Elegant Graphs: Tetsuya Fujita’s Tornado Scale

In the April 1972 issue, National Geographic Magazine wrote, “His intensive studies during the past 20 years of the behavior of these devastating funnels have earned him the sobriquet, Mr. Tornado”.

Dr. Tetsuya “Ted” Fujita began his work analyzing the damage at Hiroshima and Nagasaki. From the rubble pattern he calculated the height at which the bombs had exploded. After the bombs, he pursued meteorology in Japan. He made careful observations of thunderstorms recording wind velocity, temperature and pressure and began writing papers on the subject. At age of 33 he moved to the University of Chicago. He lived to be 78 and along the way laid the foundation for the modern science of tornado meteorology.

One of his major contributions was the development of the Fujita Scale for Tornadoes. Fujita combined his careful analysis of damage and wind velocities for a scale to classify tornadoes. It has been modified by the United States National Oceanic and Atmospheric Administration (NOAA) and the current standard is the Enhanced Fujita Scale (“EF”).

The study of tornadoes has always attracted a wide range of aficionados — from the storm chasers who stream live videos to researchers who study the thermodynamics and hydrodynamics of wind shear with laboratory models and numerical simulations. In between, we have a labor of love by Tom Grazulis who created the Tornado Project and has collected data on 60,000 US tornadoes since 1953 and made this data available online.

From this data we plotted the average number for deaths for each of the five levels on the Fujita Scale. The result was an elegant plot:

Here is a graph, that captures the devastation of the tornadoes with elegant simplicity. It is as if he had this graph in front of him as he decided what the scale ought to be; one can just marvel at the extraordinary insight by Dr. Tetsuya Fujita.

Sixty three years of data fits a forty year old classification. No matter what our field of study, a graph such as this one, delights those of us who seek pleasure in data analysis and find such correlations.