Matsuno–Gill Tropical Heating Model

The Matsuno–Gill tropical heating model is the canonical analytical framework in geophysical fluid dynamics and tropical meteorology that explains how localized diabatic heating in the deep tropics excites planetary-scale horizontal wind and pressure patterns across the equatorial troposphere. Synthesizing Taroh Matsuno's (1966) normal-mode theory of equatorially trapped shallow-water waves with Adrian E. Gill's (1980) closed-form steady solutions for damped, heat-induced tropical circulations, the model demonstrates that a spatially localized convective heat source on the equatorial beta-plane generates an asymmetric zonal and meridional response governed by wave dispersion rather than local hydrostatic buoyancy alone.

When deep cumulonimbus convection releases latent heat over warm tropical ocean basins such as the Indo-Pacific Warm Pool, mid-tropospheric warming drives upward vertical motion, lower-tropospheric mass convergence, and upper-tropospheric mass divergence. In the Matsuno–Gill framework, this localized mass sink projects onto two equatorially trapped wave modes of opposite group velocity: (1) a fast, non-dispersive equatorial Kelvin wave that propagates eastward, establishing a broad zonal trough and low-level easterly trade winds (the lower branch of the Walker circulation) across thousands of kilometers to the east of the heating region; and (2) a slower equatorial Rossby wave doublet (n = 1 mode) that propagates westward at one-third the Kelvin wave speed, generating twin off-equatorial cyclonic gyres straddling the equator and strong low-level equatorial westerlies to the west of the heat source. Together with antisymmetric extensions for off-equatorial monsoon heating, the Matsuno–Gill solution forms the dynamical backbone for understanding the El Niño–Southern Oscillation (ENSO), the Madden–Julian Oscillation (MJO), tropical teleconnections, and exoplanetary tide-locked atmospheric superrotation.

Matsuno's (1966) Equatorial Beta-Plane Wave Dispersion

Prior to the mid-1960s, theoretical meteorology focused on mid-latitude quasi-geostrophic dynamics, where the Coriolis effect parameter f = 2Ω sin φ is large (f₀ ≈ 10⁻⁴ s⁻¹) and horizontal winds remain in geostrophic balance with pressure gradients. Near the equator (φ → 0), the Coriolis parameter vanishes and changes sign between hemispheres, turning the equator into a waveguide that traps low-frequency disturbances.

In his 1966 paper "Quasi-geostrophic motions in the equatorial area" (Journal of the Meteorological Society of Japan), Taroh Matsuno introduced the equatorial beta-plane approximation (f = βy, where β = ∂f/∂y = 2Ω / a ≈ 2.29 × 10⁻¹¹ m⁻¹ s⁻¹ and y is northward distance from the equator) into the linearized shallow-water equations. Solving the meridional structure equation in parabolic cylinder (Hermite) functions, Matsuno derived the dispersion relation for equatorially trapped waves—revealing equatorial Rossby waves, inertia-gravity waves, the mixed Rossby–gravity (Yanai) wave (n = 0), and the unidirectional eastward-propagating equatorial Kelvin wave (n = −1).

Gill's (1980) Steady Heat-Induced Circulation Theory

Observational syntheses by Jacob Bjerknes (1969) and T. N. Krishnamurti (1971) highlighted a pronounced zonal asymmetry in the tropical atmosphere: deep convective heating over Indonesia and the western Pacific is flanked to the east by subsidence and easterly trade winds spanning over 10,000 km across the Pacific, whereas to the west of the convective center, the response is confined to a narrower region featuring twin low-level cyclonic vortices and equatorial westerlies over the eastern Indian Ocean.

In 1980, Adrian E. Gill published "Some simple solutions for heat-induced tropical circulation" in the Quarterly Journal of the Royal Meteorological Society. Building on linear studies by Peter J. Webster (1972), Gill posed a boundary-value problem for a localized, stationary diabatic heating anomaly Q(x, y) in a resting equatorial atmosphere subject to Rayleigh friction and Newtonian cooling of equal rate ε. Expanding the velocity and geopotential fields in Matsuno's parabolic cylinder basis functions, Gill obtained closed-form analytical solutions showing that the 3:1 ratio between the eastward Kelvin wave speed (c) and the westward n = 1 equatorial Rossby wave speed (c/3) explains the observed east–west asymmetry of the Walker circulation and monsoon troughs.

