Comparisons of Nowcasting Techniques for Oceanic Convection

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner
Tuesday, 19 January 2010: 1:30 PM
B314 (GWCC)
Huaqing Cai, NCAR, Boulder, Colorado; and C. Kessinger, N. Rehak, D. Megenhardt, and M. Steiner

The importance of oceanic convection nowcasting to the aviation community has been highlighted after several recent oceanic aviation incidents/accidents. As one component of the Oceanic Convection Diagnosis and Nowcasting system, the nowcasting system produces short-term predictions of future locations of oceanic convection. Considering possible storm impacts on aviation, identifying the best nowcasting system for convection over the ocean seems imperative.

There are a number of different nowcasting techniques available at the National Center for Atmospheric Research (NCAR); however, a complete comparison study of those techniques has not yet been done. In this paper, we are focused on four different nowcasting techniques, namely, persistence, the Thunderstorm Identification, Tracking, Analysis and Nowcasting (TITAN) object-based tracking system, gridded forecasts created using a motion vector field derived from TITAN and the Global Forecast System (GFS) numerical model steering level winds, and the random forecast technique. The persistence forecast serves as a base line for the other three techniques. Traditional verification scores such as the Critical Success Index (CSI) are used in the process of evaluation. Each technique will be described in detail and their pros and cons discussed. Based on the result of this evaluation, one of these four techniques will be implemented in the realtime Oceanic Convection Diagnosis and Nowcasting system. It is anticipated that some of the nowcasting convective products will be uplinked to the cockpit in the future.