PhD Defense - Sam Trahan
Sam successfully defended his PhD dissertation on July 28, 2011.
Multi-Scale Analysis of Observations of Tropical Cyclones with Applications to High-Resolution Hurricane Modeling
Numerical modeling of tropical cyclones is one of the primary forecasting tools used in predicting tropical cyclone track and intensity. There have been improvements in track forecasting during the past decade, but the skill of intensity forecasts has not improved. It has recently become clear that hurricane forecast models may need to go to higher spatial resolution in order to correctly represent small scale processes that are now thought to play a role in hurricane intensity. The research described in this thesis makes contributions to the problem of intensity forecasting through a combination of high resolution hurricane model development and statistical analysis of observations of hundreds of tropical cyclones. The contributions to model development include the extension of an operational forecast model to 3km resolution, a scale where convection begins to be resolved and convective parameterizations become unnecessary. With higher spatial resolution, it becomes important to find new ways of characterizing the statistical properties of the observed wind and mass fields on a range of scales, in order to evaluate whether the model is producing realistic small scale structure. Statistical methods are developed for analyzing in-situ and satellite observations of tropical cyclones that provide guidance for diagnosing small scale variability in hurricane models over a range of scales. Some statistical characteristics were found to be "universal" in that they were found to be the same for most storms, evidence was found for scale invariance, and statistical signatures of small scale turbulence were identified. In addition, a satellite data analysis reveals a connection between storm intensity and height of convection. This work lays the groundwork for some new directions in future investigations into the small scale dynamical processes that play a role in TC intensity.