Improvement in Estimation of Phytoplankton Size Class in Arabian Sea Using Remote Sensing Measurements

Abstract

Phytoplankton is the principal component of the ocean ecosystem, whose biomass (represented as total chlorophyll-a (Chl-a) concentration) and size structure are the key ecological indicators. With an aim to retrieve phytoplankton size class (PSC) in Arabian Sea (a region with limited measurements), this study focuses on modification of two existing algorithms, one for measuring Chl-a and the other for size fraction. PSC is estimated using a regionally tuned three-component model with Chl-a concentration taken as input. Parameters of the model are modified for Arabian Sea (AS) using <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">in-situ</i> data to estimate three size classes namely micro-, nano- and pico-plankton (model is referred as AS-PSC-2021). Accuracy of PSC retrieval depends on Chl-a concentration, which is estimated from spectral remote sensing reflectance ( <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">R_rs</i> (λ)) using Quasi Analytical Algorithm (QAA) along with non-linear optimization. This hybrid approach for Chl-a estimation is named “QAA-optimized model”. <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">In-situ</i> measurements of Chl-a concentration, <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">R_rs</i> (λ) and PSC fractions collected during April to May-2017 inter-monsoon period) from Arabian Sea are used along with the previous measurements from winter-monsoon periods of the year 2009, 2010 and 2011 for model development and validation. The PSCs were estimated using pigment concentrations obtained from High Performance Liquid Chromatography (HPLC). The Chl-a concentration obtained from QAA-optimized model was verified with <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">in-situ</i> data (N=44) and found to perform well (slope=0.95, bias=0.04, RMS=0.13 and R <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> =0.90). Also, the regionally tuned PSC model (AS-PSC-2021) showed a significant improvement in retrieving the size fraction when compared to previous PSC algorithm proposed for the region. The AS-PSC-2021 model was able to reproduce the variability in all the three size classes showing the dominance of pico- and micro-plankton at lower and higher Chl-a concentration, respectively, but more importantly the dominance of nano-plankton at mid Chl-a range. It is important to have a regionally tuned algorithm for key ecological parameters such as Chl-a and PSCs to assess the ecosystem in a perspective of ocean warming and climate change.