An Automatic Cloud Tracking System Based on the Cross-Covariance Method

摘要

An automatic cloud tracking system based on the computation of the cross-covariance between satellite images has been developed. The six steps of the System for Automatic Wind Extraction from Geostationary Satellite-Data (SAWEGS) are preprocessing, cloud tracking, height assignment, earth location, wind vector computation, and quality control. The system embodies various unique aspects developed to address the problems encountered with an automated technique in general and this technique in particular. These aspects include histogram-based temperature slicing, cloud edge and surface enhancement, tracking trackability tests, and use of numerical analysis temperature profiles for height assignment. Examples of applying SAWEGS to visible data, to infrared data, and to a difficult tropical cyclone case show the good quality and coverage of resulting vectors as well as the remaining difficulties. (Author)