Stommel Gyre Model

Stommel Gyre Model (often termed the Stommel westward intensification model) is the foundational analytical theory of large-scale wind-driven ocean circulation formulated by the American physical oceanographer Henry Melson Stommel in 1948 (Transactions, American Geophysical Union). Prior to Stommel's derivation, physical oceanographers could not explain why the subtropical ocean gyres of the North Atlantic and North Pacific are strongly asymmetric in the east–west direction—exhibiting slow, broad equatorward drift across thousands of kilometers of the basin interior alongside narrow, high-velocity poleward western boundary currents such as the Gulf Stream, the Kuroshio, the Brazil Current, and the Agulhas Current. By solving the steady, linearized, depth-integrated barotropic vorticity equation in a rectangular ocean basin subject to sinusoidal zonal wind stress and linear bottom friction (r), Stommel proved that westward intensification is a direct dynamical consequence of the meridional variation of the planetary Coriolis parameter (β = df/dy). On a non-rotating Earth (f = 0) or a uniform f-plane (β = 0), wind-driven gyre streamlines remain symmetric about the central meridian of the basin; on a planetary β-plane (β > 0), planetary vorticity advection crowding streamlines against the western continental margin creates a narrow frictional Stommel boundary layer of characteristic width δ_S = r/β.

Observational hydrography in the early twentieth century established that subtropical ocean gyres circulate anticyclonically under midlatitude westerlies and tropical trade winds, yet their mass transport is strongly concentrated along western basin margins:

• In the North Atlantic, the Gulf Stream transports 30×10⁶ m³ s⁻¹ (30 Sverdrups, Sv) through the Florida Straits and 85–150 Sv off Cape Hatteras within a 50–100 km band, whereas the compensating southward interior drift spans >4,000 km at 1–5 cm s⁻¹.
• In the North Pacific, the Kuroshio carries 40–65 Sv northward along Taiwan and Japan at speeds exceeding 1.5 m s⁻¹.

In 1947, Harald U. Sverdrup showed that within the frictionless ocean interior, meridional mass transport V is governed by the surface wind-stress curl via the Sverdrup relation βV = curl_z(τ) / ρ₀, driven by Ekman pumping and Ekman transport. Because the Sverdrup relation is first-order in x, it can satisfy the no-normal-flow boundary condition at only one meridional wall (the eastern boundary x = L_x) and cannot close the basin mass budget. Stommel (1948) resolved this by introducing bottom frictional dissipation into the planetary vorticity balance across a closed rectangular basin.

Linearized Barotropic Momentum and Streamfunction Equations

Stommel idealized a homogeneous ocean basin of horizontal dimensions 0 ≤ x ≤ L_x (eastward) and 0 ≤ y ≤ L_y (northward), uniform depth H, and density ρ₀. For small Rossby number, the steady depth-integrated horizontal momentum equations with linear Rayleigh bottom drag (coefficient r, in s⁻¹) and wind stress τ = (τ^x, τ^y) are:

− f V = − gH ∂η / ∂x + τ^x / ρ₀ − r U

+ f U = − gH ∂η / ∂y + τ^y / ρ₀ − r V

where (U, V) = (Hu, Hv) denote depth-integrated volume transports (m² s⁻¹), η(x, y) is sea-surface elevation, and f(y) = f₀ + βy is the Coriolis parameter on the β-plane, with β = (2Ω / R_E) cos φ₀ ≈ 2×10⁻¹¹ m⁻¹ s⁻¹. Nondivergent horizontal transport (∂U/∂x + ∂V/∂y = 0) defines a transport streamfunction ψ(x, y) with U = −∂ψ/∂y and V = +∂ψ/∂x, subject to ψ = 0 on all four walls (x = 0, L_x; y = 0, L_y).

The Stommel Vorticity Equation

Cross-differentiating the momentum equations eliminates the pressure gradient (−gH ∇η) and yields the Stommel barotropic vorticity equation:

r ∇²ψ + β ∂ψ / ∂x = (1 / ρ₀) ( ∂τ^y / ∂x − ∂τ^x / ∂y ) = (1 / ρ₀) curl_z(τ)

Each term reflects a key component of the potential vorticity budget:

1. Planetary vorticity advection (β ∂ψ/∂x = βV): Meridional flow across latitude circles alters the planetary vorticity f(y) of water columns.
2. Frictional vorticity dissipation (r ∇²ψ = rζ): Bottom drag damps relative vorticity ζ = ∇²ψ.
3. Wind-stress curl forcing ((1/ρ₀) curl_z(τ)): Anticyclonic wind-stress curl injects negative vorticity via convergent Ekman transport and downward Ekman pumping.

Analytical Solution for Sinusoidal Zonal Wind Stress

To represent subtropical trade winds (y = 0) and midlatitude westerlies (y = L_y), Stommel prescribed zonal wind stress τ^x(y) = −τ₀ cos(πy/L_y) and τ^y = 0, giving curl_z(τ) = −(τ₀π/L_y) sin(πy/L_y) < 0. For weak friction (r ≪ βL_x), matched asymptotic expansion yields:

ψ(x, y) ≈ ( τ₀ π / (ρ₀ β L_y) ) (L_x − x) ( 1 − e^−x/δ_S ) sin(π y / L_y)

where the characteristic Stommel boundary-layer thickness is:

δ_S = r / β

For typical ocean values (r ≈ 10⁻⁶ s⁻¹, β ≈ 2×10⁻¹¹ m⁻¹ s⁻¹), δ_S ≈ 50 km—roughly 1 percent of a 5,000 km basin (L_x). Because (1 − e^−x/δ_S) rises from 0 at x = 0 to 1 across x ∼ δ_S, all equatorward Sverdrup transport accumulated across the wide interior returns poleward inside the narrow western boundary layer of width δ_S.

