**2. Materials and Methods**

#### *2.1. Satellite Imagery*

The Sentinel-2A/B twin mission was used for mapping the lagoon thanks to the open data access policy and high spatial resolution (10-20-60 m). The European Commission and the European Space Agency (ESA), in the frame of the Copernicus programme, developed this optical constellation in order to support its operational requirements. Sentinel-2 is a multispectral, wide-swath imaging platform used for monitoring land surfaces, water coverage, soil, and vegetation. In addition, it can also support Copernicus water monitoring over coastal regions and inland waterways. The Sentinel-2A and Sentinel-2B mission, with a global revisit frequency of five days at the Equator, is based on a constellation of two operationally identical satellites in the same orbit and phased at 180◦ to each other. The ESA User Handbook describes the temporal, spectral, spatial, and radiometric features of the visible and near-infrared (NIR) bands of both Sentinel-2A and -2B satellites [38]. The stated quality standards for absolute geolocation of the Sentinel-2 scenes (two pixels, 20 m) are within the ESA requirements [39]. The images covering Mar Menor (zone 30 and tile SXG; acquisition time 11:00 UTC) during the study period in 2021 were downloaded from the ONDA DIAS (https://www.onda-dias.eu/cms/es/; accessed on 20 May 2021). These products are the top-of-atmosphere (TOA) datasets at Level-1C (L1C) after the radiometric and geometric corrections.

In addition, the freely available Landsat-8 visible and NIR spectra imagery from the United States Geological Survey (USGS) and the National Aeronautics and Space Administration (NASA) were also used for comprehensive monitoring. We downloaded the Level 1 data from the Earth Explorer (https://earthexplorer.usgs.gov/), orthorectified and terrain corrected at a 30 m spatial resolution, with a 16-day revisit frequency [40,41]. The region of interest was covered by the tiles located in paths 198–199 and row 34 (acquisition time 10:30–10:45 UTC). A low cloud coverage (<40%) filtering was applied over Mar Menor for further analysis of the scenes. When Sentinel-2 and Landsat-8 products are combined, the average revisit time in Mar Menor is ~4 days. From the control period before and after the 2021 crisis, 48 images were downloaded and processed (29 and 19 images for Sentinel-2 and Landsat-8, respectively). However, clouds and severe sunglint contamination diminished the number of usable images, with only 18 final scenes further evaluated (12 and 6 images for Sentinel-2 and Landsat-8, respectively) to characterize the spatial and temporal distribution of water quality. Table 1 shows the acquisition dates and the quality of the scenes.

Bottom-of-atmosphere (BOA) Level-2A products were generated with one of the most commonly used atmospheric correction softwares (ACOLITE, version 20210114.0), which supports preprocessing of Landsat-8 and Sentinel-2 satellites. This software incorporates an image-based model, without the need for in situ atmospheric datasets. The Royal Belgian Institute of Natural Sciences (RBINS) developed this free toolbox to correct Level-1 to Level-2 data products over marine, inland, and coastal waters [42]. The Dark Spectrum Fitting (DSF) atmospheric correction algorithm was applied [43,44]. The notably enhanced spectral resolution of Landsat-8 and, in particular, Sentinel-2 satellites, are key to obtain goodquality products by means of the DSF model [44]. Correction of the sunglint over the surface reflectance was performed by means of the additional image-based sunglint correction, since during the study period acute sunglint effects were observed at these latitudes (specific setting parameters: dsf\_path\_reflectance = tiled, l2w\_mask\_threshold = 0.05). The remote sensing reflectance (Rrs, sr<sup>−</sup>1) products along the visible and NIR spectrum were calculated after resampling to 10 m and 30 m pixel size for Sentinel-2 and Landsat-8, respectively.


**Table 1.** List of imagery used in this study during the latest ecological crisis in 2021 corresponding to the Sentinel-2 and Landsat-8 satellites.

The standard products to monitor the biogeochemical conditions in Mar Menor during the ecological crisis with Sentinel-2 and Landsat-8 satellites were seawater turbidity (FNU) and chlorophyll-a (chl-a, mg/m3). We selected both indicators as required by the EU Water Framework Directive (WFD) for the evaluation of the good ecological status of the coastal lagoon. The Nechad et al. semianalytical algorithm (red band, Rrs 665 nm) was applied to estimate turbidity with both satellites [45]. This model has already been validated in different environments [46–48] and was previously used in Mar Menor during an extreme weather event in September 2019 [13]. These semianalytical algorithms allow a more global performance since they are based on the inherent optical properties of the seawater. The commonly used OC3 algorithm was applied to calculate the concentration of seawater chl-a [49]. The standard masking procedures were accomplished, eliminating clouds, cloud shadows, land, and the low performance of the sunglint and atmospheric corrections. Turbidity and chl-a maps were at 30 m and 10 m spatial resolution for Landsat-8 and Sentinel-2, respectively, with the generation of the final products after 3–4 h following image acquisition for each sensor.
