neural network methods for wind velocity profiles from satellite data

摘要

A method has bene developed for retrieval of wind velocity profiles from satellite sounder radiances using a neural network technique. Each wind velocity vector represents a mean value for a layer approximately centered at that standard height or pressure level. The input includes radiances from the Geostationary Operational Environmental Satellite (GOES), plus ancillary information such as latitude and longitude. A set of "co-incident" rawinsonde soundings provide the "truth" data set. In the initial results rroot mean square errors for the standard levles from 850 through 300 hPa have about the same magnitude as those for current methods such as trackign of moisture features in GOES imagery. Although the results to date appear highly promising, considerable further work is needed before we have an operationally capable technique.