Dissertations, Theses, and Capstone Projects

Date of Degree


Document Type


Degree Name



Earth & Environmental Sciences


Alexander Gilerson

Committee Members

Fred Moshary

Nir Krakauer

Paul DiGiacomo

Robert Foster

Subject Categories

Data Science | Electromagnetics and Photonics | Environmental Monitoring | Oceanography | Other Electrical and Computer Engineering


Ocean Color, Atmospheric Correction, Remote Sensing Reflectance, satellites, uncertainty, spatial resolution


Ocean Color radiometry uses remote sensing to interpret ocean dynamics by retrieving remote sensing reflectance (π‘…π‘Ÿπ‘ ) from satellite imagery at different scales and over different time periods. π‘…π‘Ÿπ‘  spectrum characterizes the ocean color that we observe, and from which we can discern concentrations of chlorophyll, organic and inorganic particles, and carbon fluxes in the ocean and atmosphere. π‘…π‘Ÿπ‘  is derived from the total radiance at the top of the atmosphere (TOA). However, it only represents up to ten percent of the total signal. Hence, the retrieval of π‘…π‘Ÿπ‘  from the total radiance at TOA involves the application of atmospheric correction (AC) algorithms, which include accurate modeling of Rayleigh and aerosol scattering, glint, and water variability. Each of these components yields uncertainties in the retrieved value of π‘…π‘Ÿπ‘ , especially in the blue bands. It is important to understand the main sources of uncertainties in π‘…π‘Ÿπ‘ , as uncertainties propagate into the retrieval of water parameters, which in turn inform climate models. In this study, a model was developed that quantifies the uncertainties of the main components in the current AC algorithm and used to analyze holistically the influence of these components on the π‘…π‘Ÿπ‘  uncertainties spatially and temporally in different water types taking advantage of the spectral differences between the components.

The uncertainties were determined by comparing satellite and in situ data, with the in situ data obtained from the AErosol RObotic NETwork - Ocean Color (AERONET-OC) around the Northern Hemisphere and the Marine Optical BuoY (MOBY), Lanai, Hawaii. The satellite sensor data are from the Visible Infrared Imaging Radiometer Suite (VIIRS) on the S-NPP platform, the Ocean and Land Colour Instruments (OLCI) on Sentinel 3A and 3B, and the Operational Land Imager (OLI) on Landsat 8.

Results showed that the Rayleigh component (molecular scattering and surface effects) is the main source of π‘…π‘Ÿπ‘  uncertainties for all water types, followed by water variability, which is more influential in coastal areas. The contributions of other components, including aerosol scattering, are usually smaller. In addition, wind speed ranges can influence results, especially in coastal regions. Across spatial scales, water variability played a dominant role in π‘…π‘Ÿπ‘  uncertainty and increased proportionally to the ground sampling distance.