First-Baroclinic Mode Vertical Decomposition

The tropical troposphere between the sea surface (z = 0) and the tropopause (z = H_T ≈ 15–16 km) is stratified with buoyancy frequency N ≈ 1.0 × 10⁻² s⁻¹. Deep moist convection deposits latent heat throughout the middle troposphere with a vertical profile sin(πz / H_T) peaking near 400–500 hPa. Projecting the primitive equations onto this first baroclinic vertical mode (m = π / H_T) separates vertical and horizontal structure:

• Horizontal velocity (u′, v′) and geopotential (Φ′) perturbations vary as cos(πz / H_T), so lower-tropospheric winds and pressures have the opposite sign to upper-tropospheric (200 hPa) winds and pressures.
• Vertical velocity (w′) and potential temperature (θ′) perturbations vary as sin(πz / H_T), peaking in the mid-troposphere.
• Horizontal motion obeys the shallow-water equations with equivalent depth h_e = N²H_T² / (gπ²) ≈ 200–250 m (or 25–50 m for moist-coupled modes), yielding gravity-wave phase speed c = √(g h_e) ≈ 45–50 m s⁻¹ (or 15–25 m s⁻¹ for moist modes).

Nondimensional Equatorial Beta-Plane Equations

Gill (1980) scaled horizontal lengths by the equatorial Rossby radius of deformation L_e = √( c / (2β) ) (≈ 1,000 km for c = 45 m s⁻¹) and time by T_e = 1 / √(2βc) (≈ 6.2 hours). In nondimensional units—where (x, y) are eastward and northward coordinates, (u, v) are lower-tropospheric wind components, p is lower-tropospheric pressure perturbation, w is mid-tropospheric upward vertical velocity, Q is the diabatic heating rate (acting as a lower-layer mass sink −Q), and ε is the Rayleigh friction and Newtonian cooling rate—the steady linearized equations are:

1. Zonal Momentum Balance: εu − ½ y v = −∂p/∂x
2. Meridional Momentum Balance: εv + ½ y u = −∂p/∂y
3. Thermodynamic / Continuity Balance: εp + ∂u/∂x + ∂v/∂y = −Q(x, y)

Under the long-wave approximation (L_y ≪ L_x), εv is neglected in the meridional momentum equation so zonal winds are in geostrophic balance across the equator (½ y u = −∂p/∂y), while mid-tropospheric vertical velocity is w = εp + Q.

Characteristic Variables and Parabolic Cylinder Basis

Introducing the Riemann characteristic variables q = p + u and r = p − u (so p = ½(q + r) and u = ½(q − r)) transforms the system into:

• εq + ∂q/∂x + ∂v/∂y − ½ y v = −Q,  εr − ∂r/∂x + ∂v/∂y + ½ y v = −Q,  ∂q/∂y + ½ y q + ∂r/∂y − ½ y r = 0

The meridional operators Λ₊ = ∂/∂y + ½ y and Λ₋ = ∂/∂y − ½ y act as ladder operators on the parabolic cylinder functions D_n(y) = e^−y²/4 He_n(y) (D₀ = e^−y²/4, D₁ = y e^−y²/4, D₂ = (y² − 1)e^−y²/4, D₃ = (y³ − 3y)e^−y²/4), satisfying dD_n/dy + ½ y D_n = n D_n−1 and dD_n/dy − ½ y D_n = −D_n+1. Expanding q, r, v, and Q in D_n(y) decouples the system into ordinary differential equations in x for each mode n:

1. Equatorial Kelvin Mode (n = −1, q₀): With v = 0, q₀(x) satisfies dq₀/dx + εq₀ = −Q₀(x), propagating eastward at speed c with decay length L_K = c / ε (or c / (2ε) when friction and cooling are tallied separately).
2. Mixed Rossby–Gravity Mode (n = 0, q₁ and v₀): Under the long-wave approximation, r₀ = 0 and q₁ = 0 outside the forcing region, while v₀(x) = −Q₁(x) responds locally to antisymmetric heating Q₁(x)D₁(y).
3. Equatorial Rossby Modes (n ≥ 1, q_n+1 and r_n−1): The long-wave balance relates r_n−1 = (n + 1)q_n+1 and yields dq_n+1/dx − (2n + 1)εq_n+1 = (n Q_n−1 + Q_n+1)/(2n + 1). Each Rossby mode n ≥ 1 propagates westward at speed c / (2n + 1)—c / 3 for n = 1 and c / 5 for n = 2—with decay scale L_Rn = c / [(2n + 1)ε].

