Tornado
A tornado is a rapidly rotating column of air in contact with both the surface of the Earth and a cumulonimbus cloud, generated primarily through baroclinic vorticity tilting and vertical stretching within supercell thunderstorms.
A tornado is a rapidly rotating column of air extending from the base of a convective cumulonimbus cloud to the ground, often made visible by a condensation funnel of water droplets and an ascending cloud of surface dust and debris. Representing the highest-vorticity wind phenomenon in the terrestrial atmosphere, tornadoes develop vertical vorticity values (ฮถ = โv/โx โ โu/โy) on the order of 0.1 to 1.0 sโ1โroughly four orders of magnitude greater than the planetary Coriolis parameterโwith horizontal wind speeds ranging from 29 m sโ1 (65 mph) in weak vortices to more than 135 m sโ1 (300 mph) in rare multiple-vortex events, accompanied by localized barometric pressure drops of 50 to 100+ hPa within the vortex core.
Although non-supercell tornadoes (such as landspouts and waterspouts) can form via the vertical stretching of pre-existing boundary-layer shear zones beneath developing cumulus congestus towers, the vast majority of damaging and long-track tornadoes are spawned by supercell thunderstormsโdeep, rotating convective storms sustained by a persistent mid-level updraft vortex known as a mesocyclone.
Synoptic Ingredients and Supercell Mesocyclogenesis
CAPE, Deep-Layer Wind Shear, and Storm-Relative Helicity
Supercell thunderstorms and tornadogenesis require the spatial coincidence of four synoptic and mesoscale thermodynamic ingredients across the troposphere:
- Boundary-Layer Moisture and Buoyancy (CAPE): High low-level specific humidity (surface dew points Td โฅ 16โ20 ยฐC) beneath steep mid-tropospheric lapse rates (7โ9 K kmโ1) generates large Convective Available Potential Energy (CAPE):
CAPE = โซ g [(Tv,parcel โ Tv,env) / Tv,env] dz
integrated from the Level of Free Convection (LFC) to the Equilibrium Level (EL). Typical tornadic environments exhibit CAPE values of 1,500 to 4,500+ J kgโ1, yielding theoretical maximum updraft velocities (wmax โ โ(2 ยท CAPE)) of 55 to 95 m sโ1.
- Low Lifting Condensation Level (LCL) and Manageable Convective Inhibition (CIN): A capping inversion (700โ850 hPa warm layer) prevents premature widespread cumulus convection during early afternoon heating until boundary-layer convergence breaks the cap. Meanwhile, a relatively low Lifting Condensation Level (LCL < 800โ1,200 m above ground) reduces sub-cloud evaporative cooling, preventing downdraft outflows from becoming overly cold and negatively buoyant.
- Strong Vertical Wind Shear and Storm-Relative Helicity (SRH): A rapid increase in wind speed and clockwise veering of wind direction with height (in the Northern Hemisphere) across the lowest 0โ6 km (bulk shear ฮV0โ6 km โฅ 20โ30 m sโ1) and lowest 0โ1 km generates horizontal environmental vorticity (ฯh = k ร โV/โz). The alignment between storm-relative inflow velocity (V โ C, where C is the storm motion vector) and horizontal vorticity is quantified by Storm-Relative Helicity (SRH) over depth h (typically 0โ1 km or 0โ3 km):
SRH = โ โซ (k ร (V โ C)) ยท (โV/โz) dz
When 0โ1 km SRH exceeds 150 to 300+ m2 sโ2, the storm ingests helical, streamwise vorticity that tilts directly into vertical updraft rotation (mesocyclogenesis), separating the updraft from the precipitation downdraft.
Three-Stage Dynamics of Supercell Tornadogenesis
While mid-level mesocyclone rotation (at 3โ7 km altitude) is readily explained by the upward tilting of environmental horizontal wind shear by the buoyant updraft, only ~15% to 25% of radar-detected supercell mesocyclones produce a ground-level tornado. Because air rising in a pure updraft is moving away from the ground, generating rotation at the surface requires a downdraft-assisted baroclinic mechanism (Markowski & Richardson, 2009):
- Stage 1 โ Forward-Flank and Rear-Flank Downdrafts (FFD and RFD): As precipitation falls out of the rotating supercell updraft, evaporative cooling of rain and melting of hail create two distinct downdraft regions: the Forward-Flank Downdraft (FFD) ahead of the updraft and the Rear-Flank Downdraft (RFD) wrapping cyclonically around the back (southwest) side of the mesocyclone. This wrapping precipitation curtain produces the classic hook echo signature on weather radar.
