A mystery in the Pacific is complicating climate projections
Posted on 3 May 2023 by Guest Author
This is a re-post from Yale Climate Connections by Bob Henson
Nothing has a bigger influence on year-to-year variations in the global climate than the El Niño-Southern Oscillation, commonly called ENSO. And the tropical waters at the heart of ENSO aren’t behaving exactly as climate scientists expected they would in a warming world, with potentially major implications for Atlantic hurricane seasons, droughts in the U.S. Southwest and the Horn of Africa, and other weather phenomena around the world.
ENSO is a recurring ocean-and-atmosphere pattern that warms and cools the eastern tropical Pacific through El Niño and La Niña events that last from one to three years. Once El Niño or La Niña emerges, the odds reliably shift toward hotter, colder, wetter, or drier conditions for various parts of the globe, from Oceania to North America to Africa. But though ENSO’s effects are well known, the phenomenon itself is notoriously tough to predict. And its slippery nature is complicating crucial multi-decade projections of climate.
Many aspects of human-caused climate change are playing out as long predicted, including overall warming of the global atmosphere and oceans as well as the intensification of rainfall extremes and the drying of many subtropical areas.
Not so for ENSO. Top global climate models have predicted for more than 20 years that the tropical Pacific would gradually shift toward an “El Niño-like” state, with the surface waters warming more rapidly toward the east than toward the west.
Instead, just the opposite is going on. The western tropical Pacific has warmed dramatically, as predicted, but unusually persistent upwelling of cool subsurface water has led to a slight drop in average sea surface temperature over much of the eastern tropical Pacific.
Figure 1. Trend in sea surface temperatures across the equatorial Pacific Ocean from 1982 through 2022. Red (blue) shading in the map indicates trends toward more positive (negative) sea surface temperatures. (Image credit: NOAA/climate.gov, data from NCEI OISSTv2.1)
The result is a strengthening west-to-east temperature contrast that increasingly resembles La Niña. Scientists expect that El Niño events will continue to occur – such as the one predicted to arrive later this year – but they will take place on a backdrop of an ocean that looks more like La Niña.
All this is far more than an esoteric science matter. According to an ENSO Blog entry posted Jan. 26 at climate.gov: “How the sea surface temperature trend pattern will change has profound, worldwide implications. … If you are trying to make decisions based on projections of the future, you need to know the answer. And, at this moment, there is some significant (perhaps even growing) debate that surrounds it.”
Among the impacts that could be notably different in a La Niña-dominated world:
- Atlantic hurricane activity is substantially higher on average during La Niña than during El Niño.
- The U.S. Sun Belt, including much of California, tends to get less rainfall and mountain snow during La Niña, with widespread drought becoming more likely. (There are occasional exceptions, such as the very wet winter of 2022-23 in the Southwest.)
- The highly vulnerable Horn of Africa is more drought-prone during La Niña, while rains in the African Sahel tend to be more reliable.
- Typhoons are more likely to slam China, the Philippines, and Vietnam – and less likely to strike Guam, Japan, and Taiwan – during La Niña.
- Heavy rains and floods often plague eastern Australia during La Niña.
- The southwest monsoon in India is often wetter than average.
There are hints — though not yet enough cases to pass statistical muster — that La Niña events themselves are becoming more frequent, which would go hand in hand with the evolving Pacific backdrop. For example, since mid-2003, there have been 10 “years” (July to June) when La Niña conditions predominated, but only six for El Niño. Among years when the moderate-strength threshold was reached, seven were La Niña but only two were El Niño.
The La Niña-like temperature contrast in the tropical Pacific has now been strengthening for so long that it’s gotten tougher to ascribe to natural variability. It could be chalked up to a cool phase of the Pacific Decadal Oscillation, another recurring weather pattern that can span 20 to 30 years. Yet that phenomenon is largely influenced by ENSO itself.
Climate experts are thus ramping up efforts to diagnose what it is that the state-of-the-art models could be missing and whether the long-expected El Niño-like trend may yet turn up later in the century.
Figure 2. An aerial view of a property surrounded by flood water on Dec. 9, 2022, in Louth, Australia, about 400 miles northwest of Sydney. Flooding across vast swathes of southeastern Australia during the La Niña summer of 2022-23 turned properties into islands. The highest river levels in decades occurred at many flood-hit communities across the state. A few months earlier, flooding was even more widespread over eastern Australia, leading to one of the nation’s worst weather disasters on record and more than $4 billion in damage. (Image credit: Jenny Evans/Getty Images)
Above the La Niña-like Pacific, a strengthened Walker Circulation
At the core of ENSO is the Walker Circulation, the large-scale flow of air over the Pacific tropics. On average, air rises above the very warm waters of the western Pacific and sinks across the eastern Pacific, where chilly water predominates. Completing the Walker Circulation loop (as shown in Figure 3 below) are westward flow a few miles above the Pacific tropics and eastward flow – the famed trade winds – at the surface.
