Species identification of mixed algal bloom in the Northern Arabian Sea using remote sensing techniques.

ABSTRACT

Oceanic waters of the Northern Arabian Sea experience massive algal blooms during winter-spring (mid Feb-end Mar), which prevail for at least for 3 months covering the entire northern half of the basin from east to west. Ship cruises were conducted during winter-spring of 2001-2012 covering different stages of the bloom to study the biogeochemistry of the region. Phytoplankton analysis indicated the presence of green tides of dinoflagellate, Noctiluca scintillans (=N. miliaris), in the oceanic waters. Our observations indicated that diatoms are coupled and often co-exist with N. scintillans, making it a mixed-species ecosystem. In this paper, we describe an approach for detection of bloom-forming algae N. scintillans and its discrimination from diatoms using Moderate Resolution Imaging Spectroradiometer (MODIS)-Aqua data in a mixed-species environment. In situ remote sensing reflectance spectra were generated using Satlantic™ hyperspectral radiometer for the bloom and non-bloom waters. Spectral shapes of the reflectance spectra for different water types were distinct, and the same were used for species identification. Scatter of points representing different phytoplankton classes on a derivative plot revealed four diverse clusters, viz. N. scintillans, diatoms, non-bloom oceanic, and non-bloom coastal waters. The criteria developed for species discrimination were implemented on MODIS data and validated using inputs from a recent ship cruise conducted in March 2013.