FIELD NOTE 16 / EL NIÑO & LA NIÑA

An ocean moves.
The weather follows.

Give the winds a push. Follow the response below the surface—and far beyond the Pacific.

Speed Month 0.0

Near-average Pacific Neutral thermocline

Colour shows temperature; rising tracers on the right illustrate upwelling. Gold marks the thermocline; the dashed line is its neutral position. Tap the ocean to sample the model, or focus the section and use the arrow keys.

The upper warm layer is deeper in the west. Cold water rises nearer the surface in the east.

Eastern Pacific anomaly: +0.00 °C
Eastern thermocline depth: 70 m
Actual surface wind: 6.0 m/s westward

At neutral conditions, trade winds maintain a deeper warm layer in the western Pacific. Try a short wind pulse.

Watch the delayed response.

The ocean changes shape.

EastWest

Plots retain the latest 60 model months. The ±0.5 °C lines are display guides, not an operational ENSO diagnosis.

Change a control while the model runs. Only the labelled restart controls clear its history.

A PACIFIC YOU CAN EXPLORE

Move the winds.

Ocean temperatures and the depth of the warm layer respond over months. They carry a memory of earlier winds.

Restart from a starting state

These restart time and restore typical forcing with coupling on. El Niño and La Niña are illustrative starting states, not historical events.

Background trade-wind forcing100%

Weaker · 20%Stronger · 180%

100% applies no extra forcing. With coupling on, the actual winds also respond to the ocean. This changes future evolution; it does not reset the ocean.

Ocean–atmosphere feedback

Coupled: a warmer eastern ocean weakens the model’s easterly winds, which can reinforce the warming.

A TEMPORARY DISTURBANCE

Give it a nudge.

Each pulse lasts two model months. Westerly opposes the usual trade winds; easterly strengthens them.

Warm-water reservoir anomaly: 0.0 m
Upwelling indicator: 100% of neutral

This is a conceptual model with chosen coefficients. It demonstrates feedback and ocean memory; its dates, temperatures and event strength are not forecasts. The regional map uses documented tendencies, separately from the ocean calculation.

Optional sound is a gentle wind-like texture that follows the model wind. The top-right preference is shared with all experiments.

Follow the weather connection ↓

FROM TROPICAL RAIN TO DISTANT WEATHER

One Pacific. Many connections.

Moving tropical convection changes atmospheric circulation. The jet stream and storm tracks can respond thousands of kilometres away.

Showing tendencies for near-average Pacific conditions.

Choose a numbered marker or a region. Marker colours indicate direction, not magnitude or probability. Unmarked areas are not necessarily unaffected.

Explore a region

December–February

Canadian Prairies · Saskatoon

Temperature: No broad tendency shown
Rainfall: No broad tendency shown

These are broad seasonal associations, not predictions for a city or a particular storm. Grey means no broad tendency is supplied here.

NOAA seasonal impact maps ↗

REAL OBSERVATIONS / SASKATOON

Same Pacific.
Different winters.

Each dot is one observed December–February season. Compare the tropical Pacific index with winter temperature at Saskatoon airport.

Select a winter in the table to highlight its dot.

Observed winters · select one to inspect
Winter	Pacific ONI	Saskatoon anomaly

Bundled records: 2009–10 to 2024–25. Saskatoon anomalies use the airport’s 1991–2020 monthly normals. Pacific values are historical ERSSTv6 ONI; the two indices use different reference methods. This small, recent sample illustrates variability and does not estimate long-term ENSO probabilities. Download the comparison CSV ↗

Where these observations come from

Temperature: Environment and Climate Change Canada’s monthly observations for Saskatoon RCS, station 47707, climate ID 4057165. Each winter combines the preceding December with January and February. We weight monthly mean temperatures by their number of valid daily means. Months need at least 90% coverage; missing days are not filled. The reference combines the published airport December (−13.0 °C), January (−15.4 °C) and February (−13.4 °C) normals with the same weights. The station and the airport normal composite are not an adjusted homogeneous climate series.

Pacific: NOAA’s December–February Oceanic Niño Index from ERSST version 6, using Niño 3.4 sea-surface temperature anomalies and NOAA’s changing reference periods. The warm/cool filters use that winter’s ±0.5 °C threshold only; they do not apply the duration or atmospheric requirements of an official ENSO event classification. NOAA now uses the Relative Oceanic Niño Index (RONI) for operational monitoring; this comparison deliberately uses the historical ONI record.

Sources retrieved 22 September 2026. Observations stay separate from the conceptual model; changing a model control never changes them. Recent warming, other climate modes and daily weather all contribute to the scatter. These dots cannot attribute a particular winter to ENSO.

Make a prediction. Change one thing. See what happens.

Model notes & references ↗

FOLLOW THE SCIENCE

The ocean remembers.
The atmosphere responds.

