In short: wind shear is the change in wind speed and direction with height. It is the difference between a storm that rains itself out in half an hour and one that lasts for hours. Deep shear, through the lowest 6 km, decides whether supercells are possible. Shear near the ground, measured as storm-relative helicity, decides how much low-level spin those supercells can work with.

If CAPE is the fuel, shear is the steering and the spin. Storms in weak shear go straight up, and the rain and cold air falling out of them smother their own updraft. Storms in strong shear lean. Their rain falls beside the updraft instead of into it, so the updraft keeps breathing warm air.

How shear makes spin

1 2 3 4
  1. 1
    Wind near the ground. Slowed by friction.
  2. 2
    Wind about 6 km up. Much stronger, and on most severe days blowing from a different direction too.
  3. 3
    Horizontal spin. Faster air on top rolls the layer over, like a pencil between two hands. Meteorologists call it horizontal vorticity.
  4. 4
    Bulk shear. The difference between the top and bottom wind, direction included. This side view shows only the speed part.
Shear stores spin in the air before any storm exists. A strong updraft that rises into this rolling layer tips the spin upright, which is where a supercell's rotation comes from (more in Supercells & Mesocyclones).

That rolling layer is the key idea. When an updraft punches into it, the updraft lifts the middle of the tube and tips it on end, and the storm starts to rotate. The step-by-step version is in Supercells & Mesocyclones.

Bulk shear: the deep-layer number

Bulk shear, which the SPC calls bulk wind difference, is the wind at the top of a layer minus the wind at the bottom, treated as an arrow, so a change in direction counts as well as a change in speed. The standard layer is the ground to 6 km (about 20,000 feet), roughly the lower half of a severe storm.

What the numbers usually mean, as a rough guide:

0 to 6 km shearTypical storms
Under about 20 ktOrdinary single cells: up, rain, gone, in under an hour
About 20 to 35 ktMulticell clusters and lines, with new cells growing on the edges of old ones
35 to 40 kt and upSupercells become common (SPC’s guide), if storms stay separate

In MesoRadar: the National tab, group HRRR Shear & Rotation, Bulk Shear 0 to 6 km (Pro). The Skew-T tool lists Shear 0-1 / 0-6 for the point you tap.

Bulk shear, 0 to 6 km

Quick read

How much the wind changes, in speed and direction together, between the ground and about 6 km (20,000 ft) up. The map shows how big that change is, not which way it points. Read it where storms are likely to form, together with CAPE.

  1. 20
  2. 30
  3. 40
  4. 50
  5. 65
  6. 80
  7. 100

Knots. Clear below 10 kt. What each range usually means is in the table above.

For forecasters

Definition
Bulk wind difference: the length of the vector difference between the wind near 6 km and the wind near the ground. In the HRRR field used here, the post-processor takes the mean wind from 5.5 to 6 km above ground minus the mean wind from 0 to 0.5 km; SPC uses the 6 km wind minus the surface wind.
Units
kt (computed in m/s, × 1.944)
Layer
Fixed 0 to 6 km above ground, not the effective layer.
How it is made
Computed by MesoRadar as the length of the HRRR's own 0 to 6 km shear vector (GRIB fields VUCSH and VVCSH, level "0-6000 m above ground"). Direction is not drawn.
Caveats
A fixed layer is the wrong ruler for elevated or shallow storms; SPC's effective bulk wind difference (in the Skew-T tool as Effective shear) fits the storm's inflow better.
Sources
SPC mesoanalysis help, 0 to 6 km shear; HRRR GRIB2 inventory (wrfsfc file).

Effective shear: matching the layer to the storm

A fixed 0 to 6 km layer is the wrong ruler for a storm that is shallow, or for one feeding on air above a cool surface layer. Effective bulk wind difference fixes that. It starts at the base of the layer the storm can actually ingest, the effective inflow layer, and runs halfway up to the storm’s top. SPC: supercells become more likely as it climbs through 25 to 40 kt and beyond.

