Remote sensing of chlorophyll concentration in the Chesapeake Bay through discrimination of phytoplankton contribution to total light absorption coefficient

Robust derivation of chlorophyll-a concentration, [Chl-a], has been difficult in the Chesapeake Bay owing largely to abundant absorbing substances that coexist but do not covary with phytoplankton, e.g., nonalgal particles (NAP) and colored dissolved organic matter (CDOM). Recently, a generalized stacked-constraints model (GSCM) for partitioning the total light absorption coefficient of water with pure-water contribution subtracted, a_nw(λ), into phytoplankton, a_ph(λ), NAP, a_d(λ), and CDOM, a_g(λ), components has been developed [Zheng et al., J. Geophy. Res., 2015]. This model permits the separation between algal and nonalgal optical constituents without imposing highly restrictive constraints on variability in their spectral shapes, thereby facilitating the effective derivation of [Chl-a] from a_ph(λ). In the present study, we leveraged field [Chl-a] data provided through the Chesapeake Bay Program (http://www.chesapeakebay.net/) and matched them up with satellite overpasses. Using the matchup data, we evaluated the feasibility of inverting satellite remote-sensing reflectance to obtain [Chl-a] through the route of reflectance-derived a_nw(λ) and GSCM-partitioned a_ph(λ). Our approach sheds light on providing improved [Chl-a] product for coastal waters such as the Chesapeake Bay.