Multispectral Cloud Identification Study.
Abstract
Tactical and meteorological considerations demand a knowledge of cloud cover on a global basis. Since areas exist which are either data poor or denied by hostile forces, it is reasonable to infer cloud parameters in these regions from satellite-borne passive sensors. Both channel selection and inversion methodology to maximize the amount of cloud information are currently open questions. This study deals with the problem of inferring both cloud and 'nuisance' parameters given four frequencies in the microwave region (10 GHz, 19 GHz, 37 GHz and 94 GHz) and ten frequencies in the thermal IR, near IR and visible region (.55, .72, 1.0, 1.6, 2.1, 3.8, 6.7, 10.5, 11.5, and 12.5 micrometers). Using simulated data, several studies were made. Results indicate that the iterated extended Kalman filter is a robust method of cloud parameter estimation, and that while some parameters such as cloud top height and liquid water content are well estimated, others such as cloud thickness are not. Finally, simple decision trees may be formulated to distinguish the scenes analyzed.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 01, 1978
- Accession Number
- ADA077873
Entities
People
- Donald E. Gustafson
- Eliot S. Blackman
- Mary G. Fowler
- William H. Ledsham