Aerosol and Clouds
Solar zenith and viewing geometry-dependent errors in satellite retrieved cloud optical thickness: Marine stratocumulus case
Article first published online: 3 JAN 2009
Copyright 2009 by the American Geophysical Union.
Journal of Geophysical Research: Atmospheres (1984–2012)
Volume 114, Issue D1, 16 January 2009
How to Cite
2009), Solar zenith and viewing geometry-dependent errors in satellite retrieved cloud optical thickness: Marine stratocumulus case, J. Geophys. Res., 114, D01202, doi:10.1029/2008JD010579., and (
- Issue published online: 3 JAN 2009
- Article first published online: 3 JAN 2009
- Manuscript Accepted: 9 OCT 2008
- Manuscript Revised: 17 SEP 2008
- Manuscript Received: 10 JUN 2008
- remote sensing
 The error in the domain-averaged cloud optical thickness retrieved from satellite-based imagers is investigated using a cloud field generated by a cloud model and a 3D radiative transfer model. The objective of this study is to identify the optimal geometry for the optical thickness retrieval and quantify the error. The cloud field used in the simulation is a relatively uniform (retrieved shape parameter of a gamma distribution averaged over all simulated viewing and solar zenith angles is 18) and nearly isotropic stratocumulus field. The retrieved cloud cover with a 1-km pixel resolution is 100%. The domain-averaged optical thickness error is separated into two terms, the error caused by an assumption of a horizontally uniform cloud over a 1-km pixel (internal variability) and error caused by neglecting the horizontal flux through the boundary of subpixels (external variability). For the cloud field used in this study, the external variability term increases with solar zenith angle and the sign changes from negative to positive while the internal variability term is generally negative and becomes more negative as the solar zenith angle increases. At a small solar zenith angle, therefore, both terms are negative, but the error partially cancels at a large solar zenith angle. When the solar zenith angle is less than 30°, both terms are small; the error in the viewing zenith angle and domain-averaged cloud optical thickness derived from the relative azimuth angle smaller than 150 is less than 10%. However, if the optical thickness is derived from nadir view only for overhead sun, the domain-averaged optical thickness is underestimated by more than 10%. When the solar zenith angle increases to 60°, the internal variability term exceeds 10%, especially viewed from the forward direction, but the domain and viewing zenith angle averaged optical thickness error can be less than 10% in the backward direction. When the solar zenith angle is 70°, both terms are greater than 10%. The shape parameter of a gamma distribution derived from retrieved optical thicknesses increases with the viewing zenith angle but decreases with solar zenith angle. On the basis of this simulation and Terra Moderate Resolution Imaging Spectroradiometer (MODIS) viewing geometry and solar zenith angle at the sampling time over the northeastern Pacific, the error in the domain-averaged retrieved optical thickness of uniform stratocumulus over northeastern Pacific is less than 10% in March and September.