MODIS-derived green Noctiluca blooms in the upper Gulf of Thailand: Algorithm development and seasonal variation mapping

Abstract

In recent decades, red tides of non-toxic harmful algal blooms have frequently occurred in monsoon-influenced tropical areas, particularly the green form of Noctiluca scintillans (hereafter green Noctiluca). However, our understanding of the mechanism of red tide formation is hindered by spatial and temporal constraints of field data. In this study, we used moderate resolution imaging spectroradiometer (MODIS) ocean color data along with a locally developed algal-bloom classification algorithm to investigate the seasonal variability of dominant red tides across the upper Gulf of Thailand (uGoT). During our July 2018 observation, a super green Noctiluca bloom with extraordinarily high chl-a (>1,469 mg m-3) displayed a distinct spectral reflectance characteristic among red tides in blue-to-green and red-to-near infrared wavelengths. According to the distinctive in situ hyperspectral characteristics of uGoT algal blooms, we developed a classification algorithm for MODIS normalized at 488, 531, and 667 nm, which successfully discriminated green Noctiluca in three levels of blooms, namely, super (100%), strong (>80%), and weak (>40%), from other algal blooms (i.e., dinoflagellates, diatoms, cyanobacteria, and mixed red tide species) as well as non-bloom oceanic and coastal waters using MODIS data, as confirmed by uGoT red tide reports. Monthly MODIS-based discrimination composites from 2003 to 2021 revealed seasonal variability in the surface distribution and bloom frequency of green Noctiluca and other red tides according to the Asian monsoon seasons: the southwest monsoon (May–September) and the northeast monsoon (October–January of the following year). Green Noctiluca blooms occurred farther from the shore and estuaries than other red tides (dinoflagellates and cyanobacteria), and were much more frequent than other red tides between the Tha Chin and Chao Phraya River mouths during the non-monsoon period (February to April). The frequency and distribution of green Noctiluca blooms, as well as other algal blooms, varied with the monsoon season. By comparing MODIS-derived algal blooms to monsoon-induced factors (i.e., sea surface winds, precipitation, and river discharge), we present an unprecedented overview of the spatial and temporal dynamics of red tides throughout the uGoT under Asian monsoon conditions. This research contributes to our understanding of the impact of climate change on phytoplankton dynamics.