California grows more than 90 percent of the tomatoes, broccoli and almonds consumed in the U.S., as well as many other foods. These crops require a lot of water. In the spring of 2015, after four years of little winter rain, the state was in a severe drought. Reservoirs were far below capacity, and underground aquifers were being heavily tapped. Mountain snowpack, an important source of meltwater throughout the spring and summer, was nearly gone in many areas.

Not surprisingly, then, when the National Oceanic and Atmospheric Administration announced that an El Niño climate pattern was setting up over the Pacific Ocean, California farmers and their neighbors took note. Conventional wisdom said that El Niño brings plentiful rains to the Golden State.

El Niño is the warm half of a cycle of warming and cooling in the tropical Pacific Ocean's surface waters. The cycle recurs about every three to seven years; the cool half is called La Niña. When either phenomenon arises, it generally prevails for six months to a year. During El Niño, the warm waters heat the air above them, causing changes to the atmospheric circulation that affect the entire world. NOAA, where I conduct climate research, can usually see an El Niño or La Niña coming in advance of when it will have its strongest influence on global weather.

Californians' hopes were high, and yet the effects that usually occur with an El Niño there and elsewhere do not happen all the time. During the 20 El Niño seasons since NOAA began tracking them in 1950, only about half brought above-average precipitation to California during its rainy season: December, January and February. In some cases, the effects are the opposite of what is expected. Forecasters have become good at predicting a developing El Niño or La Niña, but they still struggle with predicting the regional weather changes that might result.

In early 2015, as California dried out further, forecasters faced several burning questions: Would the coming El Niño be a big one? Would it save California? For that matter, would it amp up hurricanes in the Pacific and reduce hurricanes in the Atlantic, bake Australia, fuel forest fires in Indonesia, or make the upcoming winter disappear in the Northeast, as some El Niños had done in the past? And could we know ahead of time?

Being able to answer such questions would greatly help farmers, forecasters, emergency planners and the general public prepare for extreme weather, and investigators are trying hard to pin down the data that are needed. Yet as the tale of how the most recent, extreme El Niño unfolded demonstrates, the science is tricky.

March 2015: El Niño Is Here

Early signs of a developing El Niño occur under the ocean surface. Winds across the tropical Pacific typically blow from east to west—the trade winds that reliably have carried sailing ships across the great ocean. These winds keep the surface water in the eastern and central Pacific slightly cooler and pile up warm water in the western side, toward Indonesia. Occasionally these winds can weaken, allowing slow waves of warmer western waters to begin to travel back eastward along the equator toward South America over many months. That can kick off an El Niño or feed one that has already begun.

To me and other meteorologists, it looked like the developing 2015 El Niño was going to be a big one. Over the past several months we had seen sea-surface temperatures that were warmer than average in the tropical Pacific, including the Niño 3.4 region in the central Pacific, which we track as a leading indicator. El Niño, though, is part of a phenomenon that couples changes in the ocean to changes in the atmosphere—the El Niño/Southern Oscillation—so we were also monitoring the atmosphere for signs that it was responding to those increased ocean temperatures.

Water even just a few degrees higher than usual holds a tremendous amount of heat, which warms the air above the ocean, coupling the changes in the ocean to the atmosphere. During El Niño, the warmer central-eastern Pacific takes over as the engine affecting an atmospheric pattern called the Walker Circulation. With a strong source of rising moist air now much farther east, the surface winds weaken, sometimes reversing altogether and blowing west to east. This atmospheric reaction is the Southern Oscillation, and it is essential to El Niño, helping it sustain and strengthen itself.

Click or tap to enlarge

Credit: George Resteck (illustration), Jen Christiansen (map); SOURCE: NATIONAL OCEANIC AND ATMOSPHERIC ADMINISTRATION (map)

In March the effects of the warmer tropical Pacific had taken hold. The Walker Circulation was weakening. We also saw bursts of westerly wind over the tropical Pacific, which can encourage warmer surface waters to move eastward. Heat deeper in the Pacific Ocean was high, too, which could help extend the atmospheric coupling. After 12 months of watching, NOAA issued an El Niño Advisory. Game on.