- Stage 2 โ Baroclinic Vorticity Generation and Near-Surface Tilting: Along the horizontal temperature and buoyancy gradient (โB) between the cool downdraft air and the warm, moist ambient inflow, solenoidal baroclinic torque (โ ร (B k) = โhB ร k) generates horizontal vorticity rings along parcel trajectories descending within the RFD and along the forward-flank gust front. As these air parcels descend very close to the ground and turn upward into the base of the updraft, the horizontal vorticity is tilted into the vertical within a few hundred meters of the surface.
- Stage 3 โ Vortex Stretching and Conservation of Angular Momentum: Once near-surface vertical vorticity (ฮถ > 0) is established underneath the main updraft and the upward-directed dynamic pressure perturbation (โp'dyn/โz < 0) of the overhead mesocyclone, horizontal convergence (โโh ยท v > 0) stretches the vertical air column according to the inviscid vertical vorticity equation:
Dฮถ/Dt = (ฮถ + f) (โโh ยท v) + (ฯh ยท โhw)
Because circulation ฮ = 2ฯ*rV*ฮธ is nearly conserved as converging rings of air contract from a mesocyclone radius of r โ 2,000 m down to a tornado core radius of r โ 50โ100 m, tangential wind speeds (Vฮธ) amplify by a factor of 10 to 20, forming a concentrated vortex in cyclostrophic balance (Vฮธ2/r = (1/ฯ) โp/โr). Crucially, if the RFD outflow is too cold (high evaporative cooling from a high LCL), the dense cold pool resists vertical lifting; tornadogenesis is favored when RFD parcels have only weak negative buoyancy ("warm RFD" regime).
Vortex Structure, Corner Flow, and the Enhanced Fujita Scale
Rankine Vortex, Corner Flow Collapse, and Suction Vortices
A simple combined Rankine vortex models a tornado with solid-body rotation inside the radius of maximum winds (Vฮธ โ r for r < Rmax) and irrotational potential decay outside (Vฮธ โ 1/r for r > Rmax). In reality, surface friction disrupts cyclostrophic balance in the lowest 10 to 100 meters: friction slows the tangential wind Vฮธ, reducing the outward centrifugal force while the inward radial pressure gradient remains strong. This imbalance drives a shallow, high-velocity radial inflow jet toward the central axis that erupts upward in the corner flow region, where the tightest wind speeds occur just 10 to 50 meters above the ground.
The topology of the vortex is governed by the swirl ratio (S), the ratio of tangential velocity to radial inflow velocity (S = r0 ฮ / (2Q), where Q is volumetric updraft flux):
- Low Swirl Ratio (S < 0.2): A single laminar-like central updraft jet exists throughout the column.
- Moderate Swirl Ratio (S โ 0.4โ0.8): An adverse axial pressure gradient develops aloft, triggering vortex breakdown where a central axial downdraft penetrates toward the surface and produces maximum near-ground wind intensification.
- High Swirl Ratio (S > 1.0): The central downdraft reaches the ground, expanding the tornado into a wide multiple-vortex tornado in which two to six smaller, intense sub-vortices (suction vortices) orbit inside the parent eyewall, adding their rotational speed to the parent translation and tangential speeds.
The Enhanced Fujita (EF) Scale and Doppler Radar Diagnostics
Because routine surface anemometers rarely survive direct hits by strong tornadoes, operational meteorological agencies classify tornado intensity post-event using the Enhanced Fujita (EF) Scale (adopted by the U.S. National Weather Service in 2007 to calibrate Tetsuya Theodore Fujita's 1971 scale). The EF scale assigns a 3-second gust wind estimate across 28 Damage Indicators (DIs) (such as single-family frame homes, engineered steel buildings, and hardwood trees), each evaluated across multiple Degrees of Damage (DoD):
| EF Rating | 3-Second Gust (mph) | 3-Second Gust (km hโ1 / m sโ1) | Share of US Tornadoes | Characteristic Structural Damage |
|---|---|---|---|---|
| EF0 (Light) | 65 โ 85 mph | 105 โ 137 km hโ1 (29 โ 38 m sโ1) | ~53% | Peels roof shingles, gutters, and vinyl siding; breaks large tree branches. |
| EF1 (Moderate) | 86 โ 110 mph | 138 โ 177 km hโ1 (38 โ 49 m sโ1) | ~31% | Strips roof decking, overturns mobile homes, shatters exterior windows. |
| EF2 (Significant) | 111 โ 135 mph | 178 โ 217 km hโ1 (50 โ 60 m sโ1) | ~11% | Tears complete roofs off frame houses, uproots or snaps large mature trees. |
| EF3 (Severe) | 136 โ 165 mph | 218 โ 266 km hโ1 (61 โ 74 m sโ1) | ~3.5% | Destroys upper stories and exterior walls of well-built houses; debarks trees. |
| EF4 (Devastating) | 166 โ 200 mph | 267 โ 322 km hโ1 (74 โ 89 m sโ1) | ~0.8% | Levels well-constructed frame homes to their foundations; lofted automobiles. |
| EF5 (Incredible) | > 200 mph | > 322 km hโ1 (> 90 m sโ1) | < 0.1% | Sweeps anchor-bolted frame homes clean from foundations; deforms steel frames. |
Real-time warning operations rely on dual-polarization S-band Doppler weather radar (NEXRAD WSR-88D). In addition to resolving the reflectivity hook echo and the radial velocity Tornado Vortex Signature (TVS)โan adjacent gate-to-gate couplet of strong inbound and outbound Doppler velocitiesโdual-polarization radar measures the co-polar correlation coefficient (ฯhv) and differential reflectivity (ZDR). When a tornado lofts irregular, non-meteorological surface debris (wood, insulation, leaves, soil), ฯhv drops sharply below 0.80 (compared to >0.97 in pure rain) colocalized with a high-velocity couplet, producing a definitive Tornadic Debris Signature (TDS) confirming a tornado on the ground even at night or inside rain-wrapped curtains.
