Bjerknes Feedback

The Bjerknes feedback is the foundational positive coupled ocean–atmosphere dynamical instability that amplifies initial sea surface temperature (SST) perturbations across the equatorial Pacific Ocean into mature warm (El Niño) or cold (La Niña) phases of the El Niño-Southern Oscillation (ENSO). Formulated in 1969 by Norwegian-American meteorologist Jacob Bjerknes, the mechanism resolved a long-standing paradox in tropical climatology by demonstrating that the oceanic El Niño warming along the coast of Peru and Ecuador and Sir Gilbert Walker's atmospheric Southern Oscillation sea-level pressure seesaw between Darwin, Australia, and Tahiti are not independent phenomena, but two mutually reinforcing manifestations of a single coupled air–sea feedback loop.

In the climatological mean state of the equatorial Pacific, persistent easterly trade winds drive westward surface currents and divergent equatorial Ekman transport, piling warm surface water into the Western Pacific Warm Pool (where SSTs exceed 29 °C and the oceanic thermocline reaches depths of 150 to 200 m) while shoaling the thermocline in the Eastern Pacific Cold Tongue (where SSTs drop to 20–23 °C and the 20 °C isotherm lies only 30 to 50 m below the surface). This zonal SST difference (Δ*T*_x ≈ 6 to 8 K across 15,000 km) drives the thermally direct Walker Circulation, which in turn sustains the surface easterlies. Because the mean state is maintained by this closed coupling, any perturbation that weakens or strengthens the zonal SST gradient triggers a three-leg positive feedback chain linking atmospheric boundary-layer momentum transfer, equatorial wave propagation, and upper-ocean heat advection.

Leg 1: Atmospheric Wind-Stress Response to Zonal SST Gradients

The first leg of the Bjerknes feedback governs how the tropical troposphere responds to a sea surface temperature anomaly (T′_E) in the central and eastern equatorial Pacific (commonly monitored in the Niño-3 region, 5°S–5°N, 150°W–90°W, or Niño-3.4, 170°W–120°W). When a warm SST anomaly (T′_E > 0) reduces the east–west zonal temperature gradient, enhanced surface evaporation and moist static energy destabilize the atmospheric boundary layer over the central Pacific, shifting deep cumulonimbus convection and latent heat release eastward.

In accordance with the Gill (1980) Matsuno–Gill equatorial beta-plane model, an equatorial diabatic heating anomaly excites a pair of off-equatorial cyclonic Rossby gyres to the west and an atmospheric Kelvin wave to the east, generating a low-level westerly wind-stress anomaly (τ′_x > 0) across the central Pacific (typically within the Niño-4 box, 160°E–150°W). Linearized around the climatological mean, this atmospheric coupling is expressed as:

τ′_x = μ · T′_E

where μ (in N·m⁻²·K⁻¹, typically ~0.01 to 0.02 N·m⁻²·K⁻¹) is the atmospheric wind-stress coupling sensitivity coefficient.

Leg 2: Oceanic Thermocline Slope, Ekman Upwelling, and Zonal Advection

The second leg translates the central Pacific zonal wind-stress anomaly (τ′_x) into upper-ocean dynamical adjustments across the equatorial waveguide. On sub-annual timescales, the vertically integrated zonal pressure gradient in the upper ocean balances the surface zonal wind stress via the Sverdrup balance:

g′ · (∂*h* / ∂*x*) ≈ τ_x / (ρ₀ · H)

where h is the depth of the 20 °C isotherm representing the thermocline, g′ = g(Δρ/ρ₀) ≈ 0.03 to 0.05 m·s⁻² is the reduced gravity across the pycnocline, ρ₀ ≈ 1,025 kg·m⁻³ is seawater density, and H is the mean mixed-layer/upper-layer depth. A westerly wind-stress anomaly (τ′_x > 0) excites downwelling equatorial oceanic Kelvin waves that propagate eastward along the equator at phase speeds c_K = (g′*H*)^1/2 ≈ 2.5 to 3.0 m·s⁻¹ (crossing the Pacific basin in 60 to 75 days), deepening the eastern Pacific thermocline (h′_E > 0), while westward-propagating Rossby waves shoal the western Pacific thermocline (h′_W < 0).

