Monitoring dissolved organic carbon by combining Landsat-8 and Sentinel-2 satellites: Case study in Saginaw River estuary, Lake Huron

• Published in: The Science of the total environment Vol. 718; p. 137374
• Main Authors: Chen, Jiang, Zhu, Weining, Tian, Yong Q., Yu, Qian
• Format: Journal Article
• Language: English
• Published: Netherlands Elsevier B.V 20.05.2020
• Subjects: Dissolved organic carbon (DOC); Lake Huron; Landsat-8; Remote sensing; Sentinel-2; Dissolved organic carbon (DOC); Landsat-8; Lake Huron; Remote sensing; Sentinel-2
• ISSN: 0048-9697; 1879-1026
• DOI: 10.1016/j.scitotenv.2020.137374

Dissolved organic carbon (DOC) in aquatic environments is an important cycled pool of organic matter on the Earth. Satellite remote sensing provides a useful tool to determine spatiotemporal distribution of water quality parameters. Previous DOC remote sensing studies in inland water suffered from either low spatial resolution or low temporal frequency. In this study, we evaluated the potential of jointly using Landsat-8 and Sentinel-2 with high spatial resolution to estimate DOC concentrations in Saginaw River plume regions of Lake Huron. Firstly, CDOM (colored dissolved organic matter) was estimated from images using the known models and then DOC can be derived in terms of the good correlations between DOC and CDOM. The results show that Landsat-8 and Sentinel-2 had acceptable accuracy and good consistency in DOC estimation so that jointly using them can improve the observation frequency. In different seasons from 2013 to 2018, DOC was typically higher in spring and autumn but lower in summer. Monthly spatiotemporal variations of DOC in 2018 were also observed. The image-derived DOC spatiotemporal variations show that DOC was covaried with Saginaw River discharge (r = 0.82) and also weakly and negatively correlated with water temperature (r = −0.6). This study demonstrated that using Landsat-8 and Sentinel-2 together can offer the potential applications for monitoring DOC and water quality dynamic in complex inland water. [Display omitted] •Landsat-8 and Sentinel-2 DOC models showed acceptable accuracy.•Landsat-8 and Sentinel-2 DOC estimates were consistent within 4.9%.•Monthly spatiotemporal variations of DOC in Lake Huron were observed.•DOC concentrations were significantly influenced by the riverine discharge.