Bright Oceans: Hyperspectral differentiation of whitecaps, plastics, sea ice, and floating vegetation and other flotsam
Heidi M. Dierssen, Professor, Department of Marine Sciences and Department of Geography, University of Connecticut
Geography Colloquium
Friday, November 1st, 2019
12:20 PM - 01:15 PM
Storrs Campus
AUST 434
Viewed from space, the ocean is a dark surface underlying a thick atmosphere which reflects sunlight back to the sensor. Ocean surfaces, however, are not always dark. Various optically active constituents both in the water and on the sea surface can enhance the surface albedo and make the oceans appear much “brighter” than normal. These bright waters serve to enhance water reflectance and can include the presence of whitecaps and foam, sea ice, plastics, and other flotsam at the sea surface. A user of ocean color imagery might presume that bright water features would appear as enhancements in the ocean color reflectance signal; however, standard ocean color processing aims to mask or quantitatively remove these bright features as part of the atmospheric correction. Differentiating between these sea surface constituents, many of which appear “white” or spectrally flat in the visible spectrum (400 to 700 nm), is the objective of this study with the intent to understand how the unique reflectance features of these bright reflecting targets can be used to better conduct atmospheric correction of imagery and to develop new parameters from ocean color imagery. Hyperspectral measurements of these various constituents from visible to the short wave infrared will be provided and various methods to use narrowband features specific to different targets will be presented. While some targets such as marine microplastics may be extremely challenging to differentiate unless they are densely concentrated within a pixel, other targets such as whitecaps, sea ice and floating vegetation are prevalent in satellite imagery and have unique spectral features that allow us to differentiate them from clouds and aerosols in the atmospheric correction routines.
For more information, contact: Nat Trumbull at trumbull@uconn.edu