Cloud Analysis from Bi Spectral Satellite Data,
Abstract
A horizontal differencing bi-spectral technique has been developed which includes an iteration scheme for reducing errors in computed cloud amount. The technique requires that cloud surfaces over the area of application be horizontally homogeneous, and as developed, assumes that the observed maximum and minimum brightness counts represent cloud and clear filled resolution points respectively. These values are then used to normalize the data in computing total cloud amount. The computed results of the horizontal differencing bi-spectral method, as applied to real data sets, have been compared to the results obtained from a modified frequency distribution method and the general bi-spectral method. The results of this comparative analysis indicate that the computed cloud amounts of the horizontal differencing method are less variable than for the frequency distribution and general bi-spectral methods, and are thus better suited for objective analyses. The computed cloud temperatures of the horizontal differencing method were also shown to be more realistic than those computed by the general bi-spectral method.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 01, 1978
- Accession Number
- ADA065474
Entities
People
- Christopher Mendola
- Stephen K. Cox
Organizations
- Colorado State University