Geostrophic, Gradient and Thermal Wind
CSIR NET Part C meteorology: natural coordinates, geostrophic, inertial, cyclostrophic and gradient winds, thermal wind, thickness, veering and backing.
- 18 min
On this page · 15 sections
Prerequisites
- The scale analysis that justifies balance, from Forces, Equations of Motion, Scale Analysis and Dimensionless Numbers.
- The hypsometric equation and thickness, from Composition, Thermal Structure, Hydrostatics and the Standard Atmosphere.
- The jet streams and pressure belts described in Temperature, Pressure and Wind Belts.
The Hook: A Jet Stream from a Temperature Map
Natural Coordinates
For balanced flow it is convenient to use natural coordinates \( (s, n) \): \( \mathbf t \) is the unit vector along the flow, \( \mathbf n \) points to the left of the flow, and \( R \) is the radius of curvature of the trajectory, positive when the flow curves to the left (counter-clockwise, cyclonic in the NH).
The horizontal velocity is \( \mathbf V = V\mathbf t \) with \( V \ge 0 \), and the acceleration is
$$\frac{D\mathbf V}{Dt} = \frac{DV}{Dt}\,\mathbf t + \frac{V^2}{R}\,\mathbf n$$The Coriolis force \( -f\mathbf k\times\mathbf V = -fV\mathbf n \) acts entirely in the normal direction, and the pressure gradient force has components \( -\partial\Phi/\partial s \) and \( -\partial\Phi/\partial n \). So:
$$\frac{DV}{Dt} = -\frac{\partial\Phi}{\partial s},\qquad \frac{V^2}{R} + fV = -\frac{\partial\Phi}{\partial n}$$For flow parallel to the height contours, \( \partial\Phi/\partial s = 0 \) and the speed is constant. The normal equation is a three-way balance among centripetal acceleration, Coriolis and pressure gradient. Dropping one term at a time gives three special cases; keeping all three gives the gradient wind.
| Balance | Terms kept | Valid when |
|---|---|---|
| Geostrophic | Coriolis = PGF | \( Ro \ll 1 \), straight flow |
| Inertial | centripetal = Coriolis | no pressure gradient |
| Cyclostrophic | centripetal = PGF | \( Ro \gg 1 \) (small, intense vortices) |
| Gradient | all three | curved flow, any \( Ro \) |
Geostrophic Wind
$$\boxed{\mathbf V_g = \frac{1}{f}\,\mathbf k\times\nabla_p\Phi = \frac{g}{f}\,\mathbf k\times\nabla_p Z = \frac{1}{\rho f}\,\mathbf k\times\nabla_z p}$$In components: \( u_g = -\dfrac{1}{f}\dfrac{\partial\Phi}{\partial y} \), \( v_g = \dfrac{1}{f}\dfrac{\partial\Phi}{\partial x} \).
Properties:
- It blows parallel to isobars (height contours) with low pressure on the left in the NH (Buys Ballot’s law: stand with your back to the wind and low pressure is on your left).
- Its speed is proportional to the gradient: closely packed contours, strong wind.
- On a constant-pressure chart, a given height gradient gives the same \( V_g \) at all levels at a given latitude, because density does not appear. On a constant-height chart, the same pressure gradient gives a stronger wind aloft where \( \rho \) is smaller.
- For constant \( f \), the geostrophic wind is non-divergent: \( \nabla\cdot\mathbf V_g = 0 \). With \( f \) varying, \( \nabla\cdot\mathbf V_g = -(\beta/f)v_g \), small. The divergence that drives vertical motion comes from the ageostrophic wind \( \mathbf V_a = \mathbf V - \mathbf V_g \).
- Undefined at the equator.
Ageostrophic wind and the isallobaric component
The momentum equation can be rewritten as
$$\mathbf V_a = \frac{1}{f}\,\mathbf k\times\frac{D\mathbf V}{Dt}$$so the ageostrophic wind is perpendicular to (and to the left of, in the NH) the acceleration. Two pieces are commonly identified:
- The isallobaric wind, directed towards falling pressure (down the gradient of pressure tendency), proportional to \( -\nabla(\partial\Phi/\partial t) \). It fills developing lows and brings low-level convergence to regions of rapid pressure fall.
