Creating diagnostic techniques of TC structure from satellite microwave data toward improved intensity change prediction

Increasing observations of the environment and its assimilation as modeled variables during the past few decades have contributed to the significantly enhanced predictability of TC track, though no parallel gain in intensity prediction has been realized during the same time period. Nevertheless, Piech (2007) has shown through aircraft reconnaissance measurements that there are specific regimes of hurricane core structure as a function of intensity. Murray (2009) has further demonstrated predictive skill in forecasting TC intensity, using statistical analysis of such core observations, that can match and exceed guidance provided largely by environmental predictors (e.g. SHIPS). While prior work (e.g. Shapiro and Willoughby 1981) suggests that TC structure influences intensity tendencies, a lack of regular in-situ observations remains the greatest obstacle to operational intensity forecasting by inner-core information.

The creation of a TC structure climatology can provide the first steps toward a identifying specific structure-intensity relationships. To meet this challenge, a climatology of TC satellite data (HURSAT; Knapp 2008) is analyzed as a proxy for reconnaissance flight data. The use of satellites allows TCs in all basins to be observed at an average temporal resolution of around 6 hours for microwave polar orbiters, which is consistent with the average sampling interval from recon and may facilitate operational intensity prediction. In particular, structural patterns as diagnosed by microwave channels from SSM/I are explored as representations of physical processes within the TC core.

Particularly, new objective methods are developed to analyze sizes and symmetries of the eye, inner-core, and other structural patterns within the TC. With the help of ARCHER software (Wimmers and Velden 2010) to pinpoint the TC center using the satellite image itself, characteristic features of the TC are identified in relation to the center by transforming the satellite data into storm-relative polar coordinates. From this analysis and the resulting climatology, an aide to TC intensity prediction may be developed using patterns of TC structure with respect to location and intensity. As this climatology is expanded, validation of model performance may eventually reach beyond just track and intensity, but can include diagnostics of TC internal structure regimes as well.