Date of Award

Spring 2007

Document Type


Degree Name

Doctor of Philosophy (PhD)


Ocean/Earth/Atmos Sciences

Committee Director

Richard C. Zimmerman

Committee Member

Larry P. Atkinson

Committee Member

Margaret R. Mulholland

Committee Member

Zhonghai Jin

Committee Member

Mark Havey


The primary purpose of this study was to develop an inverse method to retrieve the inherent optical properties (IOPs) and biogeochemical parameters (e.g. chlorophyll a concentration and salinity) appropriate to monitor the water quality and biogeochemical processes from remote sensing of the coastal waters in the southern Chesapeake Bay and coastal Mid-Atlantic Bight region (MAB) dominated by Case 2 waters. For this purpose, knowledge of the relationship between remote sensing reflectance (Rrs) and IOPs and the effect from bottom reflectance on Rrs, is required.

A substantial investigation of IOPs has been conducted for the coastal waters of the southern Chesapeake Bay. Although phytoplankton are the dominant contributors to IOPs of oceanic Case 1 waters, colored dissolved organic matter (CDOM) derived from non-phytoplankton sources and sedimentary particles also play very important roles in coastal Case 2 waters. Strongly influenced by riverine discharge, the shallow coastal waters of the southern Chesapeake Bay provide challenges and opportunities to develop regionally specific IOP retrieval methods from remotely sensed ocean color imagery.

A semi-analytical radiative transfer model (PZ06_Ed), based on the analysis of the simulated results of an exact radiative transfer model, Hydrolight® [Mobley, 1994], was developed to estimate the vertical distribution of downwelling plane irradiance [Ed(z)] from IOPs and sky conditions (e.g. cloud coverage and solar zenith angle). Compared to the significant overestimation of the simple Gordon [1989] model for particle-rich environments, PZ06_Ed agreed with Hydrolight® with < 6% of the root-mean-square (RMS) error. Field observations from the coastal waters of the southern Chesapeake Bay validated the predictions of PZ06_ Edwith RMS error from 10% to 14%. The SeaWiFS imagery of the diffuse attenuation coefficient (Kd) estimated from PZ06_Ed is significantly improved from the Mueller [2000] model and displays obviously the coastal processes in the lower MAB, including the riverine outflow from the Chesapeake Bay and the mixing of the Gulf Stream with the local waters.

The quadratic model (e.g. GSMO1) describing Rrs and IOPs has been widely used in bio-optics to retrieve inherent optical properties (IOPs). In this study, the derived coefficients (l1 and l2) by Gordon et al. [1988] were re-evaluated from Hydrolight® simulations and incorporated into a semi-analytical radiative transfer model (PZ06_ Rrs) that included bottom effects for optically shallow waters. Compared with Hydrolight® simulations and field observations in the Chesapeake Light Tower (CLT), Rrs calculated from PZ06_Rrs typically agreed within 5% and about 7% to 13% of RMS, respectively. Hydrolight ® simulations and field observations also confirmed that PZ06_Rrs improved the retrieval of biogeochemical-related parameters, including [Chl], adg(443), and bbp(443), compared to global ocean color algorithms (e.g. OC3M) and semi-analytic models without considering the bottom effects (e.g. GSM01-CLT).

Finally, the relatively successful inverse modeling provides a promising method to study ecosystem-level biogeochemical and physical parameters from remote sensing for coastal waters of southern Chesapeake Bay and even lower MAB.