The effective inflow layer itself is defined by buoyancy. It is the part of the sounding where a lifted parcel has at least 100 J/kg of CAPE and CIN no worse than 250 J/kg. If that layer starts above the ground, the storm is elevated.

In MesoRadar: the Skew-T tool's More fold, group Effective inflow: Base / top, Effective SRH and Effective shear (Pro).

Storm-relative helicity: the spin a storm can use

The rolling layer in the diagram helps a storm only if the storm takes that air in front-on. Picture throwing a spiraling football: the ball’s spin points along its path. When the low-level spin points along the direction the air flows into the storm, rising air comes in already spiraling, and the updraft rotates. That alignment is called streamwise spin.

Storm-relative helicity (SRH) measures how much of the low-level spin is streamwise for a storm moving in a given direction. It is reported in square meters per second squared (m²/s²) over a layer, usually 0 to 1 km and 0 to 3 km. On a hodograph it is proportional to the area swept between the storm’s motion point and the wind curve.

SPC’s reading of it:

  • There is no clean line between storms that will and will not make tornadoes. Bigger is generally worse.
  • 0 to 3 km SRH above about 250 m²/s² and 0 to 1 km SRH above about 100 m²/s² point to a higher tornado threat with supercells.
  • The lowest kilometer matters most for tornadoes. Spin in that layer is what a supercell can stretch into a tight circulation near the ground. Thompson and colleagues found 0 to 1 km SRH, and its effective-layer version, among the best separators of tornadic from non-tornadic supercells. A later study of 20,194 right-moving supercells found SRH in the lowest 500 m separated significantly tornadic storms from the rest best, and swapping it into STP cut missed events and false alarms by 18% (Coffer and colleagues 2019). The Skew-T tool’s Wind by layer table has a 0-500 m row.

In MesoRadar: Storm-Relative Helicity 0 to 1 km and 0 to 3 km, in HRRR Shear & Rotation (Pro). The Skew-T tool lists SRH 0-1 / 0-3, and Effective SRH under More.

Storm-relative helicity, 0 to 1 km (SRH)

Quick read

How much spin in the lowest kilometer a storm moving the expected way could draw into its updraft. This is the layer that matters most for tornadoes in supercells. It describes the air, not a storm: high values matter only where storms form.

  1. 50
  2. 100
  3. 150
  4. 250
  5. 350
  6. 450
  7. 600

m²/s². Clear at 0 and below. SPC: 0 to 1 km SRH above about 100 m²/s² points to a higher tornado threat with supercells; there is no clean line between storms that do and do not.

For forecasters

Definition
SRH = −∫ k · (V − C) × ∂V/∂z dz from the ground to 1 km, where V is the wind and C the storm motion.
Units
m²/s²
Layer
Fixed 0 to 1 km above ground.
How it is made
Native HRRR output (NCEP Unified Post Processor). MesoRadar reads the value from the HRRR file and only colors it; nothing is recomputed. GRIB field HLCY, level "1000-0 m above ground", computed with the post-processor's Bunkers right-mover storm motion: 7.5 m/s to the right of the 0 to 6 km mean wind, at right angles to the shear from the 0 to 0.5 km mean wind to the 5.5 to 6 km mean wind (Bunkers et al. 2000).
Caveats
The storm motion is an estimate; a storm that turns right or rides a boundary can take in far more helicity than the map shows. Negative values (spin that works against a right mover) are not colored.
Sources
SPC mesoanalysis help, 0 to 1 km SRH; Thompson et al. 2003; Bunkers et al. 2000 (storm motion).

Storm-relative helicity, 0 to 3 km (SRH)

Quick read

The same measure through a deeper layer, the lowest 3 km. It speaks more to whether a storm's updraft will rotate at all (a supercell) than to tornadoes, which depend more on the lowest kilometer.

