Untitled Document

El Niño: Predicting Cause and Effect
Scott Curtis

El Niño and Hurricanes?

Rain and El Niño Print exclusive

Have you heard? El Niño is back again. Or is it? This climate event, a warming of the eastern Pacific, occurs every two to seven years, but doesn’t always capture the headlines or grab the attention of the general public.

In February 1998, high-tide waves crashed against the seawall at an apartment complex in Capitola Beach, one of many powerful winter storms along the California coast attributed to El Niño. At least 27 counties in California were declared national disaster areas, with hundreds of millions of dollars in property losses. El Niño is the biggest player in seasonal climate predictions, and predicting the start of an El Niño itself and what effect it will have on global climate is difficult. Courtesy of Bruce Richmond, USGS.

During the winter of 1997-1998, the world was abuzz with El Niño, as the Pacific reached temperatures never before recorded. Most of the population of the United States was directly affected by that El Niño, and although scientists predicted it was on the way months in advance, they could not accurately predict its exact timing, and they well underestimated its strength.

Prediction capability has since improved, but there are still obstacles scientists must overcome to relay their predictions to decision-makers and the general public. There remains a general lack of understanding about El Niño, which is made worse by the fact that the warming of the Pacific is itself not very important to most of the world — people only care about what happens in their backyard. As a result, scientists must do a better job of communicating the likely effects from a given El Niño.

Simple beginnings

The story of El Niño begins in coastal Peru, where many people make their living fishing the Pacific Ocean. Rising waters in the far eastern Pacific bring low temperatures and abundant nutrients to the surface — conditions ideal for fish. In the winter, however, a southward-flowing warm current disrupts this balance. Peruvian fisherman have traditionally referred to this warming as El Niño or “Christ child,” as it often coincided with Christmas.

Now scientists refer to warm anomalies at any time of the year that occur over a span of several months as El Niño. During these events, the warming is enough to bring an end to the vast fish population and wreak havoc on the Peruvian economy. To top it off, the warm waters are often accompanied by copious rains, which are uncommon in this part of the world, leading to flash floods and landslides.

Over the years, as observations of the Pacific increased, scientists have been able to connect the local El Niño warming off Peru to the rest of the ocean basin. In fact the story goes deeper, literally: The warming that the fishermen experienced at the surface was also occurring at some depth. A deep layer of warm well-mixed water, called the mixed layer, is usually confined to the western Pacific near Australia, but during El Niño this water sloshes to the east. Scientists also began to realize that the ocean-circulation alone could not explain the El Niño phenomenon.

Looking for clues to predict the Asian monsoon in the 1920s, G.T. Walker found a seesaw of atmospheric pressure between the eastern and western Pacific, which he called the Southern Oscillation. Normally, pressure is low in the western Pacific and high in the eastern Pacific, with easterly winds connecting these centers. The pressure centers, however, can be strong some years and weak in other years.

In the late 1960s, J. Bjerknes made the connection between the Southern Oscillation and El Niño. During normal years, the pressure difference between the eastern and western Pacific Ocean is strong, and easterly winds pile up warm waters near Australia. During El Niño, the pressure difference is weak or even reverses. The resulting decrease in easterly wind no longer forces the water to the west, and the warm water makes its way toward South America. By the early 1980s, scientists had a pretty good picture of the El Niño-Southern Oscillation event, which is often shortened to ENSO, but understanding the cause of El Niño is still somewhat of a mystery, making prediction a tricky business.

Taking the ocean’s temperature

An advantage in trying to predict an El Niño is the slow evolution of sea-surface temperature. Weather is often chaotic and changes quickly, making predictions beyond a few days difficult. But the ocean changes at a much slower pace, so it takes a season or more for an El Niño to fully mature.

Satellite images show temperatures of the upper layer of the ocean for January 1997 — a normal month — and November 1997 — an El Niño month. During normal conditions, winds (white arrow) cause an accumulation of warm waters in the western Pacific, leaving cooler waters in the east. Rain preferentially forms over the warmest waters. During the 1997-1998 El Niño, the wind reversed direction and the warmest water and rainfall moved eastward, affecting the Peruvian people and weather around the world. Satellite images courtesy of NASA Goddard Space Flight Center.

The disadvantage in trying to predict an El Niño is that the phenomenon is part of an atmosphere-ocean interaction (a coupled system). You need two to tango. The number of atmospheric and oceanic predictors are large and many are either not well-observed or have a tenuous link to the phenomenon. In addition, computer models that truly couple the atmosphere and ocean are relatively new and primitive.

