Cloud Resolving Hurricane prediction using sssimilated Doppler radar based in space

Hurricane intensity forecasts have improved only modestly over the past 3 decades while track forecasts have steadily improved. Such forecasts are necessary if false alarm rates causing unnecessary evacuation are to be reduced significantly, and if flow surge patterns and damage patterns are to be predicted with sufficiently more accuracy than they are now. A major reason why intensity change prediction has not improved significantly is because of the dependence of intensity forecasts on the energy-containing structures internal to the larger storm vortex. The paradigm for extra-tropical numerical prediction , whereby energy is drawn from time and space scales larger than the storm simply doesn't hold for tropical storms. Instead, internal storm structures are driven from energy released on sub storm scales not resolved by traditional observations. These sub-storm-scale features take the form of internal and external rainbands, primary and secondary eyewalls, embedded convective storms and other asymmetrical moisture and thermal structures. Moreover, because these structures are near or less than the Deformation radius, the most important observational feature defining these structures is typiucally the momentum field, rather than the mass (or thermal) structure. In addition, important storm linkages between internal structures and external features such as a synoptic forcing and resistance patterns exhibit equally elaborate structural variation and temporal behavior. These structural features typically cannot be resolved by traditional observational platforms both because of insufficient spatial resolution and insufficient temporal resolution. Moreover, aircraft reconnaissance also cannot capture their essence because of their inability to observe the entire system of sufficient frequency across the entire storm and surrounding region. Nevertheless, research model simulations of tropical storms suggest that models are able to simulate these structures in a realistic manner. Interestingly, model resolution of at least 1.25 km horizontal resolution has been shown to be necessary to capture some of the most important of these features, such as eye wall replacement cycles, that determine intensity and intensity structure. This reality creates the dimema: We cannot improve intensity forecasting significantly given our current and projected observation paradigm. However, recent technological advances have shown that it is possible to change this paradigm. It has been shown that it is now possible to place a Doppler radar in geostationary orbit sufficient to take fully three-dimensional observations of the wind and hydrometeor structure in regions where condensate or particulate matter exists. Fortunately, almost all small scale structures in the tropical atmosphere are associated with suspended hydrometeors opening the possibility of cloud resolving simulation of space boprne Doppler radar as the near perfect solution to tropuical weather prediction. In this talk we will discuss the problems with the current paradigm and experiments conducted over the past 3 years showing that the assimilation of space-borne Doppler radar will provide the data necessary to release us from the paradigm of numerical weather prediction in which we are now trapped.