Tornadoes are measured primarily by examining the damage they leave behind, not by directly recording wind speeds inside the funnel. Because placing instruments in the path of a tornado is extremely difficult and dangerous, meteorologists rely on a combination of post-storm damage surveys, radar data, satellite imagery, and occasionally ground-level sensors to estimate a tornado’s intensity, size, and path.
The Enhanced Fujita Scale
The standard rating system for tornadoes in the United States is the Enhanced Fujita (EF) Scale, which became operational on February 1, 2007. It assigns a tornado a rating from EF0 to EF5 based on estimated three-second wind gusts and the damage those winds produce:
- EF0: 65–85 mph. Minor damage to chimneys, signs, and shallow-rooted trees.
- EF1: 86–110 mph. Moderate damage. Roof surfaces peeled off, mobile homes overturned.
- EF2: 111–135 mph. Considerable damage. Roofs torn from frame houses, large trees uprooted.
- EF3: 136–165 mph. Severe damage. Entire stories of well-built homes destroyed, heavy cars thrown.
- EF4: 166–200 mph. Devastating damage. Well-constructed houses leveled, structures with weak foundations blown away.
- EF5: Over 200 mph. Incredible damage. Strong frame houses swept from foundations, steel-reinforced concrete structures critically damaged.
The original Fujita Scale, introduced in 1971, had known problems with overestimating wind speeds at higher ratings. The Enhanced version corrected this by tying wind estimates to specific types of structures and the observable stages of damage to each one.
How Damage Surveys Work
The core of tornado measurement is the post-storm damage survey. When a tornado strikes, National Weather Service meteorologists travel to the affected area and systematically catalog what was damaged, how severely, and where. For significant events (those causing fatalities, numerous hospitalizations, or extensive property damage), teams conduct both ground-level and aerial surveys.
Surveyors use a system of Damage Indicators and Degrees of Damage. A Damage Indicator is a type of structure or object: a one-story home, a strip mall, a metal building, a hardwood tree. Each indicator has a set of Degrees of Damage describing progressively worse outcomes, from threshold damage to total destruction. By matching what they see on the ground to these predefined categories, surveyors estimate the wind speed at that location.
Since 2021, the NWS has used an operational digital tool called the Damage Assessment Toolkit. It includes a mobile app built on geographic information system software that lets surveyors log geo-located damage photos and metadata in the field. This data feeds into a centralized system, making it easier to map the tornado’s path and compare observations across survey teams.
Measuring Path Length and Width
Beyond intensity, every documented tornado gets two physical measurements: path length and maximum path width. Path length is the total distance from where the tornado first touches down to where it lifts. Width is the widest extent of damage at any point along that path. Lengths are recorded to the nearest hundredth of a mile, and widths to the nearest yard, though they’re frequently rounded to the nearest 5 or 10 yards.
These measurements come from a combination of radar tracking, eyewitness accounts, media reports, damage photos, and the ground and aerial surveys themselves. In rural or remote areas where ground access is limited, high-resolution satellite imagery fills the gaps. Sensors like the Sentinel-2 satellite can capture multispectral images at 10-meter resolution, which is detailed enough to trace vegetation damage across farmland and forests. This approach has revealed tornado path features that aren’t always apparent from the ground, including subtle shifts caused by local terrain.
Radar Measurement
The nationwide network of fixed Doppler radar stations (WSR-88D) detects rotation in thunderstorms and can identify tornado signatures. But these radars sit at fixed locations, and their beams rise higher above the ground the farther away a storm is. At long range, they can miss what’s happening near the surface, which is exactly where tornado winds are strongest.
Mobile Doppler radar units solve this problem. These truck-mounted systems can be driven into position near a developing storm and scan the lowest levels of the atmosphere. NOAA’s National Severe Storms Laboratory operates mobile radars, including a dual-polarized X-band system called NOXP that operates on a 3-centimeter wavelength. Mobile radars have captured some of the highest wind speed readings ever associated with tornadoes, providing direct velocity data that supplements the damage-based EF rating.
Radar doesn’t measure wind at the ground itself. It measures the speed of raindrops, debris, and other particles carried by the wind at whatever height the radar beam intersects the storm. This means radar-derived wind speeds are estimates of what’s happening aloft, which may differ from conditions at ground level.
Ground-Level Sensors
Getting instruments directly into a tornado’s path is the holy grail of tornado measurement, and it has been done, though rarely. Since 2002, researchers have deployed hardened probes designed to survive a direct hit. These small, heavy devices sit on the ground in a tornado’s projected path and record temperature, pressure, and relative humidity at high speed, typically 10 readings per second.
Over nine tornado intercepts, these probes recorded peak pressure drops ranging from 5 to 100 hectopascals, a range that spans from a modest tug to a dramatic pressure plunge equivalent to what you’d experience ascending thousands of feet in seconds. Instrumented chase vehicles (mobile mesonets) positioned near the tornado’s edge have measured wind speeds of 40 to 50 meters per second, roughly 90 to 112 mph. Researchers recommend a sampling rate of at least 10 readings per second to capture the rapid pressure and wind fluctuations inside a tornado’s circulation.
These direct measurements are valuable for research, but they’re rare and opportunistic. A probe only records data if the tornado passes directly over it, making this method impractical for routine measurement.
Infrasound Detection
Tornadoes produce sound waves below the range of human hearing, called infrasound. Research has shown that these low-frequency acoustic signals begin during tornado formation and continue through the storm’s life. One measurement from a tornado in Lakin, Kansas picked up an elevated broadband signal between 10 and 15 Hz.
Infrasound-based systems have the potential to detect and locate tornadoes even when radar coverage is limited, such as in mountainous terrain or at long range. This technology is still largely experimental, but it represents a fundamentally different approach: instead of measuring damage after the fact or scanning the atmosphere with radar, it listens for the tornado’s own acoustic signature.
The T-Scale Used Outside the US
The EF Scale is the standard in the United States, but the UK and parts of Europe use the TORRO scale (T-scale), which runs from T0 to T10 and covers a wider range of wind speeds. A T0 tornado starts at 39 mph, a T2 at 73 mph, and a T7 reaches 187–212 mph. The T-scale also incorporates typical track length, width, and area for each intensity level, making it a more integrated measure of tornado size and strength. In practice, most tornadoes outside the US are weaker than those in Tornado Alley, so the lower end of the T-scale sees the most use.
Canada adopted a version of the EF scale in 2013, and other countries are developing their own adaptations. The challenge everywhere is the same: tornado measurement depends heavily on what structures and vegetation happen to be in the path. A powerful tornado that crosses open farmland with nothing to damage may receive a lower rating than its true wind speed warrants, simply because there was nothing to survey.