Analytical Solution for Equatorially Centered Heating

To model a localized deep-convective heat source centered on the equator—such as the Maritime Continent warm pool or a central-Pacific El Niño warming anomaly—Gill prescribed a separable heating distribution restricted to |x| < L with a Gaussian meridional profile D₀(y) = e^−y²/4: Q(x, y) = F(x) D₀(y), where F(x) = cos(k x) for |x| < L (k = π / (2L)) and F(x) = 0 for |x| > L. Because the heating projects onto D₀(y), the wave hierarchy reduces to two active modes: the eastward-propagating Kelvin mode (q₀) and the westward-propagating n = 1 symmetric Rossby mode (q₂ and r₀ = 2q₂).

East of the Heat Source (x > L): The Kelvin-Wave Walker Limb

To the east of the convective patch (x > L), the westward-propagating Rossby wave vanishes (q₂ = 0), leaving a damped equatorial Kelvin wave: q₀(x) = q₀(L) e^{−ε(x − L)} < 0, with u = p = ½ q₀(x) e^−y²/4 < 0, v = 0, and w = εp < 0. This solution reproduces the lower-tropospheric branch of the Walker circulation: air subsides across an eastern sector of scale L_K = c / ε, flows westward along the equator as surface easterly trade winds into the eastern flank of the heat source, and ascends within the convective zone, while at 200 hPa in the upper troposphere the signs reverse to produce equatorial westerlies returning mass eastward.

West of the Heat Source (x < −L): The n = 1 Rossby Doublet and Westerly Inflow

To the west of the convective patch (x < −L), the Kelvin wave amplitude is zero (q₀ = 0), and the circulation is governed by the damped n = 1 equatorial Rossby wave (q₂(x) = q₂(−L) e^{+3ε(x + L)} < 0, with r₀ = 2q₂):

• u(x, y) = ½ q₂(x) (y² − 3) e^−y²/4,  p(x, y) = ½ q₂(x) (y² + 1) e^−y²/4,  v(x, y) = 3ε q₂(x) y e^−y²/4

This profile exhibits three diagnostic features: (1) equatorial westerlies (y = 0), where (y² − 3) = −3 multiplies q₂ < 0 to yield low-level westerly winds (u > 0) blowing into the western flank of the heat source, flanked by off-equatorial easterlies poleward of |y| = √3 L_e; (2) twin off-equatorial cyclonic gyres, where (y² + 1)e^−y²/4 attains two off-equatorial pressure minima at |y| = √2 L_e (≈ 12°–14° N and S), forming symmetric twin low-level cyclones (mirrored at 200 hPa by twin upper-level anticyclones); and (3) 3:1 zonal asymmetry, because the Rossby mode decays westward three times faster (e^+3εx) than the Kelvin mode decays eastward (e^−εx), concentrating stronger inflow per unit area into the western equatorial westerlies.

Off-Equatorial Heating and the n = 0 / n = 2 Response

During boreal and austral summer, maximum convective latent heating shifts off the equator into the summer hemisphere. Gill (1980) modeled this configuration with an antisymmetric heating component Q_asym(x, y) = F(x) y e^−y²/4. Because Q₁(x) is odd in y, it excites zero Kelvin wave response (q₀ ≡ 0) and zero disturbance east of the heating zone (x > L). Instead:

1. Within the Heating Sector (|x| < L): The n = 0 mixed Rossby–gravity mode drives cross-equatorial meridional flow (v₀ > 0) from the winter hemisphere into the summer hemisphere, turning eastward under the Coriolis force—capturing the Somali Jet and Findlater Current.
2. West of the Heating Zone (x < −L): The n = 2 antisymmetric Rossby mode (q₃(x) e^+5ε(x+L), decay scale L_K/5) generates a monsoon trough / cyclonic vortex in the summer hemisphere (centered near y ≈ +2 L_e, paired with the 200 hPa Tibetan Anticyclone), an anticyclonic ridge in the winter hemisphere, and low-level monsoon westerlies between the equator and 15° N.

The Damping Timescale Problem and the Lindzen–Nigam (1987) Critique

Matching the horizontal compactness of tropical surface wind anomalies in the Matsuno–Gill model requires a Rayleigh friction and Newtonian cooling rate of ε ≈ (1 to 3 days)⁻¹, whereas radiative cooling timescales in the free troposphere are 10 to 30 days. In 1987, Richard S. Lindzen and Sumant Nigam proposed a boundary-layer formulation (Journal of the Atmospheric Sciences, 1987), arguing that horizontal sea-surface temperature (SST) gradients imprint virtual temperature gradients onto the marine boundary layer (below 700–850 hPa), driving low-level winds and convergence against surface drag (where a 1- to 2-day damping timescale is realistic). J. David Neelin (1989) demonstrated that the Lindzen–Nigam boundary-layer equations and the Matsuno–Gill first-baroclinic mode are mathematically isomorphic, and that tropical surface winds combine boundary-layer pressure gradients with deep free-tropospheric wave adjustment.