  1. 50
  2. 100
  3. 150
  4. 250
  5. 350
  6. 450
  7. 600

m²/s², on the same colors as the 0 to 1 km map. Clear at 0 and below. SPC: above about 250 m²/s² points to a higher tornado threat with supercells.

For forecasters

Definition
SRH (as above) integrated from the ground to 3 km.
Units
m²/s²
Layer
Fixed 0 to 3 km above ground.
How it is made
Native HRRR output (NCEP Unified Post Processor). MesoRadar reads the value from the HRRR file and only colors it; nothing is recomputed. GRIB field HLCY, level "3000-0 m above ground", with the same Bunkers right-mover storm motion as the 0 to 1 km map. The Supercell Composite map uses this same field.
Caveats
As for 0 to 1 km. SPC's effective SRH (in the Skew-T tool) uses the layer a storm can actually ingest instead of a fixed depth.
Sources
SPC mesoanalysis help, 0 to 3 km SRH; Bunkers et al. 2000.

Storm motion changes the answer

SRH depends on how the storm is moving, and supercells do not move with the mean wind. As they mature they often split into two storms: a right mover that turns to the right of the mean wind and a left mover that turns to the left. In the usual United States setup, with winds turning clockwise with height, the right mover is the one that inherits the low-level spin and is favored. The left mover can still drop large hail.

Forecasters estimate both with the Bunkers method: start from the average wind through 0 to 6 km, then step about 7.5 m/s (15 kt) to the right or left of the shear direction. It is an estimate. A storm riding a boundary, or turning hard right as it matures, can sweep up far more helicity than the map says, which is one reason chasers watch storms that deviate.

Two words worth knowing:

  • Veering: wind turning clockwise with height, for example from the southeast at the ground to the west aloft. It usually brings warm air in, and it gives a curved hodograph.
  • Backing: wind turning counterclockwise with height, which shrinks helicity for a right mover. Watch the trap in the word: “backed surface winds” means the ground wind blows from a more easterly direction than usual, under the same winds aloft. That adds turning in the lowest kilometer and raises SRH, which is why forecasters like to see it.

In MesoRadar: the Skew-T tool's RM moving to row gives the right mover's estimate in your wind units; the More fold adds LM moving to, Mean wind from and SRH 0-3 (LM), the left mover's helicity (Pro).

Will storms stay separate?

Supercell numbers assume storms stay apart. Whether they do depends a lot on the boundary that starts them. When the deep-layer shear and the mean wind through the cloud layer point along the front or dryline, storms stay in the boundary’s lift, crowd together, and the day often turns into a line within about 3 hours. When they cross the boundary at a large angle, storms move off it and stay separate longer (Dial, Racy and Thompson 2010). Before you trust a high STP, compare the direction of the 0 to 6 km shear, and of the storms’ expected motion, with the boundary on the map. The Bulk Shear map shows only its strength; the hodograph in the Skew-T tool shows its direction.

Updraft helicity: the model's own storms

Everything above describes the environment. Updraft helicity is different: it is computed inside the storms a model simulates. The HRRR is fine enough (about 3 km) to draw individual thunderstorms, and for each one it measures how strongly the updraft rotates between 2 and 5 km up.

How to use it:

  • Each frame is an hourly maximum. The HRRR keeps the strongest value each storm reached in the hour before the frame time, so a streak is a simulated storm’s path over that hour, not where it sits at that moment.
  • Long, narrow streaks mean the model is making long-lived rotating storms, its version of supercells. Scattered dots mean brief or weak ones.
  • It is a storm-mode clue. It says what kind of storms the model expects, roughly where, and roughly when. It does not place tornadoes, and a given run can be off by tens of miles or hours. Compare several hourly runs before trusting one.
  • No streaks with big CAPE and shear often means the model keeps storms from forming, perhaps under a strong cap. That is worth knowing too.

