Topic Brief: Understanding the El Niño-Southern Oscillation
The El Niño-Southern Oscillation (ENSO) is a natural variation in sea surface temperature and winds across the Pacific Ocean that influences global weather patterns. ENSO events typically last more than six months and can alter the likelihood of extreme weather, including heatwaves, drought, floods and bushfires.
What is 21st Century Weather doing?
We understand how ENSO affects seasonal climate, but its impact on Australia’s day-to-day weather systems is less clear. The Centre is working to advance understanding in:
- the relationship between El Niño and the clustering of precipitation across the tropics,
- the connection between ENSO and rapid swings between extreme wet and dry conditions,
- the influence of ENSO on weather systems driving global and regional rainfall, including those associated with different types of events and multiyear events, and
- the impact of the projected warming pattern on projected ENSO teleconnection changes.
The mechanisms driving ENSO’s variability on decadal timescales are not fully understood and may involve a combination of ocean dynamics, atmospheric variability and stochastic processes. The Centre is also working to clarify why transitions from El Niño to La Niña are more reliable than the other direction, how ENSO interacts with and responds to conditions in other ocean basins, and how confident we are in artificial intelligence predictions of early ENSO events.
Climate models capture many aspects of ENSO, and their representation has improved overall in recent generations. However, systematic biases remain, including imbalances in feedback processes. The Centre is working to answer how climate change is influencing short-term ENSO forecasts, long-term variability and projections under high-emission and low-emission net-zero futures.
How El Niño and La Niña affect Australia
El Niño usually brings hotter and drier conditions to Australia, while La Niña tends to bring more rainfall and cooler temperatures. Yet, Australia’s climate and weather are driven by multiple factors, and no two El Niño or La Niña events are exactly alike. They vary in duration, sea surface temperature patterns, wind changes, and interactions between the Pacific Ocean with the Indian and Southern Oceans. As a result, the phase of ENSO alone does not guarantee a wet or dry year.
An El Niño has formed in 2026, but how do we know this?
El Niño is characterised by warmer-than-normal sea surface temperatures (SSTs) in the eastern and central tropical Pacific as well as a concurrent weakening of easterly trade winds across the Pacific Ocean and a shift in the location of storm activity (called the Walker Circulation) away from Australia. Therefore, an El Niño can be monitored and identified by SST anomalies—warmer-than-usual temperatures—in the central equatorial Pacific known as Niño3.4 (5°N-5°S, 170°W-120°W).
The Bureau of Meteorology (BoM) declared an El Niño in June 2026. All conditions expected for an El Niño have been met (see Figure 1): sustained warming in the central and eastern Pacific, accompanied by atmospheric criteria such as the Southern Oscillation Index, which measures the position of the Walker Circulation, and trade wind speed.
To assess these criteria and define whether an El Niño is developing, scientists gather information from a range of sources:
Observations from continuous monitoring, such as satellites, and in situ measurements using buoys. Satellites track SSTs, sea surface height and surface winds across the entire tropical Pacific Ocean basin, while moored buoy arrays, Argo floats and drifters measure subsurface temperatures and ocean currents.
A range of seasonal forecasting and long-range forecasting models. These models help us anticipate whether conditions in the tropical Pacific will strengthen, persist or weaken months in advance. Forecasters typically look at an ensemble of international models rather than a single one to make predictions.
Dynamical models are based on fundamental laws of physics that describe how the atmosphere and ocean interact and evolve over time. Supercomputers solve thousands of equations to simulate how ENSO may evolve. As such, these models can capture events or processes that are physically possible but have not been observed before. However, they require far more computational power than a statistical model.
Statistical models use historical relationships between past and present ocean-atmosphere conditions and can also provide valuable information on how ENSO will evolve. They are computationally cheap and perform well when ENSO behaves like it has in the past, but struggle with unusual or unprecedented states, such as a very strong El Niño event.
Measuring ENSO in a warming climate
The historical sea surface temperature (SST) threshold that identifies ENSO events may not be a useful indicator of events in a warming climate. For this reason, BoM and the US National Oceanic and Atmospheric Administration (NOAA) now use a relative index to measure the SST anomaly in the Niño3.4 region. This change helps identify ENSO events in a warming climate with more confidence.
Figure 1. Warm surface waters accumulate in the eastern Pacific, and the atmosphere responds with the weakening of surface winds, causing even
more warming in the east.
How does El Niño affect the weather?
ENSO acts on a global scale. A physical change in the Pacific Ocean, such as the extent and longitude of warming, alters where heat enters the atmosphere. Like a ripple in a pond, the variation in SST and winds across the Pacific Ocean alters large-scale atmospheric circulation patterns and, in doing so, influences the likelihood of different weather systems occurring. In other words, as this heat perturbs atmospheric circulation, it creates different pressure patterns outside the tropical Pacific, which in turn drive different weather systems, more or less clouds and rain, and temperature changes. The indirect influence of ENSO on regional weather patterns via atmospheric circulations is known as teleconnections.
