HR: 0800h
AN: OS21C-1276    [Abstracts]
TI: The Bidirectional Reflectance Distribution Function (BRDF) of the Ocean: Effects of Wave Representation
AU: * Eide, H
EM: heide@stevens-tech.edu
AF: Stevens Institute of Technology, Department of Physics and Engineering Physics, Hoboken, NJ 07030 United States
AU: Stamnes, K
EM: kstamnes@stevens-tech.edu
AF: Stevens Institute of Technology, Department of Physics and Engineering Physics, Hoboken, NJ 07030 United States
AU: Ottaviani, M
EM: mottavia@stevens-tech.edu
AF: Stevens Institute of Technology, Department of Physics and Engineering Physics, Hoboken, NJ 07030 United States
AB: The specular reflection of the Sun off the ocean, or sun glint, is of major concern for ocean remote sensing. Typically, data from in and around the sunglint region are discarded because of the unknown contribution to the measured radiances or because of sensor saturation. On the other hand, accurate knowledge of the sunglint properties enables retrievals of atmospheric parameters. The challenge of the ocean retrieval problem is to get the ``water leaving radiance'', $L_{w}$, by subtracting the Rayleigh scattering, aerosol scattering, water vapor, ozone, and sun glint from the measured radiances at the top of the atmosphere (TOA). Thus, the task is to correct for both the atmospheric contribution and for surface effects. Two simplifying assumptions that are frequently employed in ocean remote sensing are that the ocean BRDF is isotropic and that one can de-couple the radiative properties of the atmosphere from those of the surface. Our previous studies have shown that neglecting the inherit coupling between the atmosphere and surface can lead to large errors in the retrievals. In order to do retrievals over bright, as well as darker surfaces, it is necessary to account for this coupling between the surface and the atmosphere. In the present study we use models for the reflection of light off the ocean surface to calculate the ocean BRDF. The differences between the various models are investigated as is the effect of using different types of wave statistics (e.g. Cox Munk). We present results from calculations where we vary the wind speed and direction as well as other parameters affecting the ocean surface. The error introduced in ocean retrievals by assuming an isotropic BRDF is assessed, and methods for improved treatment of sunglint are suggested.
UR: http://odin.mat.stevens-tech.edu/
DE: 5757 Remote sensing
DE: 3360 Remote sensing
DE: 4275 Remote sensing and electromagnetic processes (0689)
DE: 4560 Surface waves and tides (1255)
DE: 0933 Remote sensing
SC: Ocean Sciences [OS]
MN: 2004 AGU Fall Meeting