Automated mesoscale winds determined from satellite imagery

A new automated technique for extracting mesoscale fields from GOES visible/infrared satellite imagery was developed. Quality control parameters were defined to allow objective editing of the wind fields. The system can produce equivalent or superior cloud wind estimates compared to the time consuming manual methods used on various interactive meteorological processing systems. Analysis of automated mesoscale cloud wind for a test case yields an estimated random error value one meter per second and produces both regional and mesoscale vector wind field structure and divergence patterns that are consistent in time and highly correlated with subsequent severe thunderstorm development.