- The inertial-advective part associated with entering and leaving jet streaks: ageostrophic flow towards low pressure where air accelerates into a jet entrance, and towards high pressure in the exit. The resulting divergence pattern (upper divergence in the right-entrance and left-exit quadrants in the NH) is a standard forecasting rule.
Near the ground, friction adds a cross-isobar component towards low pressure; see Boundary Layer, Turbulence and the Ekman Layer.
Inertial Flow
With no pressure gradient, \( V^2/R + fV = 0 \), so
$$R = -\frac{V}{f}$$In the NH (\( f > 0 \)), \( R < 0 \): the parcel turns clockwise (anticyclonically) in a circle of radius \( V/f \), completing it in \( 2\pi/f \), the inertial period. Inertial oscillations are rarely clean in the atmosphere, but they appear in the ocean mixed layer after a storm passes, and in the nocturnal low-level jet (the boundary-layer wind, freed from friction after sunset, oscillates about the geostrophic wind and becomes supergeostrophic during the night).
Cyclostrophic Flow
When \( Ro \gg 1 \) the Coriolis term is negligible:
$$\boxed{\frac{V^2}{R} = -\frac{\partial\Phi}{\partial n}\ \Rightarrow\ V = \left(-R\frac{\partial\Phi}{\partial n}\right)^{1/2} = \left(\frac{R}{\rho}\left|\frac{\partial p}{\partial r}\right|\right)^{1/2}}$$The pressure gradient always points inward (towards low pressure at the centre); there is no cyclostrophic high. Both rotation senses are dynamically allowed. Tornadoes are mostly cyclonic because they inherit rotation from the parent storm, which inherits it from larger scales; dust devils occur with either sense in roughly equal numbers.
Gradient Wind
Keeping all three terms, \( V^2/R + fV + \partial\Phi/\partial n = 0 \) is a quadratic in \( V \):
$$\boxed{V = -\frac{fR}{2} \pm\left(\frac{f^2R^2}{4} - R\frac{\partial\Phi}{\partial n}\right)^{1/2}}$$Only roots that are real and non-negative are physical. Writing \( V_g = -(1/f)\,\partial\Phi/\partial n \), the balance becomes
$$\frac{V_g}{V} = 1 + \frac{V}{fR}$$Regular low and regular high (NH)
Using \( |R| \) for the radius:
| Case | Formula | Comparison with \( V_g \) |
|---|---|---|
| Regular low (cyclonic) | ( V = -\dfrac{f | R |
| Regular high (anticyclonic) | ( V = \dfrac{f | R |
Why: around a low, centripetal acceleration points inward and is supplied partly by the PGF; Coriolis need only supply the rest, so \( V \) is less than geostrophic. Around a high, the inward centripetal acceleration must be provided by Coriolis against an outward PGF, so Coriolis (and \( V \)) must exceed the geostrophic value.
The limit on anticyclones
For the high, the square root is real only if
$$\left|\frac{\partial\Phi}{\partial n}\right| \le \frac{f^2|R|}{4}\qquad\Leftrightarrow\qquad V_g \le \frac{f|R|}{4},\qquad V \le \frac{f|R|}{2}$$The pressure gradient must vanish as \( R\to 0 \): highs have weak gradients and light winds near their centres, while lows can have arbitrarily steep gradients. This asymmetry is visible on every weather chart.
All four roots
Mathematically there are also anomalous lows (clockwise flow round a low, needing a strong gradient, rare and small-scale) and anomalous highs (supergeostrophic flow with the “+” root, not observed on synoptic scales). Part C sometimes asks which combinations are physically allowed:
| Pressure centre | Rotation (NH) | Name | Observed? |
|---|---|---|---|
| Low | anticlockwise | regular low | very common |
| High | clockwise | regular high | very common |
| Low | clockwise | anomalous low | rare, small (anticyclonic tornadoes and dust devils) |
| High | clockwise, “+” root | anomalous high | not observed on synoptic scale |
| High | anticlockwise | none | impossible: no balance exists |
Thermal Wind
Derivation
Differentiate the geostrophic wind in isobaric coordinates with respect to \( \ln p \) and use the hydrostatic equation \( \partial\Phi/\partial\ln p = -RT \):
$$\frac{\partial\mathbf V_g}{\partial\ln p} = \frac{1}{f}\mathbf k\times\nabla_p\frac{\partial\Phi}{\partial\ln p} = -\frac{R}{f}\,\mathbf k\times\nabla_p T$$Integrating from \( p_0 \) (lower) to \( p_1 \) (upper):
$$\boxed{\mathbf V_T = \mathbf V_g(p_1) - \mathbf V_g(p_0) = \frac{R}{f}\ln\frac{p_0}{p_1}\,\mathbf k\times\nabla_p\bar T = \frac{g}{f}\,\mathbf k\times\nabla(Z_1 - Z_0)}$$where \( \bar T \) is the mean temperature of the layer. The thermal wind is not a wind but a vertical shear vector of the geostrophic wind.