The object of the majority of ENSO forecast models, whether they be statistical or dynamic, is to forecast temperature anomalies in the Pacific. Oftentimes, they are compared based on their prediction of the “Niño 3.4 index,” which is the average sea-surface temperature anomaly for a location in the central Pacific. This location marks a significant geographical shift from the roots of El Niño in coastal Peru, but the central Pacific is where the climate change of the Pacific is most pronounced, or in other words where ENSO is strongest.

Statistical models combine a set of climate variables through various statistical techniques to produce either a field of data or just one value (for example, the Niño 3.4). Dynamical models use the physics of the atmosphere and ocean to try and represent reality. They usually generate fields of temperature as well as other variables.

Right now, statistical and dynamical models do equally well (or poorly, depending how you look at it) in predicting El Niño. Likely, dynamical models will one day be superior to statistical models and show a level of skill, as computer power increases and we better understand how the ocean and atmosphere communicate with each other.

Global consequences

The relationship between El Niño and the people of Peru is unique in the sense that the El Niño itself (the warm Pacific) and a subsequent effect of El Niño (flooding) both affect the population. Most of the world, however, is not directly affected by a warm ocean near the equator, but instead views El Niño through the window of its global effects.

In the 1980s, with increased globalization and satellite images, scientists began piecing together the effects of El Niño. The reason that El Niño is not just a Pacific problem is because the warming is widespread and the heat that is generated (equivalent to about 4,000 times the energy of the Dec. 26 earthquake that set off the tsunami in the Indian Ocean) causes a ripple effect in the atmosphere, changing the large-scale circulation and storm tracks.

Areas that are usually dry become wet and vice versa. Heavy rains accompany the warming and weakening of the high pressure in the eastern Pacific. Indonesia and northern Australia experience drought and often raging fires, as the western Pacific low pressure center weakens. The horn of Africa becomes wet, and northeast Brazil dries out. In the United States, more rain falls on California and the Southeast. (The recent increase in floods and landslides in California, however, does not appear to be related to the El Niño.) As the El Niño peaks in strength, many of the effects are felt in the tropics first and then “ripple” poleward. Other regions are preferentially affected during certain seasons, and some parts of the planet show no connection to El Niño at all.

El Niño is often perceived as producing only bad weather. Little is ever said about the positive effects of El Niño — one being the decrease in the number of Atlantic hurricanes. In fact, hurricane prediction models use the state of El Niño as one of the input variables. While the 2004 hurricane season was one of the worst on record, the developing 2004-2005 El Niño should be partly credited for a quiet October and November (see sidebar). Furthermore, one study in the Bulletin of the American Meteorological Society calculated that more human lives were saved than lost, and there were more economic gains than losses as a result of the 1997-1998 El Niño.

While some effects are well-known, others are pure invention, which lead to misconceptions within the general public. For example, El Niño is blamed for weird weather and economic and human losses that occur during the event, prompting people to ask, “exactly what isn’t El Niños fault?” While there does seem to be evidence for an increase in worldwide precipitation extremes during El Niño, it is impossible to blame a single storm or outbreak of severe weather on the phenomenon. This is similar to the fallacy that one warm summer is a sign of global warming.

Finally, it is important to remind ourselves that every El Niño is unique. The magnitude and areal extent of warm waters changes from event to event. These factors, in turn, determine if and how the atmosphere changes and whether a certain spot on Earth will feel some effect. Even with these caveats, El Niño is the biggest player in seasonal climate predictions, which give forecasts in percent chance of being dry or wet, warm or cold, over a season or more.

The 2004-2005 El Niño?

The 1990s can be considered the decade of El Niño. Warm conditions were observed almost every month from 1990 to early 1995, and the strongest El Niño on record occurred in 1997 and 1998. This event led scientists to explore the connections between global warming and El Niño. In the process, some scientists began to observe 20-year variations in the strength and number of El Niños. If their hypotheses are correct, we should be entering a period where El Niños will be infrequent and weak.

This pattern seems to be holding true so far, as after the 1997-1998 El Niño, there was a minor El Niño during the winter of 2002-2003. Now there seems to be an even weaker El Niño under way. In December 2004, the Niño 3.4 index was 0.85. The Southern Oscillation Index (SOI) was minus 1.8, and the ENSO Precipitation Index (ESPI) was 0.06. These last two measures indicate how much the atmosphere is teaming up with the ocean. (For comparison, the December 1997 value of Niño 3.4 was 2.78 degrees Celsius, SOI was minus 2.10, and ESPI was 2.19.) Data in January showed a weakening of the current El Niño. This data thus suggest that there may be little to no response in the atmosphere, and that the global weather effects in the coming months will be minimal.