The evolution of the El Niño-Southern Oscillation
El Niño Southern Oscillation is a very complex system that emerges from interactions between the ocean and the atmosphere. These events can look quite different from one another spatially and temporally. Spatially, the warm sea surface temperature anomaly can be closer to the Central Pacific or Eastern Pacific Ocean, and this difference can affect the resulting weather impacts. El Niño events can also evolve over time in linear or staggered ways (Figure 2).
We can only predict El Niño development probabilistically. In other words, we discuss the likelihood of an El Niño forming. This likelihood can change over time as the various components evolve. Even with a high likelihood, an El Niño may not form. This is how probabilities work: a less likely event might still come to fruition. This is why we must be cautious with early forecasts.
The 2026 El Niño has formed. However, how big it will be and the consequences it will bring remain uncertain as the event evolves, in this case, as of August 2026.
Figure 2. Central equatorial Pacific SST anomalies for El Niño events between 1980 and 2026 (1982, 1986, 1987, 1991, 1994, 1997, 2002, 2004, 2006, 2009,
2015, 2018, 2023 & 2026). Years highlighted by colour are 1982 (magenta), 1991 (red), 2002 (orange), 2015 (cyan), and 2026 (dark blue), while the
average event is presented in black and all other event years are plotted as gray.
Why doesn’t a strong El Niño always lead to major impacts?
The weather is governed by physical laws that constrain its behaviour. However, these physical laws can induce chaotic behaviour, some of which–including weather patterns–act on such short timescales that it can appear to us as randomness.
The unpredictable nature of the interactions between the atmosphere and the ocean that shape ENSO means that the El Niño phenomenon contains natural climate variability that manifests differently for each individual event, so no two El Niños are the same, suggesting that we shouldn’t expect their climatic impacts to be exactly the same. El Niños differ in intensity, geographical location and structure, duration and timing, their interaction with other climate drivers and predictability.
In addition, the randomness and nonlinearity of ocean processes that shape El Niño’s development can lead to differences in the strength and spatial pattern of sea surface temperature (SST) anomalies. The different SST expressions affect weather patterns differently across the Pacific. For example, El Niño events where the maximum equatorial ocean warming is in the central Pacific (i.e., closer to Australia) affect Australian rainfall more consistently than when the ocean warming is in the east Pacific.
El Niño background conditions make drier weather systems more likely and wetter systems less likely in northern and eastern Australia. Nonetheless, this doesn’t stop a wet system or guarantee dry conditions; those changes occur through weather systems. For example, even with severe El Niño background conditions, a synoptic-scale weather system needs to be present to create the perfect storm of dangerous fire weather or heatwave conditions.
The main reason the magnitude of an El Niño is largely unrelated to the magnitude of rainfall anomalies is that rainfall distributions are bounded and non-normal. In other words, if it’s already pretty dry on average, it can’t get much drier.
Beyond the variability within ENSO and its impacts, ENSO is just one ingredient in the recipe that makes up the weather we feel on the ground. In addition to the chaotic nature of weather patterns in general, other factors, such as the Indian Ocean Dipole, can interact with ENSO and modulate its influence on the weather.
What we know
We know El Niño shifts the odds toward warmer, drier conditions, but even if we knew how strong an event would get or its exact spatial structure, we could not precisely predict Australia’s rainfall. El Niño is a climate driver and therefore influences the odds of wetter or drier months or seasons, but the precise day-to-day weather is difficult to predict beyond 7-10 days, even during El Niño. That’s why a strong weather system like a tropical cyclone can disrupt our expectations for a season, like the wet El Niño in 2023.
El Niño is already in place, and most models indicate it will persist until early 2027. Based on forecasts from multiple meteorological centres, this event may be the strongest ever recorded. However, it is too early to tell exactly how strong the impacts will be. From an Australian perspective, the exact magnitude of the event is not critical. Past observations suggest that the individual weather systems Australia experiences at any given point will dictate temperature and rainfall more than the strength of an El Niño event.
Thinking further ahead toward the mid to late 21st century, models suggest that the impacts of ENSO events will increase across most regions affected by ENSO due to climate change. We also know that projected changes in teleconnections depend not only on the changes in ENSO itself, but also on the background spatial warming pattern of the tropical Pacific: whether it will warm faster in the east (closer to South America) or in the centre (closer to Australia). Climate models largely agree that the east Pacific will warm faster than the west; however this projection remains the subject of ongoing research.
Current limitations
Current-generation climate models display varying skill in representing historical warming patterns in the ocean regions relevant to ENSO, compromising future projections. Existing observational gaps aggravate this, including the lack of moorings in the western equatorial Pacific (west of 165°E) to observe surface winds and the ocean’s subsurface. Previous moorings have been removed and to date no nation has filled that gap. This region is crucial for ENSO development.
Glossary
Teleconnections: The link between a climate driver and weather in a distant part of the world.
Ensembles: Single-model ensembles consist of multiple realisations of one model. Multi-model ensembles consist of multiple realisations of climate models. Multi-model single-model ensembles have both. Perturbed parameter ensembles consist of a single model with perturbed parameters that are different in each realisation.
Contact us to learn more about our research projects at hello@21centuryweather.org.au. You can download a PDF copy of the Topic Brief here.