Statistical Evaluation of Sentinel-3 OLCI Ocean Color Data Retrievals
Karlis Mikelsons, Menghua Wang, Ewa Kwiatkowska, Lide Jiang, David Dessailly, Juan Ignacio Gossn
Abstract
We employ a previously developed statistical method to evaluate the performance of the Sentinel-3 OLCI (Ocean and Land Colour Instrument) global ocean color data relying on the temporal stability of the retrievals. We analyze the normalized water-leaving reflectance ρ <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">wN</i> (λ) spectra generated by the Multi-Sensor Level-1 to Level-2 (MSL12) ocean color data processing system from the OLCI measurements, as well as EUMETSAT-IPF-OL-2 OLCI reflectance ρ <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">wN</i> (λ) spectra. The deviations in ρ <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">wN</i> (λ) spectra from temporally and spatially averaged baseline data are statistically evaluated corresponding to various parameters, including the solar-sensor geometry, various ancillary data (i.e., surface wind speed, sea-level atmospheric pressure, water vapor amount, and ozone concentration), and other related parameters. Our results show that, under most conditions, both NOAA-MSL12 and EUMETSAT-IPF-OL-2 data processing systems produce statistically consistent ocean color products in the open ocean with respect to all corresponding parameters analyzed, but with some underestimates of ρ <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">wN</i> (λ) spectra by EUMETSAT retrievals in moderate sun glint conditions being the notable exception.