Closed-Basin Vorticity Balance

In a steady subtropical gyre forced by clockwise (negative) wind-stress curl, the wind imparts negative vorticity across the interior, balanced in the Sverdrup interior by equatorward flow (βV < 0). For fluid parcels to complete a closed circuit with unchanged vorticity, net negative wind vorticity must be canceled by positive frictional vorticity in a boundary current:

• Western Boundary Return Flow (V > 0): Velocity rises from zero at the coast to a strong northward peak offshore and decays into the interior (∂V/∂x < 0 in the outer shear zone), generating positive frictional torque (−rζ > 0) that balances both the negative wind curl and planetary vorticity gain (βV > 0).
• Hypothetical Eastern Boundary Return Flow (V > 0): Northward flow along the eastern wall would require negative relative vorticity (ζ < 0) offshore to match the coastal decay, failing to close the vorticity budget without unphysical basin-scale recirculation.

Arrested Planetary Rossby Waves

From wave dynamics (Pedlosky, 1987), large-scale vorticity perturbations propagate as planetary Rossby waves. Long Rossby waves (k ≪ l) carry energy westward at group velocity c_gx < 0 until reflecting from the western wall as short-wavelength, eastward-group-velocity Rossby waves (k ≫ l). Steep velocity gradients in these short reflected waves are damped by bottom friction r within δ_S = r/β of the western coast.

Subsequent extensions by Walter Munk (1950), Jule Charney (1955), and Nick Fofonoff (1954) replaced or supplemented Stommel's linear bottom friction with lateral eddy viscosity (A_H ∇⁴ψ) and nonlinear inertial advection (J(ψ, ∇²ψ)).

Although the Stommel 1948 model captures the first-order physics of western boundary intensification, its linear bottom-friction parameterization assumes that the wind-driven circulation extends all the way to the seafloor. In the stratified ocean, the wind-driven subtropical gyre is largely confined above the main thermocline (upper 500–1,000 m), where lateral eddy viscosity (the Munk scale δ_M) and nonlinear inertial advection (δ_I) dominate over bottom drag. Nevertheless, Stommel and Arnold Arons (1960) applied the exact same planetary vorticity logic to the deep abyssal thermohaline circulation, predicting that uniform abyssal upwelling across the global ocean interior forces poleward interior flow balanced by Deep Western Boundary Currents (DWBCs)—a theoretical prediction confirmed observationally in the North Atlantic via neutrally buoyant Swallow floats prior to publication.

Would western boundary currents reverse to the eastern boundary in the Southern Hemisphere?

No. Western boundary currents remain on the western side of ocean basins in both the Northern and Southern Hemispheres (for example, the Brazil Current in the South Atlantic, the East Australian Current in the South Pacific, and the Agulhas Current in the South Indian Ocean). This occurs because westward intensification depends on the sign of β = df/dy = (2Ω / R_E) cos φ, which is positive in both hemispheres since the Coriolis parameter f = 2Ω sin φ increases monotonically from −2Ω at the South Pole through 0 at the Equator to +2Ω at the North Pole.

What happens to the Stommel gyre if the Earth's rotation direction were reversed?

If the planetary rotation vector Ω were reversed (so the Earth rotated from east to west, Ω < 0), the meridional gradient β = df/dy would become negative (β < 0). In the Stommel differential equation r d²X/dx² + β dX/dx = 0, a negative β produces growing solutions e^+|β|x/r toward the eastern wall x = L_x, shifting all intense boundary currents to the eastern margins of the ocean basins.

Why did Stommel use linear bottom friction instead of lateral viscosity?

Stommel (1948) deliberately selected linear bottom drag (−r U) because it yields a second-order differential operator (r ∇²ψ) that can be solved analytically in closed elementary form with a single boundary condition (ψ = 0) at each wall. Two years later, Walter Munk (1950) replaced bottom drag with lateral eddy viscosity (−A_H ∇⁴ψ), which raises the equation to fourth order, allows no-slip boundary conditions (ψ = ∂ψ/∂n = 0), and adds an offshore recirculation countercurrent of width δ_M = (A_H/β)^1/3.

1. Stommel, H. (1948). "The westward intensification of wind-driven ocean currents." Transactions, American Geophysical Union, 29(2), 202–206. https://doi.org/10.1029/TR029i002p00202
2. Sverdrup, H. U. (1947). "Wind-driven currents in a baroclinic ocean; with application to the equatorial currents of the eastern Pacific." Proceedings of the National Academy of Sciences, 33(11), 318–326. https://doi.org/10.1073/pnas.33.11.318
3. Munk, W. H. (1950). "On the wind-driven ocean circulation." Journal of Meteorology, 7(2), 80–93. https://doi.org/10.1175/1520-0469(1950)007<0080:OTWDOC>2.0.CO;2007<0080:OTWDOC>2.0.CO;2)
4. Charney, J. G. (1955). "The Gulf Stream as an inertial boundary layer." Proceedings of the National Academy of Sciences, 41(10), 731–740. https://doi.org/10.1073/pnas.41.10.731
5. Stommel, H., & Arons, A. B. (1960). "On the abyssal circulation of the world ocean—I. Stationary planetary flow patterns on a sphere." Deep-Sea Research, 6, 140–154. https://doi.org/10.1016/0146-6313(59)90065-690065-6)