Reading the thermal wind
- \( \mathbf V_T \) blows parallel to the mean isotherms (thickness lines) with cold air on the left in the NH, just as the geostrophic wind has low pressure on the left.
- Temperature decreasing towards the pole gives \( \mathbf V_T \) from the west: westerlies increase with height. This is why the mid-latitude troposphere has westerlies capped by jet streams near the tropopause, where the meridional temperature gradient reverses.
- Thickness: \( Z_1 - Z_0 = (R\bar T/g)\ln(p_0/p_1) \). For 1000–500 hPa, about 29.3 × \( \bar T \) × 0.693 m; at \( \bar T = 260 \) K this is 5,272 m. Thickness lines are mean isotherms.
Temperature advection: veering and backing
Let the geostrophic wind turn with height. Because the change of wind with height is the thermal wind, and the thermal wind keeps cold air on its left:
| Wind change with height (NH) | Meaning | Advection |
|---|---|---|
| Veering (turns clockwise with height, e.g. southerly to westerly) | low-level wind blows from warm to cold across isotherms | warm advection |
| Backing (turns anticlockwise with height, e.g. westerly to southerly) | wind blows from cold to warm | cold advection |
| No turning, only speed change | wind parallel to isotherms | no advection |
Warm advection is associated with ascent (ahead of warm fronts, in the warm sector) and cold advection with descent. In the Southern Hemisphere the rule reverses. A quantitative check: the geostrophic temperature advection in a layer is \( -\mathbf V_g\cdot\nabla T \), computable once \( \nabla T \) is found from \( \mathbf V_T \).
Barotropic and baroclinic atmospheres
| Barotropic | Baroclinic | |
|---|---|---|
| Definition | density depends only on pressure, \( \rho = \rho(p) \): isobaric and isosteric (and isothermal) surfaces coincide | density depends on pressure and temperature; isobars and isotherms intersect |
| Temperature on a pressure surface | uniform | varies |
| Thermal wind | zero: geostrophic wind independent of height | non-zero: wind changes with height |
| Solenoids | none | present (see the circulation page) |
| Example | deep tropics (approximately), idealised models | mid-latitude frontal zones |
| Instability available | barotropic (from horizontal shear) | baroclinic (from vertical shear and APE) |
graph LR
A[Hydrostatic: dPhi/dlnp = -RT] --> C[Thermal wind: shear parallel to isotherms, cold on left]
B[Geostrophic: Vg = k x grad Phi / f] --> C
C --> D[Poleward temperature drop gives westerlies increasing with height]
C --> E[Veering = warm advection, backing = cold advection]
D --> F[Jet streams near the tropopause]Memory Tricks
Don’t Confuse
| Pair | Distinction |
|---|---|
| Geostrophic vs gradient wind | straight-line PGF = Coriolis vs curved flow including centripetal acceleration |
| Subgeostrophic vs supergeostrophic | around lows vs around highs (regular cases) |
| Inertial vs cyclostrophic flow | no PGF, always anticyclonic vs no Coriolis, either sense, always a low |
| Geostrophic vs thermal wind | a wind at one level vs the vertical shear between two levels |
| Veering vs backing | clockwise turn with height (warm advection, NH) vs anticlockwise (cold advection) |
| Barotropic vs equivalent barotropic | no temperature gradient on isobaric surfaces vs isotherms parallel to contours |
| Ageostrophic vs isallobaric wind | total departure from \( V_g \) vs the component driven by pressure tendency |
Common Mistakes
Practice
All questions below are original practice questions written for this page. They are not from any CSIR NET paper.
