Enhanced Fujita Scale

A scale used since February 2007 for categorizing tornados by estimates of wind speed based on the damage done. This scale took the place of the Fujita scale, on which it was intentionally based in order to preserve comparability with existing tornado databases. Values on the Enhanced Fujita Scale are stated as “EF numbers,” from EF0 to EF5. Each value is associated with a range of wind speeds as shown in the table below. To avoid giving the appearance of unjustified precision, the developers of the new scale rounded off wind speeds to the nearest speed divisible by 5 (unlike the Fujita scale). (In converting the miles per hour values to km/h and m/s for the table, we've made slight fudges to avoid gaps created by rounding.)

Enhanced F Scale				Fujita Scale (obsolete, for comparison)
EF Number	3-second Gust			F Number	Fastest ¼ mile wind
EF Number	miles per hour	km per hour	m/s	F Number	mph	km/hr	m/s
EF0	65-85	105-137	29-38	F0	40-72	64-116	18-32
EF1	86-110	138-177	38-49	F1	73-112	117-180	33-50
EF2	111-135	178-218	50-60	F2	113-157	181-253	51-70
EF3	136-165	219-266	61-73	F3	158-207	254-334	71-92
EF4	166-200	267-322	74-89	F4	208-260	335-419	93-116
EF5	over 200	over 322	>89	F5	261-318	420-512	117-142

Why a new scale?

Some of the problems with the Fujita scale:

Tornados rated F4 and F5 often did not reach the wind speeds the Fujita scale claimed¹.
Since the estimates were based primarily on damage to structures, it wasn't possible to make an estimate in locations where there were no structures.
The estimates didn't take into account the differing fragility of various structures clearly enough.

1. J. E. Minor, J. R. McDonald, and K. C. Mehta.
The tornado: An engineering oriented perspective.
NOAA Technical Memorandum, ERL NSSL-82.
Norman, OK : National Severe Storms Laboratory, 1977.

L. T. Phan and E. Simiu.
The Fujita tornado intensity scale: a critique based on observations of the Jarrell tornado of May 27, 1997.
NIST Technical Note 1426.
Gaithersburg, MD: U.S. Department of Commerce, 1998.

How the new scale works

The first step in revising the scale was creating classes of clearly and narrowly defined objects a tornado might damage. These are called Damage Indicators. The first version of the scale has 28 such indicators, including such things as automobile showrooms, motels, one or two family residences, high schools, transmission line towers and hardwood trees¹.

For each Damage Indicator, experts defined degrees of damage (DOD). The DOD's represent visible steps from scarcely observable damage to utter destruction. Different Damage Indicators have different numbers of degrees of damage; 5 for hardwood trees, for example, and 10 for one- or two-family homes. For each Degree of Damage three wind speeds are given: the expected speed, a lower bound and an upper bound. Here, for example, are the Degrees of Damage (DOD) for single-wide manufactured homes.

DOD	Damage description	Expected wind speed, mph	Lower bound, mph	Upper bound, mph
1	Threshold of visible damage	61	51	76
2	Loss of shingles or partial uplift of one-piece metal roof covering	74	61	92
3	Unit slides off block piers but remains upright	87	72	103
4	Complete uplift of roof; most walls remain standing	89	73	112
5	Unit rolls on its side or upside down; remains essentially intact	98	84	114
6	Destruction of roof and walls leaving floor and undercarriage in place	105	87	123
7	Unit rolls or vaults; roof and walls separate from floor and undercarriage	109	96	128
8	Undercarriage separates from unit; rolls, tumbles and is badly bent	118	101	136
9	Complete destruction of unit; debris blown away	127	110	148

An observer at the site can assign any speed between the upper and lower bound, if circumstances warrant it. For example, suppose a trailer is found on its side. That is DOD 5, with an expected wind speed of 98 mph. But suppose the inspector discovers the trailer had unusually substantial ground anchors. He or she might then assign a higher speed, anything up to 114 mph. On the other hand, if the ground anchors were weak or nonexistent, he or she might assign a speed of 90 mph or even lower, down to 84 mph.

In assessing a tornado, the inspectors try to find more than one Damage Indicator. These can be averaged and compared to the table at the top of the page to assign an EF number to the tornado.

Ideally, the correlations between wind speed and degrees of damage would be done by, for example, testing structures in a wind tunnel. That was far beyond the means of those revising the scale, and so instead they resorted to a Delphi-like process in which they polled panels of tornado-watching experts.

1. For a complete list of the damage indicators, see www.spc.noaa.gov/faq/tornado/ef-scale.html

for further reading

The key document, with color photographs depicting some degrees of damage for some of the damage indicators:

A Recommendation for an Enhanced Fujita Scale (EF-Scale). Revision 2.
Lubbock (TX): Wind Science and Engineering Center, Texas Tech University, 10 October 2006.

Available on the web at www.depts.ttu.edu/nwi/Pubs/FScale/EFScale.pdf

NOAA has a web page devoted to the Enhanced Fujita Scale.

www.spc.noaa.gov/efscale/

The website of the National Wind Institute at Texas Tech, Fujita’s final home base, has more information on the scale and its application than anywhere else. Unfortunately, as far as the Fujita scales go, it is poorly organized and requires heavy use of the site's search function.

www.depts.ttu.edu/nwi/Pubs/FScale/Fujita.php

A video on development of panels to resist lumber flung by EF5 winds:

http://vimeo.com/48300343

Thoughts on improving the EF scale:

Roger Edwards and Harold E. Brooks.
Possible impacts of the enhanced Fujita scale on United States tornado data.
Preprints, 25th Conference on Severe Local Storms (2010)
available online at http://www.spc.noaa.gov/publications/edwards/eftrends.pdf

Roger Edwards, James G. LaDue, John T. Ferree, Kevin Scharfenberg, Chris Maier and William L. Coulbourne.
Tornado intensity estimation. Past, present, and future.
Bulletin of the American Meteorological Society, vol. 94, no. 5, pages 641-653 (May 2013).
doi:10.1175/BAMS-D-11-00006.1