Dependence of Physiochemical Features on Marine Chlorophyll Analysis with Learning Techniques. (arXiv:2304.12325v1 [q-bio.QM])

Marine chlorophyll which is present within phytoplankton are the basis of
photosynthesis and they have a high significance in sustaining ecological
balance as they highly contribute toward global primary productivity and comes
under the food chain of many marine organisms. Imbalance in the concentrations
of phytoplankton can disrupt the ecological balance. The growth of
phytoplankton depends upon the optimum concentrations of physiochemical
constituents like iron, nitrates, phosphates, pH level, salinity, etc. and
deviations from an ideal concentration can affect the growth of phytoplankton
which can ultimately disrupt the ecosystem at a large scale. Thus the analysis
of such constituents has high significance to estimate the probable growth of
marine phytoplankton. The advancements of remote sensing technologies have
improved the scope to remotely study the physiochemical constituents on a
global scale. The machine learning techniques have made it possible to predict
the marine chlorophyll levels based on physiochemical properties and deep
learning helped to do the same but in a more advanced manner simulating the
working principle of a human brain. In this study, we have used machine
learning and deep learning for the Bay of Bengal to establish a regression
model of chlorophyll levels based on physiochemical features and discussed its
reliability and performance for different regression models. This could help to
estimate the amount of chlorophyll present in water bodies based on
physiochemical features so we can plan early in case there arises a possibility
of disruption in the ecosystem due to imbalance in marine phytoplankton.