Michael Dixon and Gerry Wiener
Research Applications Program, National Center for Atmospheric Research, Boulder, Colorado
Journal of Atmospheric and Oceanic Technology
Vol. 10, No. 6, December 1993
pp. 785-797
A methodology is presented for the real-time automated identification, tracking and short-term forecasting of thunderstorms based on volume-scan weather radar data. The emphasis is on the concepts upon which the methodology is based. A "storm" is defined as a contiguous region exceeding thresholds for reflectivity and size. Storms defined in this way are identified at discrete time intervals. An optimization scheme is employed to match the storms at one time with those at the following time, with some geometric logic to deal with mergers and splits. The short-term forecast of both position and size is based on a weighted linear fit to the storm track history data. The performance of the detection and forecast were evaluated for the summer 1991 season, and the results are presented.
This methodology provides the framework necessary to identify storms within three-dimensional radar data and to track them as physical entities. The storm and track data are suitable for scientific analysis for the purposes of both understanding and forecasting the physics of storm development and movement.
The method was successfully applied during real-time operations for a summer season in Colorado, and the human observers felt that typically it performed well. The accuracy of the forecasts is encouraging and is comparable with that of human-based forecasts that take additional factors such as low-level convergence and storm initiation into account. The intention is to use the system to assist forecasters in future field projects, who will in turn provide feedback on how well the system performs, and what enhancements should be made.