INCREASING THE USEFULNESS OF A MESOCYCLONE CLIMATOLOGY

The advent of the WSR-88D Doppler radar and its associated vortex detection algorithms has had a positive impact on severe weather forecasting. These algorithms, while intended for use in real-time situations, also provide a beginning point for the large-scale, longterm quantitative study of the detected phenomenon. Here, mesocyclone detections from a realization of the NOAA National Severe Storms Laboratory (NSSL) Mesocyclone Detection Algorithm (MDA) (Stumpf et al. 1998) are used to illustrate the opportunities, the challenges, and some possible remedies in producing a mesocyclone climatology. As examples of the challenges to be addressed in assembling such a climatology, radar beam geometry and inherent characteristics limit even a “perfect” detection algorithm. These limits include larger bin sizes (as range from the radar increases, resulting in lower resolution data), ground clutter, and spurious echoes due to Anomalous Propagation (AP) during certain atmospheric conditions. Proper dealiasing of the radial velocity data is especially important in making accurate mesocyclone detections. Unfortunately, the current clutter filters and dealiasing algorithms leave much to be desired in this regard. Often, spurious data from ground clutter returns are dealiased to produce false circulations that are identified by the MDA as strong mesocyclones. While a human forecaster usually recognizes these spurious detections and discounts them during real-time operations, a study involving climatologies synthesized automatically from algorithm output must find other means of dealing with these erroneous detections. Past work involving the creation of mesocyclone climatologies utilizing the MDA has been conducted by Mitchell et al. (2000), the results of which reveal similar challenges. The work reported here differs from this earlier study by attempting to improve the quality of the mesocyclone data sets. Several postMDA filtering techniques have been developed to aid in the removal of false mesocyclone detections. These techniques are presented here along with some initial results. 2. DATA & METHODOLOGY