At this time, the Climate Prediction Center predicts that the 2004-2005 El Niño will be essentially finished this spring. The spring season forecast for the United States, however, contains some of the typical lingering effects from a weak El Niño: for example, wetter than normal conditions in the Southwest. This is good news for this part of the country, which has been suffering for a long time under drought conditions.

Predicting the future

Often, El Niño prediction is a matter of reading the tea leaves. A warming here, a wind disturbance there — each might be harbingers of a coming El Niño. And recently developed models seek to find the right combination using new methods.

For example, my colleagues and I have used changes in Indian Ocean precipitation, estimated by satellites, to determine whether an El Niño will appear in the coming year. This has been met with success in the prediction of the last El Niño in 2002, as it appears that climate shifts in the Indian Ocean precede wind disturbances in the western Pacific (see Geotimes, May 2002). How ironic that Walker looked to the Pacific’s Southern Oscillation to predict Indian rainfall, and now scientists are using Indian Ocean rainfall to predict the ENSO. Other simple models have used wind data from the eastern Indian Ocean to western Pacific Ocean and the heat content of the mixed layer in the western Pacific.

Still, El Niño prediction has a long way to go. Given a slight warming of the Pacific, computer models will often attempt to generate full-blown El Niño events. Prediction of the 1997-1998 El Niño was considered unskilled. Although the majority of statistical and dynamical models in use at the time did “forecast” an El Niño, the timing was off by several months, and no model predicted the magnitude of the event.

Besides improving the predictions themselves, scientists are beginning to study how society uses predictions, a function that is influenced by many factors. A study of the fishing sector in five port towns in Peru found that the perceived accuracy of the predictions, access to media outlets and the timing of hearing the forecast were important in the usefulness of the 1997-1998 El Niño prediction.

Once the forecast is heard, decisions and actions can be made based on how the El Niño is expected to affect a certain region. While scientists will continue to rely on what El Niño has done in the past to predict what will happen in the future, attention must be paid to other environmental factors that might affect climate. For example, the impacts of the 1982-1983 El Niño, which was almost equal in strength to the 1997-1998 event, may have been moderated by the eruption of Mexico’s El Chichón volcano, whose ash sent tons of climate-cooling pollutants into the stratosphere (see Geotimes, February 2004).

Better predictions of the effects of El Niño will initially come down to more detailed observations of our planet, whether from instruments on the ground or in space. We have a good understanding of how the climate will change over a season, but we don’t know for certain what extremes in precipitation and temperature can be expected during the season. This is an area that many researchers, including myself, are just beginning to tackle.

El Niño and Hurricanes?

While many people associate El Niño with increased storminess, the number of Atlantic hurricanes actually decreases during El Niño. The reason for this decline, according to the National Hurricane Center, has to do with the way El Niño changes the atmospheric circulation.

Tropical storms form over almost every ocean on Earth, but only the Atlantic storms are strongly affected by El Niño because of a change in the air flow over the Atlantic. During an El Niño, the eastern Pacific is warmer than normal, leading to rising air. When the air reaches the upper levels of the atmosphere, it spreads out, and some of that air moves northeastward over the Caribbean, Gulf of Mexico and Atlantic, mostly south of 30 degrees north latitude. This upper-air flow pattern is in a different direction than the surface winds at these locations, creating what is known as wind shear.

Hurricanes hate shear. One of the ingredients for hurricanes is tall, straight columns of clouds. Wind shear tilts the columns of clouds, inhibiting hurricane development. Thus, while the 2004 Atlantic hurricane season was particularly bad in August and September, the reason it was mostly quiet in October and November may have been due to the start of the current weak El Niño.


Back to top

Curtis is an assistant professor of geography at East Carolina University in Greenville, N.C., and a member of the university’s newly formed Atmospheric Science Laboratory. Prior to that, he worked at NASA Goddard Space Flight Center. E-mail:

"Volcanic forcing of El Niños," Geotimes, February 2004
"Wait and see for El Niño," Geotimes, May 2002

Back to top

Geotimes Home | AGI Home | Information Services | Geoscience Education | Public Policy | Programs | Publications | Careers

© 2018 American Geological Institute. All rights reserved. Any copying, redistribution or retransmission of any of the contents of this service without the express written consent of the American Geological Institute is expressly prohibited. For all electronic copyright requests, visit: