Properties of aerosol and surface derived from OLCI/Sentinel-3A using GRASP approach: Retrieval development and preliminary validation

The Ocean and Land Color Instrument (OLCI) onboard the Copernicus Sentinel-3A satellite is a medium-resolution and multi-spectral push-broom imager acquiring radiance in 21 spectral bands covering from the visible to the far near-infrared. These measurements are primary dedicated to land & ocean color applications, but actually include also reliable information for atmospheric aerosol and surface brightness characterization. In the framework of the EUMETSAT funded study to support the Copernicus Program, we describe the retrieval of aerosol and surface properties from OLCI single-viewing multi-spectral Top-Of-Atmosphere (TOA) radiances based on the Generalized Retrieval of Atmosphere and Surface Properties (GRASP) algorithm. The high potential of the OLCI/GRASP configuration stems from the attempt to retrieve both aerosol load and surface reflectance simultaneously using a globally consistent high-level approach. For example, both over land and ocean surfaces OLCI/GRASP uses 9 spectral channels (albeit with different weights), strictly the same prescribed aerosol models and globally the same a priori constraints (though with some differences for observations over land and ocean). Due to the lack of angular multi-viewing information, the directional properties of underlying surface are strongly constrained in the retrieval: over ocean the Fresnel reflection together with foam/whitecap albedo are exclusively computed using a priori wind speed; over land, the Bidirectional Reflectance Distribution Function (BRDF) is slightly adjusted from a priori values of climatological Ross-Li volumetric and geometric terms. Meanwhile, the isotropic reflectance is retrieved globally under mild spectral smoothness constraints. It should be noticed that OLCI/GRASP configuration employs innovative multi-pixel concept (Dubovik et al., 2011) that enhance retrieval by simultaneously inverting large group of pixels. The concept allows for benefiting from knowledge about natural variability of the retrieved parameters.