Statistical Study on Tornadoes, Waterspouts, and Funnels Aloft in Okinawa Prefecture in Japan

Osamu Suzuki*, Hiroshi Niino**

and

Cooperative research group of Okinawa observatory 94-95***

 

*Senior Researcher, Meteorological Satellite and Observation System Research Department, Meteorological Research Institute, JMA, 1-1, Nagamine, Tsukuba, Ibaraku, 305-0052, Japan

Tel: +81-298-53-5379, Fax: +81-298-56-0644, e-mail: osuzki@mri-oma.go.jp

**Tokyo University, Japan

***Okinawa Observatory, JMA, Japan

 

A statistical study on tornadoes, waterspouts, and funnels aloft was made in Okinawa prefecture in Japan. Purpose of this study is to understand environmental conditions of the phenomena and to clarify relations between the phenomena and to see relations between the phenomena and atmospheric parameters.

According to the past statistical studies on tornadoes, average number of tornadoes per unit area is as large as that of Oklahoma state in US. Therefore, we chose Okinawa prefecture in Japan for this study. Numbers of events are 90 during 1961-94 for tornado and waterspout, and 120 during 1983-94 for funnel aloft, respectively. Numbers of events which proximity sounding were available are 36 for tornado and waterspout, and 34 for funnel aloft, respectively. In this statistics, funnel aloft is included because of similarity among the three phenomena, that is, they are vertical vortices in the real atmosphere and occur under convective clouds.

The results are summarized as follows:

  1. Synoptic conditions for Tornado/Waterspouts occurrences are in agreement with past studies
  2. Most of funnels aloft were observed in warm season. Many of them occurred near west-southwest edges of highs or in warm sector of low systems.
  3. Parent clouds of tornaodes generated in high SRH (Storm Relative Helicity integrated over 0 to 3 km) environments showed similar charateristics to those of miniature/bonsai supercells often developed in tropical cyclone environments.
  4. Significant difference in envionmental parameters are found between tornado events and funnel aloft:
    • CAPE for tornadoes varied over the interval 0-4200 J/kg. About 58% of events occurred with CAPE to be less than 1000 J/kg. On the other hand, CAPE for funnels aloft were distributed uniformly over the interval from 0 to 3500 J/kg.
    • SRH for tornadoes were distributed almost uniformly over the interval from 0 to 800 (m/s) 2, while those for funnels aloft did not exceed 90 (m/s)2.
    • Difference of equivalent potential temperature (Dqe), defiend by subtracting maximum from minimum equivalent potential temperature in vertical, for tornadoes were uniformly distributed over 0-40 K, while 89% of that for funnels aloft were distributed over 25-40 K.
    • Bulk Richardson Numbers (BRN) for tornado were smaller than that of for funnels aloft. About 67% of tornadoes occurred when BRN less than 50, while most of funnel aloft were observed when BRN greater than 50.