May 2015: Probabilities Rise

By May, NOAA had determined there was a 90 percent chance that the current El Niño would continue through the summer and an 80 percent chance that it would continue through the end of 2015. The agency was confident in its prediction because sea-surface temperatures in the equatorial Pacific remained substantially above average during April. The same was true for water below the surface, in the upper 300 meters of the ocean. The atmospheric response had strengthened, too.

But what weather effects might we see? And what would happen in California? El Niño typically exerts its strongest impacts on temperature and rain in the early winter, which was still six months away. Some signs suggested that this El Niño would be like other strong ones in the past. Drought and heat waves in Australia were rearing up (autumn was giving way to winter there). And the western Pacific cyclone season was off to a roaring start, with seven named storms by May; the average is two.

July 2015: Full Swing

As July commenced, nearly all the computer models were in agreement, and the ocean and atmosphere continued to behave according to plan. El Niño was well established, and forecasters were convinced that it would become very strong. The three-month-average sea-surface temperature in the Niño 3.4 region was expected to peak near an all-time high, matching that of the two previous, strongest El Niños on record: 1997–1998 and 1982–1983.

Southern Californians who remembered the 1997–1998 winter anticipated pounding storms and surf. During the 1982–1983 and 1997–1998 El Niños, winter conditions shifted the Pacific subtropical jet stream—a band of eastward-flowing air high in the atmosphere above the U.S. that often influences weather—south toward southern California. Powerful storms, fed with moisture from the warmer waters, provided heavy, reservoir-filling rains—as well as landslides along a soggy coast.

Farmers and residents hoped that the new El Niño would deliver plentiful precipitation by December. Forecasters gave a 60 percent chance that during the upcoming 2015–2016 winter, regions of southern California, as well as the Gulf States, would see rain amounts in the upper third of the historical record. This forecast was derived in part by monitoring several different signals, including El Niño, and comparing them with past trends to see if the odds of a certain outcome might be shifting.

October 2015: Unexpected Winds

In October, hope for California was high. We were closing in on the peak of the 2015–2016 El Niño, and it still ranked among the strongest in our records. Yet we were seeing something unexpected. The surface winds along the Pacific equator, important for maintaining the high sea-surface temperatures, had not weakened as much as they had during past strong El Niños. In 1997–1998 the winds weakened so much they reversed, blowing from west to east during October and November, moving even more warm water from the far western Pacific into the central Pacific and feeding the El Niño.

We humans have a tendency to expect that the outcome of a set of circumstances will always be the same, but variability happens in nature all the time. In coastal northern California, a strong El Niño year averages about 40 rainy days per winter, compared with about 26 during a non–El Niño winter. Yet the winter of 1965, one of the six strongest El Niños, had fewer rainy days than the non–El Niño average. In these times of global warming, we also had to wonder whether that was playing a role, too. If it was, prediction of El Niño's effects would become that much more difficult.

January 2016: A Lot Going On

By January, El Niño had put up some impressive numbers. In December the Niño 3.4 index broke the record for that month at 2.32 degrees Celsius above average, surpassing the 2.24 degrees C of December 1997. El Niño is ultimately measured on seasonal timescales, though, so the average of the sea-surface temperature anomaly (the departure from the long-term average) over three months is what we really pay attention to. From October to December 2015, the anomaly was 2.3 degrees C, tied for first place with 1997.

Outside of California, the effects of El Niño were mostly occurring as expected. Much more rain than is typical fell in eastern Africa during the “short rains” season (October through December). Southern Africa had continued dry conditions. Uruguay, southern Brazil and Paraguay experienced a lot of rain, and northern South America had been dry.

Australia's typical El Niño impact is dryness over most of the continent from about July to December, but in 2015 there had not been a clear rainfall deficit, except in parts of eastern Australia. It is possible that a record warm Indian Ocean had a strong countereffect—a reminder that the climate system has many moving parts, so expected impacts from El Niño are not guaranteed.

