Cloud type pattern recognition using environmental satellite data

Abstract

A classification analysis is conducted concerning the tropical cloud types as remotely sensed in the visual and infrared range by ITOS scanning radiometers. A statistical pattern recognition technique is used to examine the ability of coincident dual-channel and single-channel data to classify four main forms of clouds, including cumulus, stratocumulus, cumulonimbus, and cirrus.