ENSO, the Madden–Julian Oscillation, and Exoplanetary Atmospheres

The Matsuno–Gill framework underpins climate dynamics and planetary atmospheric modeling:

• Coupled ENSO Dynamics (Zebiak–Cane Model): In the 1987 Zebiak–Cane coupled model of El Niño, the atmospheric component is a steady Matsuno–Gill model driven by SST and convergence heating anomalies, supplying the surface zonal wind stress that forces oceanic Kelvin and Rossby waves across the Pacific thermocline.
• Madden–Julian Oscillation: Theories of the eastward-propagating 30- to 60-day MJO model its planetary circulation envelope as a coupled Matsuno–Gill Kelvin–Rossby wave packet slowed to ~5 m s⁻¹ by moisture-radiative feedbacks and moisture advection.
• Tidally Locked Exoplanets: Adam P. Showman and Lorenzo M. Polvani (2011) showed that permanent dayside irradiation on tidally locked exoplanets excites a standing Matsuno–Gill pattern whose tilted Kelvin and Rossby lobes pump eastward eddy angular momentum into the equator, driving equatorial superrotation (1–4 km s⁻¹ eastward jets) on hot Jupiters such as HD 189733b.

Why does localized equatorial heating produce easterly winds to the east and westerly winds to the west, rather than symmetric inflow?

Pressure drops on both sides of the heat source, propagating eastward via an equatorial Kelvin wave at speed c and westward via an n = 1 equatorial Rossby wave at speed c/3. Under equal damping rate ε, the Kelvin wave travels three times farther eastward (L_K = c/ε) than the Rossby wave travels westward (L_R = c/(3ε)), stretching the eastern inflow into a broad easterly trade-wind belt while compressing the western inflow into a shorter, stronger equatorial westerly jet flanked by twin off-equatorial cyclones.

Why are upper-tropospheric (200 hPa) winds in the Matsuno–Gill model opposite in direction to surface winds?

Deep tropical convection deposits latent heat in the mid-troposphere (400–500 hPa), exciting the first baroclinic vertical mode whose horizontal velocity and pressure perturbations follow cos(πz / H_T) (+1 at the surface and −1 at the tropopause), so lower-tropospheric convergence is mirrored by upper-tropospheric divergence.

What generates the twin tropical cyclones that often form simultaneously north and south of the equator in the western Pacific and Indian Oceans?

In the symmetric Matsuno–Gill solution, equatorial heating projects onto the n = 1 Rossby mode west of the heat source, creating two off-equatorial pressure minima and cyclonic vorticity maxima near ±√2 to ±√3 equatorial deformation radii (~10°–15° N and S) with equatorial westerlies between them—seeding twin tropical cyclones during active MJO and El Niño onset events.

How does the Matsuno–Gill model explain equatorial superrotation on tidally locked exoplanets?

On a tidally locked exoplanet with permanent dayside heating and nightside cooling, the standing Matsuno–Gill response superposes an eastward-displaced Kelvin wave with westward-displaced off-equatorial Rossby gyres, tilting the horizontal velocity axes into a chevron ("<") pattern that pumps eastward eddy momentum (−∂(u′v′)/∂y > 0) into the equatorial jet.

1. Matsuno, T. (1966). Quasi-geostrophic motions in the equatorial area. Journal of the Meteorological Society of Japan, Series II, 44(1), 25–43. https://doi.org/10.2151/jmsj1965.44.1_25
2. Gill, A. E. (1980). Some simple solutions for heat-induced tropical circulation. Quarterly Journal of the Royal Meteorological Society, 106(449), 447–462. doi:10.1002/qj.49710644905
3. Webster, P. J. (1972). Response of the tropical atmosphere to local, steady forcing. Monthly Weather Review, 100(7), 518–541. https://doi.org/10.1175/1520-0493(1972)100<0518:ROTTAT>2.3.CO;2100%3C0518:ROTTAT%3E2.3.CO;2)
4. Lindzen, R. S., & Nigam, S. (1987). On the role of sea surface temperature gradients in forcing low-level winds and convergence in the tropics. Journal of the Atmospheric Sciences, 44(17), 2418–2436. https://doi.org/10.1175/1520-0469(1987)044<2418:OTROSS>2.0.CO;2044%3C2418:OTROSS%3E2.0.CO;2)
5. Showman, A. P., & Polvani, L. M. (2011). Equatorial superrotation on tidally locked exoplanets. The Astrophysical Journal, 738(1), 71. doi:10.1088/0004-637X/738/1/71