Abstract
A multi-sensor and multi-scale monitoring tool for the spatially explicit and periodic monitoring of eutrophication in a small drinking water reservoir is presented. The tool was built with freely available satellite and in situ data combined with Unmanned Aerial Vehicle (UAV)-based technology. The goal is to evaluate the performance of a multi-platform approach for the trophic state monitoring with images obtained with MultiSpectral Sensors on board satellites Sentinel 2 (S2A and S2B), Landsat 8 (L8) and UAV. We assessed the performance of three different sensors (MultiSpectral Instrument (MSI), Operational Land Imager (OLI) and Rededge Micasense) for retrieving the pigment chlorophyll-a (chl-a), as a quantitative descriptor of phytoplankton biomass and trophic level. The study was conducted in a waterbody affected by cyanobacterial blooms, one of the most important eutrophication-derived risks for human health. Different empirical models and band indices were evaluated. Spectral band combinations using red and near-infrared (NIR) bands were the most suitable for retrieving chl-a concentration (especially 2 band algorithm (2BDA), the Surface Algal Bloom Index (SABI) and 3 band algorithm (3BDA)) even though blue and green bands were useful to classify UAV images into two chl-a ranges. The results show a moderately good agreement among the three sensors at different spatial resolutions (10 m., 30 m. and 8 cm.), indicating a high potential for the development of a multi-platform and multi-sensor approach for the eutrophication monitoring of small reservoirs.
Highlights
One of the main problems currently affecting water resources is the eutrophication of water bodies, both coastal and freshwater [1,2]; it was already known ten years ago that around 20% of European lakes were suffering from nutrient enrichment [3]
(2) To validate the application of free-distributed multispectral images from Sentinel 2 and Landsat 8 satellites to improve the continuous monitoring of small reservoirs, in an area of very high cloud cover (Galicia, NW Iberian Peninsula)
The studied reservoir is an eutrophic waterbody with recurrent cyanobacterial blooms since 2012, documented through the Galician Platform for Environmental Information (GAIA) (Table 1)
Summary
One of the main problems currently affecting water resources is the eutrophication of water bodies, both coastal and freshwater [1,2]; it was already known ten years ago that around 20% of European lakes were suffering from nutrient enrichment [3]. Most of the reservoir monitoring programs are currently based on measurements taken in the field making it difficult to capture the spatial and temporal variability of phenomena such as algal blooms, with irregular spatiotemporal patterns and potential toxicity (Cyanobacterial blooms) This circumstance hampers the possibility of taking timely and informed management decisions to prevent, counteract or mitigate the negative effects of eutrophication in drinking water reservoirs where “early warning” systems are increasingly needed. Empirical algorithms based on green and blue bands have been successfully applied to open oceans (Case 1 waters [18]) and oligotrophic inland waters [19,20], while the near-infrared (NIR) to red reflectance ratios are used in productive waters (Case 2) In the latter, chl-a is not correlated with the other optically active constituents, and the absorption by Coloured Dissolved Organic Matter (CDOM), tripton, and other pigments overlaps the absorption peak of chl-a in the blue spectral region. The use of combined algorithms has been suggested for the monitoring of these different water types and trophic states [24]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
