Recipes for the Derivation of Water Quality Parameters Using the High-Spatial-Resolution Data from Sensors on Board Sentinel-2A, Sentinel-2B, Landsat-5, Landsat-7, Landsat-8, and Landsat-9 Satellites

Abstract

Satellites have provided high-resolution ( < 100 m) water color (i.e., remote sensing reflectance) and thermal emission imagery of aquatic environments since the early 1980s; however, global operational water quality products based on these data are not readily available (e.g., temperature, chlorophyll- a , turbidity, and suspended particle matter). Currently, because of the postprocessing required, only users with expressive experience can exploit these data, limiting their utility. Here, we provide paths (recipes) for the nonspecialist to access and derive water quality products, along with examples of applications, from sensors on board Landsat-5, Landsat-7, Landsat-8, Landsat-9, Sentinel-2A, and Sentinel-2B. We emphasize that the only assured metric for success in product derivation and the assigning of uncertainties to them is via validation with in situ data. We hope that this contribution will motivate nonspecialists to use publicly available high-resolution satellite data to study new processes and monitor a variety of novel environments that have received little attention to date.

Optimizing Optical Coastal Remote-Sensing Products: Recommendations for Regional Algorithm Calibration

“…Távora et al also suggested that Patos Lagoon exhibits optical properties that differ from those considered in the semi-analytical algorithm of Nechad et al [27]. Thus, the study area could serve as a case study for the regionality of turbidity and SPM estimates, which was addressed in previous works [10,34] by regionally recalibrating the coefficients of the turbidity algorithm for Landsat 5, 7, and 8. Therefore, we also provided regionally recalibrated coefficients specific to the study area, expanding this work to include SPM and other satellites, applying a more robust recalibration method, and examining the improvements and limitations associated with this approach.…”

“…This choice was based on previous comparisons of these two methods, showing that they have distinct spectral uncertainties [11] and performances for different final products [25]. In Patos Lagoon, previous works have applied ACOLITE [10,42] and POLYMER [34], but without a direct comparison of these methods. Additionally, [10] lists the advantages and limitations of each correction.…”

“…Traditional approaches for quantifying these parameters rely on in situ measurements of turbidity or SPM and usually involve costly, laborious, and time-consuming processes, limiting the study of the spatiotemporal variability in large areas and revealing associated uncertainties [8]. Since turbidity and SPM are associated with light scattering, however, they can be studied using remote sensing [10]. This alternative provides several advantages, allowing a synoptic view of large areas and frequent revisits at a low cost.…”

“…The regional calibration of algorithms is an alternative [10,34], combining the advantages of two distinct methodologies-a global and semi-analytical relationship between the reflectance and the variable of interest-combined with empirical adjustments to account for the specificity of the study area (related, for example, to the mass-specific coefficient [32]). Recalibration can be performed using satellite reflectance data, potentially improving the algorithm performance in regions with limited in situ radiometric data, with the disadvantage of adding uncertainty associated with the atmospheric correction process [10,34]. Additionally, spectral convolution is an important step when applying remote-sensing algorithms due to the spectral differences between hyperspectral (commonly used for algorithm development) and multispectral (most satellites) sensors [10].…”

“…Recalibration can be performed using satellite reflectance data, potentially improving the algorithm performance in regions with limited in situ radiometric data, with the disadvantage of adding uncertainty associated with the atmospheric correction process [10,34]. Additionally, spectral convolution is an important step when applying remote-sensing algorithms due to the spectral differences between hyperspectral (commonly used for algorithm development) and multispectral (most satellites) sensors [10]. Spectrally convoluted coefficients are specific to each satellite sensor [35].…”