Simultaneously, because the Coriolis effect changes sign across the equator (f = β*y*, where β = 2Ω/R_E ≈ 2.28 × 10⁻¹¹ m⁻¹·s⁻¹), weakened easterly trade winds reduce poleward Ekman transport on both sides of the equator, suppressing vertical Ekman upwelling (w′ < 0) and inducing anomalous eastward surface geostrophic and Ekman currents (u′ > 0).

Leg 3: Subsurface Thermal Coupling to Mixed-Layer Temperature

The third leg closes the positive feedback loop by converting these oceanic dynamical anomalies into further eastern Pacific mixed-layer warming (∂*T*′_E/∂*t* > 0). The linearized mixed-layer temperature tendency equation in the eastern and central equatorial Pacific decomposes the Bjerknes growth rate into three distinct oceanic feedback sub-terms:

∂*T*′ / ∂*t* = − w̄ · (∂*T*′_sub / ∂*z*) − w′ · (∂*T̄* / ∂*z*) − u′ · (∂*T̄* / ∂*x*) − α_s · T′

1. Thermocline Feedback [−*w̄*·(∂*T*′_sub/∂*z*)]: Even when mean climatological upwelling (w̄ > 0) continues, a deeper eastern thermocline (h′_E > 0) causes the upwelled water entering the mixed layer from below (T′_sub = γ*h*′_E) to be substantially warmer than normal. This term dominates in the Eastern Pacific (Cold Tongue) where the mean thermocline is shallow.
2. Ekman Upwelling Feedback [−*w*′·(∂*T̄*/∂*z*)]: Reduced easterly wind stress weakens the vertical upwelling velocity (w′ < 0) across the sharp mean vertical temperature gradient (∂*T̄*/∂*z* > 0), curtailing dynamical cooling of the surface layer.
3. Zonal Advective Feedback [−*u*′·(∂*T̄*/∂*x*)]: Anomalous eastward surface currents (u′ > 0) advect warm water from the Western Pacific Warm Pool eastward across the strong mean zonal SST gradient (∂*T̄*/∂*x* < 0). This term dominates in the Central Pacific (Warm Pool edge) and drives Central Pacific ("Modoki") El Niño events.

The Jin (1997) Recharge–Discharge Oscillator and BJ Index

If the positive Bjerknes feedback operated in isolation without negative restoring mechanisms or thermodynamic damping, the equatorial Pacific would lock permanently into either a perennial warm state or a perennial cold state. In the Recharge–Discharge Oscillator formulated by Fei-Fei Jin (1997), the Bjerknes instability is coupled to basin-wide meridional Sverdrup transport of equatorial ocean heat content (represented by the zonal-mean equatorial thermocline depth anomaly h′_W):

d*T*′_E / d*t* = R · T′_E + γ · h′_W

d*h*′_W / d*t* = − ε · h′_W − α · T′_E

Here, the Bjerknes Stability Index (R or BJ index) Developed by Jin et al. (2006) sums the three positive dynamic feedbacks (thermocline, zonal advective, and Ekman upwelling feedbacks) minus the negative thermodynamic surface heat-flux damping (α_s ≈ 10 to 25 W·m⁻²·K⁻¹, driven by enhanced latent heat evaporation and deep convective cloud-albedo shielding of solar shortwave radiation) and mean dynamical advection damping. When R approaches or exceeds zero, SST anomalies grow rapidly until the westerly wind-stress curl discharges warm upper-ocean water poleward out of the equatorial band (d*h*′_W/d*t* < 0), shoaling the basin-mean thermocline, cutting off subsurface warmth, terminating El Niño, and overshooting into the opposite La Niña phase—where an identical Bjerknes feedback loop operates with reversed algebraic signs (T′_E < 0 → τ′_x < 0 → h′_E < 0).