Level 1: Foundation
- Which three forces balance in the gradient wind?
- State the condition on \( V_g \) for gradient balance around a high of radius \( R \).
- In the NH, does a wind that backs with height indicate warm or cold advection?
Answers
- Pressure gradient force, Coriolis force and centripetal acceleration (the centrifugal term in the rotating natural frame).
- \( V_g \le f|R|/4 \), equivalently \( |\partial\Phi/\partial n| \le f^2|R|/4 \).
- Cold advection.
Level 2: Part B Style
Practice Q1. At 500 hPa near 45° N, the height decreases by 80 m over 400 km. The geostrophic wind is closest to:
(a) 9.5 m s\(^{-1}\) (b) 19 m s\(^{-1}\) (c) 27 m s\(^{-1}\) (d) 38 m s\(^{-1}\)
Answer: (b) 19 m/s
\( V_g = (9.81/1.031\times 10^{-4})\times(80/4\times 10^5) = 95{,}150\times 2\times 10^{-4} = 19.0 \) m s\(^{-1}\).
Practice Q2. Around a low of radius 300 km with \( f = 10^{-4} \) s\(^{-1}\), the geostrophic wind is 15 m s\(^{-1}\). The gradient wind speed is closest to:
(a) 7.5 m s\(^{-1}\) (b) 11 m s\(^{-1}\) (c) 15 m s\(^{-1}\) (d) 22 m s\(^{-1}\)
Answer: (b) 11 m/s
\( f|R| = 30 \) m s\(^{-1}\). \( V = -15 + \sqrt{225 + 30\times 15} = -15 + \sqrt{675} = -15 + 25.98 = 11.0 \) m s\(^{-1}\). Subgeostrophic, as expected for a low.
Practice Q3. For the same \( f \) and radius (300 km), the largest geostrophic wind that can be in gradient balance around a high is:
(a) 3.75 m s\(^{-1}\) (b) 7.5 m s\(^{-1}\) (c) 15 m s\(^{-1}\) (d) unlimited
Answer: (b) 7.5 m/s
\( V_g \le f|R|/4 = 30/4 = 7.5 \) m s\(^{-1}\). For example, \( V_g = 5 \) gives \( V = 15 - \sqrt{225 - 150} = 6.34 \) m s\(^{-1}\), supergeostrophic.
Practice Q4. In the NH, the geostrophic wind is from the south at 850 hPa and from the west at 500 hPa. Between these levels there is:
(a) cold advection; the wind backs (b) warm advection; the wind veers (c) warm advection; the wind backs (d) no advection
Answer: (b)
South (180°) to west (270°) is a clockwise turn with height: veering, which means the thermal wind has a component that makes the low-level wind cross isotherms from warm to cold, i.e. warm advection.
Level 3: Part C Style
Practice Q5. At 45° N (take \( f = 10^{-4} \) s\(^{-1}\), \( R = 287 \) J kg\(^{-1}\) K\(^{-1}\)), the geostrophic wind components (\( u, v \)) in m s\(^{-1}\) are (5, 5) at 1000 hPa and (20, 5) at 500 hPa. Which statements are correct?
- The thermal wind for the layer is 15 m s\(^{-1}\) from the west, so the mean temperature decreases northward.
- The mean-layer meridional temperature gradient is about 7.5 K per 1,000 km.
- The wind veers with height, from about 225° at 1000 hPa to about 256° at 500 hPa.
- Using the layer-mean geostrophic wind, the geostrophic temperature advection is warm advection of about 3 K per day.
(a) 1 and 2 only (b) 1, 2 and 3 only (c) 1, 2, 3 and 4 (d) 3 and 4 only
Answer: (c) all four
- 1 true: \( \mathbf V_T = (20-5,\ 5-5) = (15, 0) \), a westerly shear; with cold air on the left (north), temperature decreases northward.
- 2 true: \( -\partial\bar T/\partial y = fV_T/(R\ln 2) = 10^{-4}\times 15/(287\times 0.693) = 7.54\times 10^{-6} \) K m\(^{-1}\) = 7.5 K per 1,000 km.
- 3 true: (5, 5) blows from the south-west, 225°; (20, 5) from \( 270° - \arctan(5/20) \approx 256° \). Clockwise turning: veering.
