The Japanese Enhanced Fujita Scale: Its Development and Implementation

              Tornado is a small-scale and short-lived phenomenon which is rarely detected by ordinary surface weather observation networks. When gusty winds cause some damage, the Japan Meteorological Agency (JMA) dispatches the JMA Mobile Observation Team (JMA-MOT) to the damage sites for investigating the damage situations and collecting evidences to identify the phenomenon and rate its intensity.

              Conventionally, the intensity of tornadoes in Japan has been rated in terms of the Fujita scale (F-scale), which is commonly implemented in many countries all over the world. Recently, the Enhanced Fujita scale (EF-scale) was developed and is currently in use in the United States, and a slightly modified version started to be used in Canada. These EF-scales divide damage descriptions into two factors: Damage Indicators (DIs) and Degrees Of Damage (DODs). Combinations of these factors are used to estimate the wind speed and rate the intensity of a tornado. Since both the F-scale and the EF-scale were developed based on damages to buildings or trees in the United States (or Canada), however, it was desired to develop a new EF-scale that is based on the features of Japanese structures (buildings, cars, trees and other damage indicators) so as to rate the wind speeds more accurately in Japan. To this end, JMA organized “The Advisory Committee for Rating Intensity of Tornadoes” (2013-2015) to improve the conventional F-scale, which consisted of nine experts on wind engineering and meteorology. The committee developed the Japanese Enhanced Fujita scale (JEF-scale) in December 2015 and JMA started to implement this scale from April 2016.

              DIs for JEF-scale include structures such as Japanese-style wooden residential houses or shops, vending machines, and light vehicles. For each DI, several DODs were defined based on the state-of-the-art knowledge of wind engineering. Wind speeds corresponding to the DIs/DODs are now estimated within 5 m/s, which is more accurate than those for F-scale.

              To maintain the statistical continuity with the F-scale, JEF-scale classification was determined in the way that the rating of a certain phenomenon results in the same class as that by the F-scale as much as possible. 215 damage events among the previous disasters caused by tornadoes were used to obtain the correlation between wind speeds estimated by the F-scale and those by DIs and DODs using a regression analysis to a power function, which was also adopted in the Canadian EF-scale.

              The rating procedure using JEF-scale is made in the following way: 1) Determine DI and DOD for each damaged structure on the damage sites. 2) Derive the wind speed corresponding to the combination of DI and DOD. 3) Select the maximum value among the wind speed values derived in 2). 4) Determine the JEF-scale class by comparing the value derived in 3) with the wind speed ranges of its classification.

              The implementation of the JEF-scale enables more accurate ratings of the tornado intensity. The JEF-scale will contribute to the studies on the mechanism of tornado, improvement of its forecast, and disaster prevention through the accumulation of accurate tornado data including intensities, sizes, frequencies, geographical distribution and so on. An English version of “Guideline for the Japanese Enhanced Fujita Scale,” which gives a detailed description of the JEF-scale, is in preparation and will be uploaded on the JMA web site in due course. We expect that the JEF-scale provides a useful tool for comparing intensities of tornadoes in Japan with those in the world.