Search

In a first approximation to the use of hyperspectral DESIS data to ocean colour applications in inland waters, we present here a comparison of the DESIS vs. MSI spectra from two reservoirs located in the province of Valencia. Benagerber and Tous are two small, oligo-mesotrophic reservoirs in need of constant water quality monitoring. These preliminary results have shown us the advantages and limitations of DESIS imagery for these types of dark and small lakes, but also pointed out the need for a specific over water atmospheric correction, as well as further understanding of the band combinations that can lead to accurate water quality parameter estimations.

Transparency or turbidity is one of the main indicators in studies of water quality using remote sensing. Transparency can be measured in situ through the Secchi disc depth (SD), and turbidity using a turbidimeter. In recent decades, different relationships between bands from different remote sensing sensors have been used for the estimation of these variables. In this paper, several indices and spectral bands have been calibrated in order to estimate transparency from Sentinel-2 (S2) images from field data, obtained throughout 2017 and 2018 in Júcar basin reservoirs with a great variety of trophic states. Three atmospheric correction methods developed for waters have been applied to the S2 level L1C images taken at the same day as the field data: Polymer, C2RCC and C2X. From the spectra obtained from S2 and the SD field data, it has been found that the smallest error is obtained with the images atmospherically corrected with Polymer and a potential adjustment of the reflectivities’ ratio of the blue and green bands (R490/R560), which allow the estimation of SD with a relative error of 13%. Also the C2X method presents good adjustment with the same bands ratio, although with a greater error, while the correction C2RCC shows the worst correlation. The relationship between SD (in m) and turbidity (in NTU) has also been obtained, which provides an operational method for estimating turbidity with S2. The relationship for the different reservoirs between SD and chlorophyll-a concentration, suspended solids and dissolved organic matter, is also shown.

Eutrophy in Albufera of Valencia (Eastern Iberian Peninsula) is a quite old problem since after the intense eutrophication processes throughout the 1960s. The system shifted to a turbid stable state consolidated by the virtual disappearance of macrophytes by the early 1970s. The lagoon has been studied extensively since the 1980s, but efforts to revert the system to a clear state have not yielded the expected results because cultural eutrophication due to the growth of population in its area of influence and the effects of climate change, decreasing rainfall and increasing evaporation. This has driven to an increase in water retention times in the lagoon in recent years, resulting in a phytoplanktonic shift towards potentially toxic cyanobacteria species, often forming blooms. Cyanobacterial blooms severely affect water quality for human use, ranging from recreation and fishing to drinking water resources, as indicated in the surveillance protocol of World Health Organization (WHO). The current state of the lake requires constant monitoring and remote sensing is an optimal tool for the continuous monitoring of the whole water mass. This work is included in the ESAQS project (Ecological Status of AQuatic systems with Sentinel satellites), to establish a protocol for regular and frequent monitoring of the ecological status of reservoirs, lakes and lagoons. Algorithms are developed using the images provided by the Sentinel-2 (A and B), provided with a spatial resolution of 10 m and a temporal frequency of 5 days. In this work we demonstrate that using this new earth observation satellite is possible to develop a consistent and suitable algorithm to estimate the phycocyanin concentration [PC] and establish a protocol for regular and frequent monitoring. Calibrating (R 2  = 0.841; n = 21; p < 0.001) and validating (R 2  = 0.775; n = 55; p < 0.001; RMSE% = 40) the algorithm with field data are also demonstrated., (Copyright © 2019. Published by Elsevier B.V.)