- 4 true: mean wind (12.5, 5). Advection \( = -v\,\partial\bar T/\partial y = 5\times 7.54\times 10^{-6} = 3.77\times 10^{-5} \) K s\(^{-1}\) = 3.3 K per day, warming. The southerly component carries warm air northward, consistent with veering.
Practice Q6. Which statements about balanced flow are correct?
- In a barotropic atmosphere, the geostrophic wind does not change with height.
- A strong anticyclone with steep pressure gradients near its centre is common on synoptic charts.
- For constant \( f \), the geostrophic wind has zero horizontal divergence, so vertical motion must be linked to the ageostrophic wind.
- In cyclostrophic balance, the flow may rotate in either sense, but pressure at the centre must be low.
(a) 1, 3 and 4 only (b) 1 and 2 only (c) 2 and 3 only (d) 1, 2, 3 and 4
Answer: (a) 1, 3 and 4 only
- 1 true: no temperature gradient on isobaric surfaces, so \( \mathbf V_T = 0 \).
- 2 false: the gradient-wind limit \( |\partial\Phi/\partial n| \le f^2|R|/4 \) forces weak gradients near anticyclone centres.
- 3 true: \( \nabla\cdot\mathbf V_g = 0 \) on an f-plane.
- 4 true: the PGF must provide the inward centripetal force, so the centre is low; Coriolis is absent so rotation sense is unconstrained.
Practice Q7. On a winter 200 hPa chart, a 70 m s\(^{-1}\) westerly jet core lies at 30° N over north India. Surface winds below are weak. Which statements are consistent with thermal-wind balance?
- The troposphere below the jet must have a strong poleward decrease of mean temperature.
- Above the jet core, in the lower stratosphere, the meridional temperature gradient at those latitudes has reversed or weakened, so wind decreases with height.
- The jet core must coincide with the region of maximum surface pressure gradient.
- The 1000–200 hPa thickness gradient across the jet is large.
(a) 1, 2 and 4 only (b) 1 and 3 only (c) 2 and 3 only (d) 1, 2, 3 and 4
Answer: (a) 1, 2 and 4 only
- 1 true: a large westerly shear from weak surface winds to 70 m s\(^{-1}\) requires a strong poleward temperature decrease.
- 2 true: wind maximum means shear changes sign, so the temperature gradient reverses above the core.
- 3 false: thermal wind relates the shear to the layer temperature gradient; the surface pressure gradient can be weak, as stated.
- 4 true: thickness gradients are the mean temperature gradients that produce the shear.
Quick Revision Table
| Idea | Fact or formula |
|---|---|
| Natural coordinates | \( V^2/R + fV = -\partial\Phi/\partial n \); \( \mathbf n \) to the left; \( R > 0 \) cyclonic |
| Geostrophic | \( \mathbf V_g = (1/f)\mathbf k\times\nabla_p\Phi = (g/f)\mathbf k\times\nabla Z \); low on left (NH) |
| Inertial | \( R = -V/f \), clockwise (NH), period \( 2\pi/f \) |
| Cyclostrophic | \( V = \sqrt{(r/\rho)\,\partial p/\partial r} \); low centre only |
| Gradient | \( V = -fR/2 \pm\sqrt{f^2R^2/4 - R\,\partial\Phi/\partial n} \); \( V_g/V = 1 + V/fR \) |
| Lows / highs | subgeostrophic / supergeostrophic; high limit ( V_g \le f |
| Thermal wind | \( \mathbf V_T = (R/f)\ln(p_0/p_1)\,\mathbf k\times\nabla\bar T \); cold on left (NH) |
| Thickness | \( \Delta Z = (R\bar T/g)\ln(p_0/p_1) \); 1000–500 hPa \( \approx 20.3\bar T \) m |
| Advection | veering = warm, backing = cold (NH) |
| Barotropic | \( \rho = \rho(p) \), no thermal wind; baroclinic otherwise |
Key Takeaways
What’s Next
- Circulation, Vorticity and Potential Vorticity: the rotation and spin of these balanced flows.
- Gravity, Inertia and Rossby Waves; Barotropic and Baroclinic Instability: how the baroclinic atmosphere generates cyclones.
- Air Masses, Monsoon, Jet Streams, Cyclones and ENSO: the climatology of the jets.