Why is the Bjerknes feedback symmetric in sign yet El Niño is typically stronger than La Niña?

While linearized equations treat the Bjerknes feedback as symmetric between warm and cold phases, observed El Niño SST anomalies (+2.5 to +3.0 °C in Niño-3.4 during 1997–98 and 2015–16) exceed extreme La Niña anomalies (−1.5 to −2.0 °C). This amplitude asymmetry arises from two nonlinearities: (1) deep atmospheric convection requires absolute SSTs above a radiative-convective threshold of ~27.5 °C, making westerly wind responses to eastern warming stronger than easterly responses to cooling, and (2) during La Niña, once the eastern thermocline shoals near the surface, subsurface temperatures saturate near deep-water values and cannot cool proportionally further.

Does the Bjerknes feedback operate in the tropical Atlantic or Indian Ocean?

Yes. In the equatorial Atlantic, an analogous coupled air–sea mechanism known as the Atlantic Niño (or Atlantic Zonal Mode) operates between the western Atlantic warm pool and the eastern Gulf of Guinea cold tongue during boreal summer, though it is weaker and shorter-lived due to the narrower basin width (~5,000 km vs. 15,000 km in the Pacific). In the Indian Ocean, Jacob Bjerknes's positive air–sea coupling drives the Indian Ocean Dipole (IOD) between the Sumatra–Java upwelling region and the western tropical Indian Ocean during boreal autumn.

How does anthropogenic climate change affect the strength of the Bjerknes feedback?

Coupled general circulation models (CMIP6) project two competing changes to the Bjerknes Stability Index under greenhouse warming: enhanced vertical thermal stratification of the upper equatorial ocean increases the subsurface temperature gradient (∂*T̄*/∂*z*), strengthening the positive thermocline and zonal advective feedbacks, while a weakened mean tropical Walker Circulation and stronger evaporative latent heat damping (α_s, governed by the Clausius–Clapeyron relation) oppose growth, yielding an overall increase in the frequency of extreme Eastern and Central Pacific ENSO events.

1. Bjerknes, J. (1969). "Atmospheric teleconnections from the equatorial Pacific." Monthly Weather Review, 97(3), 163–172. https://doi.org/10.1175/1520-0493(1969)097<0163:ATFTEP>2.3.CO;2097%3C0163:ATFTEP%3E2.3.CO;2)
2. Jin, F.-F. (1997). "An equatorial ocean recharge paradigm for ENSO. Part I: Conceptual model." Journal of the Atmospheric Sciences, 54(7), 811–829. https://doi.org/10.1175/1520-0469(1997)054<0811:AEORPF>2.0.CO;2054%3C0811:AEORPF%3E2.0.CO;2)
3. Jin, F.-F., Kim, S. T., & Bejarano, L. (2006). "A coupled-stability index for ENSO." Geophysical Research Letters, 33(23), L23708. https://doi.org/10.1029/2006GL027221
4. Timmermann, A., An, S.-I., Kug, J.-S., Jin, F.-F., Cai, W., Capotondi, A., Cobb, K. M., Lengaigne, M., McPhaden, M. J., Stuecker, M. F., Stein, K., Wittenberg, A. T., Yun, K.-S., Bayr, T., Chen, H.-C., Chikamoto, Y., Dewitte, B., Dommenget, D., Grothe, P., Guilyardi, E., Ham, Y.-G., Hayashi, M., Ineson, S., Kang, D., Kim, S., Kim, W. M., Lee, J.-Y., Li, T., Luo, J.-J., McGregor, S., Planton, Y., Power, S., Rashid, H., Ren, H.-L., Santoso, A., Takahashi, K., Todd, A., Wang, G., Wang, G., Xie, R., Yang, W.-H., Yeh, S.-W., Yoon, J., Zeller, E., & Zhang, X. (2018). "El Niño–Southern Oscillation complexity." Nature, 559(7715), 535–545. https://doi.org/10.1038/s41586-018-0252-6