Inter-comparison and Validation of Ocean Colour Chlorophyll Products off the Western Iberian Peninsula Coast
Brotas, Vanda¹; Sá, Carolina¹; D'Alimonte, Davide²; Kajiyama, Tamito³; Da Silva, José^4
1University of Lisbon, Centre of Oceanography, PORTUGAL; ²New University of Lisbon, FCT, Centria, PORTUGAL; ³New University of Lisbon, FCT, CITI, PORTUGAL; ⁴University of Porto, FCUP, CIIMAR, PORTUGAL

Chlorophyll a (Chl a) is the main biomass proxy for phytoplankton, and ocean colour sensors allow a synoptic and quasi-permanent following of this pigment concentration in surface waters. However, algorithms that are designed for use at global scales may be less accurate at local and regional scales, namely in coastal areas. These areas are optically complex but are most important to monitor since they are subject to major anthropogenic pressures. Therefore, it is important to understand the limitations of the different products in reference to specific areas and to validate the ocean-colour standard products with in situ data in order to match quality requirements for monitoring purposes. The present study takes only product validation in consideration once no optical data were available for atmospheric correction or algorithm verification. Remote sensing data were compared directly with in situ surface data. Water samples collected off the Portuguese coast during the period 2005 - 2012 were processed by reversed phase High-Performance Liquid Chromatography (HPLC) for pigment determination, and the Chl a concentration compared with coincident MERIS and MODIS sensors data. The performance of MERIS (algal1 and algal2) and OC3M MODIS products was evaluated using match-up data sets. Satellite products were found to overestimate Chl a concentrations of in situ values. An index of sea-water type was calculated for the matchup dataset and product accuracy was observed to be related to the type of water considered. Maps of index water type revealed seasonal variations along the coast. Analysis of non-uniform Chl a vertical profiles confirmed this parameter to be also influential on validation results. Pigments information allowed for classification of phytoplankton into size-classes and higher errors were associated to matchups where dominance in micro-sized phytoplankton was verified in situ.