.. _rrs:

Remote Sensing Reflectance (Rrs)
################################

Remote Sensing Reflectances product is based on optical satellite data from the Sentinel-2 constellation. 
It is generated in near real-time (NRT) once a week for the entire France and overseas area. 
Remote Sensing Reflectances (Rrs) product is delivered in the steradian unit (sr-1). 
As an aquatic specific data corresponding to the ratio of Luminance (L in W.m-2.sr-1) and Irradiance (I in W.m-2) at the Bottom-of-Atmosphere (BOA). 


Retrieval methodology 
*********************

Products are derived from a single Sentinel-2 L1C product after atmospheric correction from the GRS processor (Harmel et al., 2018; Pahlevan et al., 2021), 
which performs atmospheric correction and generates a cloud mask from s2cloudless processor (Skakun et al., 2022; Zupanc, 2017) 
and a simple water mask from normalized difference water index (NDWI).
GRS is developed by Magellium with the support of CNES [#f1]_ and INRAE [#f2]_. The main functions of this system correct atmospheric effects using Sentinel-2 L1C images, 
ECMWF-ERA5 [#f3]_ gaseous data and SRTM [#f4]_ DEM data. This correction takes into account the effect of topography for gaseous correction. 
It also takes into account and corrects for the specular reflection of direct sunlight called “sunglint” on the water surface.

Product limitations 
*******************

Rrs derive from the optical estimates of the reflected light of the sun, outcoming from the water. 
Therefore, no data can be produced when clouds are present in the atmosphere.
To mask clouds, for Sentinel-2-MSI images we included the processor that uses the algorithm “s2cloudless” based on the machine-learning x-boost (Skakun et al., 2022; Zupanc, 2017).
Atmospheric corrections algorithms for aquatic applications are constantly under development and improvement, and no processor has been proven to outperform the others 
(Pahlevan et al., 2021). 
Clouds can be present over the scenes but the masks are not applied directly over the Rrs produced. 
Therefore, the user needs to specify (mask==0) to eliminate land and clouds to obtain usable Rrs.
The GRS processor operates by applying a pixel by pixel gaseous and aerosol corrections based on ERA5 data, and corrects from sunglint pollution 
by deriving sunglint intensity in all bands from the signal detected on the SWIR band.
It implies that some pixels can present negative values, especially in the violet and blue bands due to overestimating the Atmospheric Optical Thickness (AOT).
Despite good aerosol estimation from various models, retrieving the Rrs over is subject to significant uncertainties, particularly over inland waters. 
Expert users usually verify the amplitudes of Rrs and spectral shapes to detect anomalous retrievals.


.. rubric:: Footnotes
.. [#f1] Centre National d'études spatials - French Space Agency 
.. [#f2] Institut national de recherche pour l'agriculture, l'alimentation et l'environnement - French Research Institute for Agriculture, Food and the Environment
.. [#f3] European Centre for Medium-Range Weather Forecasts (https://atmosphere.copernicus.eu/)
.. [#f4] Shuttle Radar Topography Mission