A preliminary result of statistics for meso-vortex-signatures in Japan detected by MRI-MDA

JMA started to issue a new tornado-related hazardous wind watch in March 2008. This watch is issued based on atmospheric indices calculated with GPV-MSM (Grid Point Value-Meso Scale Model) and radar observations. To improve accuracy of the watch, we are now developing a mesocyclone detection algorithm and making case studies of the hazardous events.

  A preliminary result of evaluation of our new mesocyclone detection algorithm (MRI-MDA) using real data will be presented. We applied our algorithm to one-year data taken from four terminal Doppler weather radars in Japan. Although not all detected meso-vortex-signatures were mesocyclones, statistics on meso-vortex-signatures is very useful to know algorithm performance. It is found that meso-vortex-signatures were detected over 30 days in the year. Core diameters, rotational velocities, and depth of rotations showed similar characteristics of those found in mini-supercells in the US. There were four tornadoes within the observation area in the 2004. Mesocyclone were observed prior to two tornadoes, and simultaneously for one tornado, and not observed for one tornado. Lead times of the two tornadoes were 17 and 35 minutes respectively.