Performance of Ocean Colour Chlorophyll a algorithms for Sentinel-3 OLCI, MODIS-Aqua and Suomi-VIIRS in open-ocean waters of the Atlantic

• Published in: Remote sensing of environment Vol. 260; p. 112444
• Main Authors: Tilstone, Gavin H., Pardo, Silvia, Dall'Olmo, Giorgio, Brewin, Robert J.W., Nencioli, Francesco, Dessailly, David, Kwiatkowska, Ewa, Casal, Tania, Donlon, Craig
• Format: Journal Article
• Language: English
• Published: New York Elsevier Inc 01.07.2021; Elsevier BV
• Subjects: Algorithms; Anthropogenic factors; Atlantic Ocean; Atmospheric correction; Atmospheric models; Chlorophyll; Chlorophyll a; Color; Gyres; Low concentrations; Microprocessors; MODIS; MODIS-Aqua; Ocean color; Oceans; OLCI; Phytoplankton; Polymers; Reflectance; Satellite imagery; Sentinel-3; Sentinel-3A and 3B Tandem Phase; Spectrophotometry; Suomi-VIIRS; Validation; Validation; Sentinel-3; Atmospheric correction; Sentinel-3A and 3B Tandem Phase; OLCI; Atlantic Ocean; MODIS-Aqua; Suomi-VIIRS; Chlorophyll a
• ISSN: 0034-4257; 1879-0704
• DOI: 10.1016/j.rse.2021.112444

The proxy for phytoplankton biomass, Chlorophyll a (Chl a), is an important variable to assess the health and state of the oceans which are under increasing anthropogenic pressures. Prior to the operational use of satellite ocean-colour Chl a to monitor the oceans, rigorous assessments of algorithm performance are necessary to select the most suitable products. Due to their inaccessibility, the oligotrophic open-ocean gyres are under-sampled and therefore under-represented in global in situ data sets. The Atlantic Meridional Transect (AMT) campaigns fill the sampling gap in Atlantic oligotrophic waters. In-water underway spectrophotometric data were collected on three AMT field campaigns in 2016, 2017 and 2018 to assess the performance of Sentinel-3A (S3-A) and Sentinel-3B (S3-B) Ocean and Land Colour Instrument (OLCI) products. Three Chl a algorithms for OLCI were compared: Processing baseline (pb) 2, which uses the ocean colour 4 band ratio algorithm (OC4Me); pb 3 (OL_L2M.003.00) which uses OC4Me and a colour index (CI); and POLYMER v4.8 which models atmosphere and water reflectance and retrieves Chl a as a part of its spectral matching inversion. The POLYMER Chl a for S-3A OLCI performed best. The S-3A OLCI pb 2 tended to under-estimate Chl a especially at low concentrations, while the updated OL_L2M.003.00 provided significant improvements at low concentrations. OLCI data were also compared to MODIS-Aqua (R2018 processing) and Suomi-NPP VIIRS standard products. MODIS-Aqua exhibited good performance similar to OLCI POLYMER whereas Suomi-NPP VIIRS exhibited a slight under-estimate at higher Chl a values. The reasons for the differences were that S-3A OLCI pb 2 Rrs were over-estimated at blue bands which caused the under-estimate in Chl a. There were also some artefacts in the Rrs spectral shape of VIIRS which caused Chl a to be under-estimated at values >0.1 mg m-3. In addition, using in situ Rrs to compute Chl a with OC4Me we found a bias of 25% for these waters, related to the implementation of the OC4ME algorithm for S-3A OLCI. By comparison, the updated OLCI processor OL_L2M.003.00 significantly improved the Chl a retrievals at lower concentrations corresponding to the AMT measurements. S-3A and S-3B OLCI Chl a products were also compared during the Sentinel-3 mission tandem phase (the period when S-3A and S-3B were flying 30 sec apart along the same orbit). Both S-3A and S-3B OLCI pb 2 under-estimated Chl a especially at low values and the trend was greater for S-3A compared to S-3B. The performance of OLCI was improved by using either OL_L2M.003.00 or POLYMER Chl a. Analysis of coincident satellite images for S-3A OLCI, MODIS-Aqua and VIIRS as composites and over large areas illustrated that OLCI POLYMER gave the highest Chl a concentrations and percentage (%) coverage over the north and south Atlantic gyres, and OLCI pb 2 produced the lowest Chl a and % coverage. •Underway spectrophotometric Chl a were collected on 3 campaigns in the Atlantic.•Performance of Sentinel-3 OLCI, MODIS-Aqua, VIIRS ocean colour products assessed.•S-3A OLCI Chl a from POLYMER performed best; MODIS-Aqua also performed well.•Both S-3A and S-3B OLCI pb 2 under-estimated Chl a and was corrected using OL_L2M.003.00 and POLYMER.