Global Climatology and Anthropocene Spatial Shifts
Roughly 1,200 to 1,400 tornadoes are recorded annually in the United Statesโthe highest frequency on Earthโbecause North America's meridian-oriented Rocky Mountains lie directly west of the warm Gulf of Mexico. Southerly low-level jets transport warm, humid tropical air northward across the Great Plains and Mississippi Valley directly beneath dry, steep-lapse-rate elevated mixed layers advected eastward off the Mexican Plateau and Rockies, creating an ideal supercell incubator. Secondary global tornado corridors occur in the Rรญo de la Plata Basin (Argentina, Uruguay, and southern Brazil) east of the Andes, the GangesโBrahmaputra Delta (Bangladesh and eastern India), southern Canada, Europe, South Africa, and eastern Australia.
Observational trend analyses since 1979 (Brooks et al., 2014; Gensini & Brooks, 2018) identify two structural shifts in North American tornado activity during the Anthropocene:
- Temporal Clustering ("Fewer Tornado Days, Bigger Outbreaks"): While the mean annual number of EF1+ tornadoes in the United States has remained roughly stationary (~500 yrโ1), the number of calendar days per year with at least one EF1+ tornado has decreased (from ~150 days yrโ1 in the 1970s to ~100 days yrโ1), while the number of days with โฅ20 or โฅ30 tornadoes has increased significantly.
- Eastward Shift from the Great Plains toward the Mid-South ("Dixie Alley"): Decadal frequency maps of Significant Tornado Parameter (STP) environments and confirmed tornado tracks show a statistical decline across the dry western High Plains (Texas, Oklahoma, Kansas) and a rising trend across the densely populated, forested Mid-South, Mississippi Valley, and Ohio Valley (Mississippi, Alabama, Tennessee, Kentucky, Illinois, Indiana), where nocturnal and cool-season high-shear/low-CAPE tornadoes pose heightened societal vulnerability.
Recent Events
In May 2013, spawned on May 31 over central Oklahoma inside an environment with extreme buoyancy (CAPE > 4,500 J kgโ1), the El Reno, Oklahoma multiple-vortex tornado expanded to a surveyed damage path width of 4.2 kilometers (2.6 miles)โthe widest tornado in recorded history. Mobile RaXPol radar measured instantaneous sub-vortex wind speeds exceeding 135 m sโ1 (302 mph) at <100 m above ground, though because the core tracked across open rural farmland without striking dense structural damage indicators, it was officially rated EF3 under the damage-based EF scale, catalyzing meteorological efforts to integrate mobile radar wind measurements into intensity standards [4].
In December 2021, driven by an unseasonably warm, moist Gulf of Mexico air mass (Td โ 18โ21 ยฐC) surging into the Mississippi and Ohio Valleys ahead of a deep synoptic trough, a nocturnal long-track supercell tracked over 400 km across Arkansas, Missouri, Tennessee, and western Kentucky during the night of December 10โ11. The primary Western Kentucky EF4 tornado (estimated peak winds 85 m sโ1 / 190 mph) remained continuously on the ground for 266.7 km (165.6 miles), devastating Mayfield and Dawson Springs and lofting debris to 11.3 km (37,000 ft) altitude on NEXRAD Tornadic Debris Signature scans, demonstrating the hazard of cool-season nocturnal outbreaks in the Mid-South [7]. Satellite shortwave-infrared observations later showed that the first violent tornadoes of the outbreak, over cropland in northeastern Arkansas, left soil-moisture scars that were not captured by visible and near-infrared channels [8].
In May 2024, on May 21, an EF4 tornado struck Greenfield, Iowa, damaging wind turbines and homes; a subsequent engineering failure analysis examined how the wind turbines failed in the storm [9].
