Effects of water-emission anisotropy on multispectral remote sensing at thermal wavelengths of ocean temperature and of cirrus clouds.

The assumption of blackbody emission (emissivity, 1.0) for a calm ocean surface can lead to significant underestimates of the sea-surface temperature (SST) derived from IR radiometric data. Taking the optical properties of the atmosphere as known, we calculate the errors stemming from the blackbody assumption for cases of a purely absorbing or a purely scattering atmosphere. It is observed that for an absorbing atmosphere the errors in SST are always reduced and are the same whether measurements are made from space or at any level in the atmosphere. As for atmospheric scattering, the SST errors are slightly reduced when one is viewing from large zenith angles but are slightly enhanced when one is viewing from the zenith. The inferred optical thickness tau of an absorbing layer can be in error under the blackbody assumption by a Deltatau of 0.01-0.08, while the inferred optical thickness of a scattering layer can be in error by a larger amount, Deltatau of 0.03-0.13. The error Deltatau depends only weakly on the actual optical thickness and on the viewing angle, but it is rather sensitive to the wavelength of the measurement. In the absence of steep slopes in the wave-slope distribution, directional emissivities are essentially unchanged by sea state when one is viewing from or near the zenith. When one is viewing from moderately large zenith angles (such as 507 degrees ), however, the departures in the directional emissivities from blackbody emission can be much larger under perturbed sea state than under calm conditions.