Publications

Sahay, A; Ali, SM; Gupta, A; Goes, JI (2017). Ocean color satellite determinations of phytoplankton size class in the Arabian Sea during the winter monsoon. REMOTE SENSING OF ENVIRONMENT, 198, 286-296.

Abstract
A regionally tuned three component "abundance" model of Brewin et al. (2012) has been used to discriminate satellite ocean color derived fields of phytoplankton biomass observable as Chlorophyll-a (Chl-a), into three size classes, i.e. microplankton (>20 mu m), nanoplankton (>2 to <20 mu m) and picoplankton (<2 mu m). The model has been applied to MODIS-Aqua and Oceansat-2, Ocean Color Monitor (OCM) derived fields of Chl-a data between Nov. and Mar. In the Arabian Sea, during the evolution of blooms of the large (>800 mu m sized) green mixotrophic dinoflagellate Noctiluca scintillans. A comparison of shipboard measured and model derived values of phytoplankton size classes (PSCs) show the superiority of the regionally tuned model over parameterizations used in the original model of Brewin et al. (2012). A total number of 39 in situ data points have been used for the tuning of the regional model and 5 different in-situ data points have been used for the comparison with in situ data in this remote region of data paucity. The absolute mean and the maximum absolute errors for all size fractions are 4.7% and 17.2% respectively, as compared to the values of 9.6% and 26% respectively obtained using Brewin et al. (2012). When applied to a weekly time series of Chl-a images, the regionally tuned model is able to capture the seasonal cycle of PSC in the Arabian Sea associated with the tail end of the fall inter-monsoon (Nov.), the winter monsoon (Dec. to Feb.) and the transition to the spring inter-monsoon Although ocean color remote sensing is a useful tool for studying phytoplankton processes in regions like the Arabian Sea that suffer from a paucity of in-situ observations, enhancing the validity and confidence in satellite ocean color derived products such as PSC, will require additional shipboard datasets. (C) 2017 Published by Elsevier Inc.

DOI:
10.1016/j.rse.2017.06.017

ISSN:
0034-4257