, and the accuracy of the model was obtained (0.55 ≤ r ≤ 0.88) with low RMSE. Data collected on October 2015 were used for second test case using 40 samples, and it is found that the coefficient of determination was 0.22 < R2 < 0.81. In estimating model of TP, BOD5, SS, pH, Chroma and NTU data coefficient of determination (R2) was consistently above 0.5, with low root mean square error. This paper concludes that the combined use of these indices provide an effective method to monitor water quality, including TP, BOD5, SS, pH, Chroma and NTU, in the Ebinur Lake watershed. The study can not only provide a basis for the identification of lakes in arid areas, but also provide a scientific basis for the application of remote sensing technology in the extraction of surface water quality.
摘要点击次数: 1460 全文下载次数: 2528