Algorithms Merging for the Determination of Chlorophyll- ${a}$ Concentration in the Black Sea

Two regional bio-optical algorithms are combined to retrieve the Chlorophyll-<italic>a</italic> (<inline-formula> <tex-math notation="LaTeX">$\text {Chl-}\textit {a}$ </tex-math></inline-formula>) concentration in the Black Sea. The first is a band-ratio algorithm that computes <inline-formula> <tex-math notation="LaTeX">$\text {Chl-}\textit {a}$ </tex-math></inline-formula> as a function of the slope of Remote Sensing Reflectance (<inline-formula> <tex-math notation="LaTeX">$\mathit {R_{\text {RS}}}$ </tex-math></inline-formula>) values at two wavelengths using a polynomial regression that captures the overall data trend, enhancing extrapolation results. The second algorithm is a Multilayer Perceptron neural net based on <inline-formula> <tex-math notation="LaTeX">$\mathit {R_{\text {RS}}}$ </tex-math></inline-formula> values at three individual wavelengths that features interpolation capabilities helpful to fit data non-linearities. A new merging scheme is then designed to benefit from the complementarity of the two approaches. Remote sensing data employed to demonstrate the merging of regional results for the Black Sea are those acquired by the Ocean and Land Color Instrument on board Sentinel-3A to acknowledge the need for data products of higher accuracy within the long-term Copernicus program.