Back to the controls ↑

El Niño and La Niña are the warm and cool phases of the El Niño–Southern Oscillation (ENSO), a coupled pattern in the tropical Pacific. “Coupled” means that ocean and atmosphere affect one another. Neutral conditions are the third state; they do not mean that the weather stops varying.

A FEW QUESTIONS TO TRY

Follow cause and consequence.

A short push. A long response.

Begin neutral, then apply a two-month westerly wind pulse. At month 6, the pulse is over. Has the ocean returned to where it started?

Let the ocean talk back.

Repeat exactly the same pulse with the atmospheric feedback off. Compare the temperature at month 6 with the coupled trial. The difference comes from feedback.

Same Pacific, different winters.

Compare 2009–10 and 2023–24. Both had a substantially warm tropical Pacific. Did Saskatoon have the same kind of winter?

01 / THE USUAL ARRANGEMENT

Winds tilt the warm layer.

The equatorial trade winds normally blow westward. Warm surface water accumulates towards Indonesia. In the east, colder water lies closer to the surface and upwelling helps keep surface temperatures lower. Moist air tends to rise over the western warm pool, feeding clouds and rain; air returns eastward aloft as part of the Walker circulation.

02 / A REINFORCING LOOP

Warm water changes the winds.

When the central and eastern Pacific warm, the east–west temperature contrast weakens. Changes in convection and pressure can weaken the trades. The eastern thermocline deepens and upwelled water cools the surface less effectively, reinforcing the warming. This is the Bjerknes feedback. La Niña involves stronger trades and an enhanced cool eastern Pacific.

03 / MEMORY BELOW THE SURFACE

A short disturbance can outlast itself.

The ocean takes time to adjust. Redistribution and loss of equatorial warm water can eventually oppose an El Niño event. The model includes a slow reservoir variable to represent this delayed response. Real events also involve equatorial ocean waves, seasonal influences and irregular wind bursts. They are not a perfectly repeating clock.

04 / THE DISTANT CONNECTION

Rain can reorganize circulation.

Condensation in tropical thunderstorms releases heat into the atmosphere. When the preferred storm region shifts, this heating can change large-scale atmospheric waves, jet streams and storm tracks. These distant connections are called teleconnections. Their effects depend on region and season; a winter association is not automatically a summer one.

The model, mathematics & limits

Three evolving quantities

This is a reduced recharge-oscillator teaching model. T is an eastern Pacific temperature anomaly (°C); H is the mean thermocline-depth anomaly (m), used as a proxy for equatorial warm-water storage; and W is a dimensionless weakening of the usual easterly wind. Positive W weakens the trades; negative W strengthens them. Time t is in model months.

dT/dt = −0.18T + 0.48W + 0.025H − 0.02T³
dH/dt = −1.2T − 0.04H
dW/dt = (Wtarget − W) / 0.6
Wtarget = clamp(F + P + 0.65CT, −3, +3)

C is 1 with feedback on and 0 with it off. F = (100 − trade-forcing percentage)/60; P is +1.4 for a westerly pulse or −1.4 for an easterly pulse, lasting two months. Temperature damping, wind-driven heating, warm-phase reservoir discharge (and cool-phase recharge), a slow reservoir and nonlinear saturation produce delayed, bounded responses. Coefficients have the units required by these equations and are chosen for an understandable demonstration, not calibrated to an observed event.

Fourth-order Runge–Kutta advances in steps no larger than 1/120 month, splitting exactly at pulse endings and history samples. Live control changes preserve T, H, W, time and history. Pause freezes all model motion; Play continues from the same state. Only the starting-state and guided-trial buttons restart.

What the section shows

Actual wind = −6 + 4W m/s (east positive). The west/east thermocline depths are 135 + H ± 65(1 − 0.65W) m, bounded to 20–290 m for display. Surface temperatures are 29 − 0.18T °C in the west and 24 + T °C in the east. A smooth vertical transition towards 9 °C draws the temperature field. The upwelling indicator is clamp(1 − 0.4W, 0.08, 2.2) relative to neutral. Rain-cloud position and circulation tracers are illustrative functions of the state.

The section does not solve fluid dynamics, local heat conservation, individual clouds, currents, sea level, salinity or the full global energy budget. H is a reservoir proxy, not a measured ocean heat content. The fixed section has no seasonal heating cycle; choosing a season changes the documented weather comparison only. The perfectly symmetric warm/cool equations simplify a real system whose phases and impacts are not mirror images.

What the map means

Its qualitative markers summarize selected NOAA seasonal associations. The jet-stream line is a schematic Pacific storm-track illustration for December–February; its exact latitude is not computed. Following the ocean uses T ≥ +0.5 °C, T ≤ −0.5 °C or the interval between as simple view selectors. These instantaneous model thresholds are not ONI, RONI or an official event diagnosis. Region markers have no implied probability or spatial boundary. The observations underneath are independent measured records.