ALBERT

Description: Atmospheric motion vectors (AMVs) derived from geostationary meteorological satellites have long stood as an important observational contributor to analyses of global-scale tropospheric wind patterns. This paradigm is evolving as numerical weather prediction (NWP) models and associated data assimilation systems are at the point of trying to better resolve finer scales. Understanding the physical processes that govern convectively-driven weather systems is usually hindered by a lack of observations on the scales necessary to adequately describe these events. Fortunately, satellite sensors and associated scanning strategies have improved and are now able to resolve convective-scale flow fields. Coupled with the increased availability of computing capacity and more sophisticated algorithms to track cloud motions, we are now poised to investigate the development and application of AMVs to convective-scale weather events. Our study explores this frontier using new-generation GOES-R Series imagery with a focus on hurricane applications. A proposed procedure for processing enhanced AMV datasets derived from multispectral geostationary satellite imagery for hurricane-scale analyses is described. We focus on the use of the recently available GOES-16 mesoscale domain sector rapid-scan (1-min) imagery, and emerging methods to optimally extract wind estimates (atmospheric motion vectors (AMVs)) from close-in-time sequences. It is shown that AMV datasets can be generated on spatiotemporal scales not only useful for global applications, but for mesoscale applications such as hurricanes as well.