Wind Extraction from Meteosat Water Vapor Channel Image Data

Abstract

Wind extraction from successive Meteostat water vapor channel (WVC) images has been performed in an operational environment at the European Space Operation Centre for several months. Motions are computed with a cross-correlation algorithm; height assignment is deduced from the observed WVC brightness temperature. The scheme is fully automated. No distinction is made between clouds and pure water vapor structures. The accuracy of the produced winds is assessed by statistical comparisons with forecast winds and radiosonde winds. Above 400 hPa, WVC derived winds are mainly produced in cloudy areas. They are more numerous and generally as reliable as cloud-motion winds extracted from thermal Infrared channel images, showing a great potential to improve observation of high-troposphere circulations. Below 400 hPa, WVC-derived winds are sparse and usually do not fit with conventional wind observations. WVC-derived winds are very sensitive to signal-to-noise ratios in the imagery.