Closer to home, the Northeast was very warm, as anticipated. Michelle L'Heureux, my fellow meteorologist at NOAA, wrote in her blog: “For the first time ever my extended family did our Christmas gift exchange outside on my aunt's patio in the Washington, D.C. area…. We abandoned hot cider in favor of tropical beverages. Some of us wore t-shirts and sandals. We played catch with the dogs.”

As in Australia, El Niño was not the only cause of the unusual weather in Washington. An atmospheric pattern known as the Arctic Oscillation—epitomized by the “polar vortex” of winds that circle the Arctic—had entered a strong state. Unlike some recent winters, when the vortex was weak and allowed cold air to pour down into the U.S., during December 2015 it had been strong, trapping the frigid air way up north and allowing warm air from the southern U.S. to drift northward.

Several other non–El Niño oceanic and atmospheric phenomena could also have influenced weather in one part of the globe or another. One is the Madden-Julian Oscillation (MJO), an area of storminess that circles the equator, traveling eastward and lasting for weeks. It can temporarily enhance the effects of El Niño but can also reduce them. As I wrote in January, “Clearly, the question of how the MJO and El Niño act to reinforce or weaken each other is still up for debate.”

Then there is the Pacific Decadal Oscillation, a relation between the surface temperatures of the eastern and western North Pacific that often prevails for 15 years or more before switching to a different state. All these patterns affect one another. And of course, climate change is a wild card that could influence any of these patterns in still unpredictable ways.

Click or tap to enlarge

Credit: Jen Christiansen; SOURCES: NOAA (temperature and precipitation data); Philip Klotzbach (hurricane data)

L'Heureux's warm Christmas, and the extremely warm November and December across eastern North America, seemed to stem from a combination of El Niño, the bottled-up cold air near the Arctic and an active Madden-Julian Oscillation, plus a large component that cannot be explained even by those factors. Despite screaming headlines in daily newspapers and bold declarations by television weather forecasters, all saying El Niño was causing the extreme weather, it is not possible to point to a single storm, or cold snap, or heat wave and say, “That's El Niño.” El Niño influences the background state, so a collection of weather events can be partially attributed to El Niño but not a single incident.

In California hopes were quickly disappearing that El Niño would help ease the drought. The rain in December and early January had been above average in northern California but around average in the southern half of the state. A brief string of storms coming from the Pacific finally arrived in the last three days of January, delivered by a so-called atmospheric river of moisture that heads directly toward the U.S. West Coast from the tropics. It dropped a fair amount of snow across the Sierra Nevada Mountains, which might, it was thought, be beneficial when it melted later in the spring.

“Of course, one wet month isn't going to erase California's drought,” wrote another of my NOAA colleagues, Tom Di Liberto, in his own climate blog. “While some interior portions of California as well as northern California have recorded above-average precipitation, areas to the south, including the heavily populated coastal corridor stretching from Santa Barbara to San Diego, have seen precipitation less than 75% of normal.” Rainstorms, he concluded, will “need to occur more often to get this year back to normal,” let alone overcome reservoir and aquifer shortfalls that have accumulated since 2011.

Even a strong El Niño is no guarantee of abundant rainfall for California. It just tilts the odds in favor of a wetter-than-average winter there.

After the warm Christmas in the mid-Atlantic, snow hit there in January, shutting down Capitol Hill for two days. Weird weather once more. The public was ripe for hype about El Niño, but again, attributing a single storm to one climate influence, especially such a complicated nor'easter, just cannot be done. Although at least six of the top-10 snowstorms on record in Washington, D.C., have occurred during El Niño conditions, a lot of components had to come together to create the 2016 blizzard. They included a cold snap, warm Atlantic waters to feed moisture to the storm and a strong frontal system. El Niño's fingerprint may have been present in some of those factors, but it is tough to separate out. During El Niño, the subtropical jet stream does tend to steer storms across the Gulf States, as well as Georgia and the Carolinas, but they typically exit to the Atlantic south of Maryland and Virginia. Yet during some storms the jet stream bends more northward to D.C., which is somewhat unusual for El Niño conditions but not unheard of.