Frequently Asked Questions
How can a tornado produce 300 mph winds on mobile Doppler radar yet receive an EF3 or EF4 rating instead of EF5?
The operational Enhanced Fujita (EF) Scale is a structural damage scale, not a direct anemometer wind measurement scale. Official National Weather Service survey teams assign an EF rating based strictly on the most severe Degree of Damage observed across 28 calibrated engineering Damage Indicators (such as well-built single-family homes or structural steel buildings). Because more than 95% of land area in the central United States consists of open agricultural fields, crops, and unengineered outbuildings, a tornado with 300 mph mobile-radar winds that tracks over open farmland (such as the 2013 El Reno tornado) or strikes only smaller residential structures for a fraction of a second as a sub-vortex sweeps past (such as the 2024 Greenfield tornado) may leave behind no surviving engineering indicator that proves >200 mph three-second structural loading at roof height.
Why do mobile homes and highway overpasses offer poor shelter during a tornado?
Even weak-to-moderate tornadoes (EF1 to EF2, 86โ135 mph) generate aerodynamic uplift forces on roofs and walls that exceed the weight and tie-down anchor strength of manufactured mobile homes, causing the structure to roll or disintegrate at wind speeds that well-anchored permanent frame homes survive with minor roof damage. Highway overpasses are also hazardous shelters: as tornado winds pass beneath the bridge deck, the Venturi constriction effect accelerates wind velocity through the narrow opening while the elevated girder beams provide no protection against high-velocity horizontal flying debris, which is the primary cause of tornado injuries and fatalities.
Why does opening windows in a house before a tornado strikes not protect the building from "exploding"?
An early twentieth-century myth held that the 50โ100 hPa barometric pressure drop inside a tornado core caused sealed houses to explode outward from internal air pressure. Modern structural wind engineering proved that normal residential buildings have sufficient air leakage through soffits, vents, and chimneys to equalize static pressure drops within seconds. Instead, houses fail aerodynamically: extreme winds blowing over the roof create upward Bernoulli lift while wind entering through a broken upwind window or garage door pressurizes the interior of the house, adding an upward internal push beneath the roof rafters that doubles the net lifting force and peels the roof off. Opening windows actually allows wind to rush inside and accelerates structural roof failure.
References
- Davies-Jones, R. (1984). "Streamwise vorticity: The origin of updraft rotation in supercell storms." Journal of the Atmospheric Sciences, 41(20), 2991โ3006. https://doi.org/10.1175/1520-0469(1984)041<2991:SVTOOU>2.0.CO;2
- Markowski, P. M., & Richardson, Y. P. (2009). "Tornadogenesis: Our current understanding, forecasting considerations, and questions to guide future research." Atmospheric Research, 93(1โ3), 3โ10. https://doi.org/10.1016/j.atmosres.2008.09.015
- Wurman, J., Kosiba, K., & Robinson, P. (2012). "In situ, Doppler radar, and video observations of the interior structure of a tornado at very close range." Bulletin of the American Meteorological Society, 93(6), 835โ848. https://doi.org/10.1175/BAMS-D-12-00114.1
- Wurman, J., Kosiba, K., Robinson, P., & Marshall, T. (2014). "The role of multiple-vortex structure in driving tornado debris and damage in the 31 May 2013 El Reno, Oklahoma, tornado." Bulletin of the American Meteorological Society, 95(1), 31โ45. https://doi.org/10.1175/BAMS-D-13-00221.1
- Brooks, H. E., Carbin, G. W., & Marsh, P. T. (2014). "Increased variability of tornado occurrence in the United States." Science, 346(6207), 349โ352. https://doi.org/10.1126/science.1257460
- Gensini, V. A., & Brooks, H. E. (2018). "Spatial trends in United States tornado frequency." npj Climate and Atmospheric Science, 1, 38. https://doi.org/10.1038/s41612-018-0048-2
- Pilguj, N., Taszarek, M., Allen, J. T., & Hoogewind, K. A. (2022). "Are trends in convective parameters over the United States and Europe consistent with tornado and severe thunderstorm observations?" Journal of Climate, 35(11), 3605โ3626. https://doi.org/10.1175/JCLI-D-21-0135.1
- Wang, J., Lin, Y., McFarquhar, G. M., et al. (2023). "Soil moisture observations from shortwave infrared channels reveal tornado tracks: A case in 10โ11 December 2021 tornado outbreak." Geophysical Research Letters, 50(6), e2023GL102984. https://doi.org/10.1029/2023GL102984
- Aslam, M. U., & Alipour, A. (2026). "Structural failure analysis of wind turbines in 2024 Greenfield tornado." Engineering Failure Analysis, 186, 110387. https://doi.org/10.1016/j.engfailanal.2025.110387