Edish University of Agricultural Science (SLU) Milj erate ecoregions, which may possibly present a diverse selection of prospective wat data MVM Environmental database. Samples had been chosen in these ecoregions as they Tasisulam Apoptosis graphic clustering information sources for lake water excellent parameters. These had frequentl provided consistent open of information happens as only particular ecoregions databases also helped give a geographic spread of information from the chl-a and turbidity have been taken ter top quality outcomes. Only samples where both tropics to northern temperate ecoregions, which could provide a diverse range of possible water sorts. Geographic of a Landsat 4, 5, as only satellite overpasses had been reported water top quality clustering of data occurs7, or 8 distinct ecoregions had frequentlyselected. This window s to permit for an sufficient quantity turbidity had been in between samples and final results. Only samples where both chl-a andof matchups taken inside days of a satel Landsat four, 5, 7, or 8 satelliterelationship with measured reflectance chosenLimited whilst maintaining a overpasses have been chosen. This window size was . to allow for an sufficient number of matchups amongst samples and satellite overpasses while oured dissolved organic matter and total suspended solids metrics had been fou preserving a partnership with measured reflectance . Limited samples of coloured diswindow and consequently suspended solids within this study. A total of window solved organic matter and totalwere not utilised metrics have been identified inside this 204 sample p and therefore have been not used in this study. A totalS1). Lake sizes ranged from five.three to 86,66 lakes have been selected (Figure 1, Table of 204 sample pairs inside 142 lakes were chosen (Figure 1, Table S1). Lake sizes ranged from 5.three to 86,661.9 ha (median = 119.three ha). = 119.3 ha). As a result of a lack of available metadata for public information records, Resulting from a lack of readily available metadata for public information records, variations in ground-based ground-based measurement processing as well as a source of possible error in measurement processing and calibration will occur and offercalibration will happen and from the remote sensing retrieval. remote sensing retrieval. possible error in theFigure 1. Places of ground-based chl-a and turbidity Figure 1. Areas of ground-based chl-a and turbidity samples.samples.Remote Sens. 2021, 13,four of2.2. Landsat Image Acquisition, Processing, and Evaluation Sample locations have been mapped for the Worldwide Reference Program (WRS-2) Landsat catalogue technique to recognize the (longitudinal) paths and (latitudinal) rows in which the samples had been located. A total of 105 pairs of Landsat Level-1 and -2 images with ten cloud coverage and inside days of sample dates had been downloaded from the USGS EarthExplorer information catalogue (https://earthexplorer.usgs.gov/, last accessed: 3 November 2021) (72 Landsat 4-5 TM, 11 Landsat 7 ETM (SLC-on), and 22 Landsat eight OLI) (Table S1). Different atmospheric correction selections are obtainable for the remote sensing of water top quality applying Landsat information (e.g., 6S, DOS, Expense, iCOR); even so, such solutions normally result in errors resulting from the violation from the dark pixel assumption in turbid waters when estimating aerosol optical thickness in the N [51,52]. Whilst the SWIR band could be utilised in lieu with the N, it frequently outcomes in reduce aerosol accuracy estimation on account of a poorer signal oise ratio . Some research have IQP-0528 Purity & Documentation instead opted for straightforward atmospheric correction of Rayleigh scatter (and not of aerosol contributions) for chl-a retrieval in turbid wate.