March 2016: The Big Three

We had seen enough by March to address the widespread claims in popular media that the U.S. and the world were experiencing the strongest El Niño ever. We had a complete data set for 2015–2016 and could compare it with data for the other two biggest El Niños: 1982–1983 and 1997–1998. It was definitely one of the strongest three going back to 1950, but ranking El Niños is difficult because strength can be measured in different ways.

The primary number we use at NOAA is the Oceanic Niño Index, which shows how the three-month-average sea-surface temperature in the Niño 3.4 region departs from the long-term average temperature. That value for November 2015 to January 2016 was 2.3 degrees C, tied with 1997–1998. We watch other areas of the oceans, too, including the eastern Pacific (warmer in 1997–1998) and the western Pacific (warmer in 2015–2016). And we consider the second part of the El Niño/Southern Oscillation, which is the all-important atmospheric response to ocean temperatures. Overall, the atmospheric response during 1997–1998 was stronger than during 2015–2016.

As for the winter weather and climate, a number of similar outcomes had arisen in all three cases, but there were also some notable exceptions. Among them: Northern California took most of that state's rain instead of southern California. Determining why this pattern was different from past strong El Niño winters will keep climate researchers busy for years to come. Many potential components are at play, including that the world's oceans today are much warmer than they were during the past two strong El Niños. And short-term chaotic effects are always present in weather systems, which ensures that even if one El Niño looks identical to another, its effect on the weather will not necessarily be the same.

Enter La Niña?

A year after NOAA first announced El Niño conditions, the episode was ending. Sea-surface temperature anomalies across most of the equatorial Pacific had decreased in February 2016, and the large amount of water below the surface that was warmer than average declined, too.

In March it was also looking likely that conditions would transition to neutral by early summer, with about a 50 percent chance that La Niña would set in by the fall. La Niña conditions have followed six of the 10 moderate and strong El Niños since 1950, but that sample size is too small for a confident forecast.

Our computer climate models also have a difficult time making accurate forecasts in spring (specifically, March through May), when El Niño and La Niña are often weakening, changing into neutral. Weather over North America during this transition season is highly variable, which often overwhelms signals from either phenomenon.

Nevertheless, most models agreed that La Niña would develop by the fall of this year. In April, NOAA issued a La Niña Watch, and by August we had determined that there was a 50 percent chance that La Niña would be in place by the fall. Computers, crunching the latest data, found that sea-surface temperatures in the Pacific would continue to drop, potentially passing the La Niña threshold (0.5 degree C below average). Also, cooler than average water was accumulating under the surface across the entire Pacific, along the equator. It is interesting to note that subsurface water in the Niño 3.4 zone reached record-cold temperatures in 1998—immediately following the strong El Niño of 1997–1998.

In the U.S., La Niña impacts are roughly the opposite of El Niño's but not precisely. La Niña alters westerly winds and the jet stream in its own ways. Among other things, it tends to create a hospitable environment for hurricanes to develop in the Atlantic. NOAA's official Hurricane Season Outlook update, which was released in August, forecast a typical to above-average North Atlantic hurricane season because other factors were at play. La Niña winters in California tend to be dry.


Despite El Niño's fearsome reputation, it does not usually trigger a greater number of weather-related disasters worldwide than other years do, but the disasters are more predictable. If government and emergency planners take heed of El Niño–related seasonal forecasts, it may be possible to move resources to certain places, ahead of disasters, to reduce the human impacts should a given El Niño produce its typical effects.

Researchers are still not clear on how global warming will interact with the El Niño/La Niña cycle. Some work suggests that warmer overall oceans will lead to more powerful El Niños, but other studies say that global warming could actually reduce their strength. Because the entire global system includes large atmospheric and oceanic phenomena that naturally flip between different states, such as the Pacific Decadal Oscillation, figuring out how El Niño has been affected by global warming so far, and how it could be in the future, has been a challenge. Researchers will continue to try to decipher possible connections.

One thing we can say for near certain is that El Niños and La Niñas will continue to occur, some stronger than others. Evidence from fossilized corals tells us that the cycle has been recurring for thousands of years. As we better understand climate system dynamics, we can better anticipate the effects of this important pattern on global weather and on people around the world.