In MesoRadar: Updraft Helicity 2 to 5 km, in HRRR Shear & Rotation (Pro).

Updraft helicity, 2 to 5 km (UH)

Quick read

How strongly the model's own simulated storms rotate between 2 and 5 km up. Long, narrow streaks mean the model is making long-lived rotating storms. It is a clue to storm type, place and timing in that model run, not a detection of anything real.

  1. 50
  2. 100
  3. 150
  4. 200
  5. 250
  6. 300
  7. 400

m²/s². Clear below 25. Research on 2 to 4 km models has flagged simulated supercells with thresholds of roughly 25 to 100 m²/s² and higher; the right threshold depends on the model and its grid spacing (Kain et al. 2008; Sobash et al. 2011).

For forecasters

Definition
UH = ∫ w ζ dz from 2 to 5 km above ground, where w is vertical velocity and ζ vertical vorticity.
Units
m²/s²
Layer
2 to 5 km above ground; the largest value at each grid point during the hour before the frame time. The first frame of a run (the analysis hour) has no hour before it and shows the value at that moment.
How it is made
Native HRRR output (NCEP Unified Post Processor). MesoRadar reads the value from the HRRR file and only colors it; nothing is recomputed. GRIB field MXUPHL, level "5000-2000 m above ground" (hourly maximum).
Caveats
The placement and timing limits are in Updraft helicity: the model's own storms, above.
Sources
Kain et al. 2008, some practical considerations regarding horizontal resolution in the first generation of operational convection-allowing NWP (Weather and Forecasting 23); Sobash et al. 2011, probabilistic forecast guidance for severe thunderstorms based on the identification of extreme phenomena in convection-allowing model forecasts (Weather and Forecasting 26).

A quick read of the shear

  1. Is the deep layer strong enough? About 35 to 40 kt of 0 to 6 km shear, or 25 to 40 kt and up of effective shear, for supercells.
  2. Is it in the right place? Shear and CAPE have to overlap where storms will form.
  3. Will storms stay separate? Compare the shear direction and the storms’ motion with the boundary: along it favors a line, across it favors separate storms (more).
  4. How much low-level spin is there? 0 to 1 km SRH and the 0 to 1 km shear.
  5. What does the model do with it? Updraft helicity streaks, across several runs.
  6. Then combine. The composite indices in SCP & STP, Explained put these together.

Key points

  • Shear is the change in wind with height; it keeps updraft and downdraft apart, so storms last
  • About 35 to 40 kt of 0 to 6 km shear is the usual supercell guide
  • SRH measures the low-level spin a moving storm can draw in; 0 to 1 km matters most for tornadoes
  • SRH depends on storm motion; right movers usually get the most
  • Effective-layer versions fit the storm’s real inflow, including elevated storms
  • Updraft helicity is the model’s simulated supercells, not a tornado forecast

Common questions

How much wind shear do supercells need?

SPC's guide is that supercells are commonly associated with 0 to 6 km shear of 35 to 40 knots and greater. Below about 20 knots, storms tend to be short-lived single cells. In between, storms tend to organize into clusters and lines.

What is storm-relative helicity?

A measure of how much of the spin in the low-level wind a storm can draw into its updraft, given the direction the storm is moving. It is measured in square meters per second squared over a layer, usually the lowest 1 km or 3 km. Larger values mean a rotating updraft is more likely, and in the lowest kilometer, a tornado is more likely if a supercell forms.

What is a good SRH value for tornadoes?

SPC says there is no clean threshold, but 0 to 1 km SRH above about 100 and 0 to 3 km SRH above about 250 m²/s² point to an increased tornado threat with supercells. Higher is generally worse, and storms must form and stay discrete for any of it to matter.

What is updraft helicity?

A number the forecast model computes inside its own simulated storms: how strongly their updrafts rotate between 2 and 5 km up. Long streaks of it trace where the model is making supercells. It is a storm-mode clue, not a map of tornadoes.

Sources