Figure 3. This schematic of average (neutral) conditions over the Pacific Ocean shows the Walker Circulation, with rising air and showers and thunderstorms to the west and descending air to the east. Trade winds (surface arrows) push warm water toward the west, allowing colder water to upwell toward the east. The depth of the upper Pacific Ocean is exaggerated in this schematic. (Image credit: climate.gov)
When the Walker Circulation is strong, as it is during La Niña, the east-to-west-blowing trade winds are boosted and there is more upwelling of cold water off Ecuador and Peru. When the Walker Circulation weakens, the trade winds slacken or even reverse, and the resulting westerly surface winds can push warm water all the way to the South American coast. The usual upwelling weakens, and an El Niño event is in place. The warm water leads to rising air thousands of miles farther east than usual, helping to suppress the Walker Circulation and torquing other weather patterns far away.
Given that El Niño is fighting against the grain of the usual atmospheric and oceanic conditions, it normally takes only about nine to 12 months before the tropics swing back toward a more typical Walker Circulation. In contrast, La Niña — with an extra-strong version of the Walker Circulation — can persist or recur for as long as three consecutive years, as it did from 2020-21 to 2022-23.
There’s no doubt that the tropical Pacific has behaved in more of a La Niña-like than El Niño-like fashion over the past several decades. The questions now center on whether the trend might still be natural variability — and if it’s not, what’s forcing it to happen.
Figure 4. The left column shows the relevant processes in the preindustrial climate and the right column shows how La Niña-like conditions could be reinforced. (This schematic does not show the final state). Technical Details: Left panel: Showers and thunderstorms are strongest in the western Pacific where sea surface temperature is the highest. Latent heating warms aloft, and evaporation cools the ocean surface. In the eastern Pacific, the thermocline is closest to the surface. Right panel: Under greenhouse-gas warming, a uniform surface heat flux into the Pacific Ocean causes the sea surface temperature to rise, but in the eastern tropical Pacific, the upwelling of cold water counters the forced warming. As a result, the zonal sea surface temperature gradient increases. (Image credit: climate.gov)
One of the first and most persistent among scientists looking into the perplexing sea surface temperature trend has been Richard Seager of Columbia University. “We first wrote a paper on this in 1997,” he said. Although less than 20 years of routine satellite data had been gathered from the tropical Pacific at that point, sea surface temperature data collected by ships extended back to the 1850s.
Since that 1997 study in Science, led by Seager’s colleague Mark Cane at Columbia, the sheer accumulation of satellite and buoy data on sea surface temperatures has supported the notion that the eastern tropical Pacific is cooling. A 2019 paper gave the ideas new prominence.
“We showed using all the available models just how robust the actual discrepancy is [between models and observations], and we also provided a simple model to show how this might be happening,” Seager said. “I think that moved the ball a little bit … I’m certainly very gratified [the idea] has gained traction now, including with funding agencies.”
Michelle L’Heureux tracks the variable nature of ENSO as closely as anyone. She’s the physical scientist who coordinates the monthly ENSO updates released by the NOAA/NWS Climate Prediction Center. L’Heureux noticed the acceleration of the Walker Circulation close to a decade ago. In 2022 she joined lead author Sukyoung Lee of Pennsylvania State University, together with Seager and others, for an overview paper that summarized the state of observations, simulations, and theories on where ENSO is heading.
“It’s exciting to see that others are now doing more work on the topic and coalescing around these ideas,” L’Heureux said.
At the National Center for Atmospheric Research, senior scientist Clara Deser, a longtime ENSO researcher, isn’t yet ready to dismiss natural variability as part of the picture. However, she adds: “There may be multiple causes for the La Nina-like cooling trend in the eastern tropical Pacific, not just a single one.”
The cold, remote Southern Ocean could be a big part of the Pacific’s puzzling trend. Much like the eastern tropical Pacific, the Southern Ocean is one of the few other areas on Earth where sea surface temperatures have cooled rather than warmed since the 1980s, in spite of pronounced surface warming over the Antarctic Peninsula and concerns about longer-term deep-ocean warming.
A number of studies have already found links between the Southern Ocean and ENSO behavior. Usually it’s the tropical Pacific that triggers weather and climate effects at middle and higher latitudes, but the Southern Ocean itself may be influencing ENSO, based on recent work that includes a 2022 study led by Columbia’s Yue Dong as well as a paper by Deser and colleagues that’s now in process.
Could longer-term warming catch up to La Niña?
Adding to the complexities at hand, it’s possible that whatever factors are now pushing the tropical Pacific toward a La Niña-like state will eventually get overwhelmed by longer-term global warming. That could lead to the El Niño-like outcome that models have long projected — and thus vindicate the models, at least in the long run.
Figure 5. Schematic showing a possible mechanism toward an El Niño-like state, as reviewed in Lee et al. (2022). The left column shows the relevant processes in the preindustrial climate and the right column shows how these same processes initially respond to greenhouse-gas warming (this schematic does not show the final state). Technical Details: Left panel: Convection is strongest in the western Pacific where sea-surface temperatures are the highest. Latent heating warms aloft, and evaporation cools the ocean surface. Right panel: Sea surface temperatures in the tropical Pacific increase leading to more condensational heating aloft and a higher tropopause, which increases dry static stability and gross moist stability, thus weakening the Walker circulation. At the same time, evaporative cooling and cloud shading are more sensitive to warming in the western Pacific, which weakens the zonal sea surface temperature gradient. (Image credit: climate.gov)
“There are some good physical reasons for why we would expect to eventually end up with eastern-Pacific warming,” said Ulla Heede, a postdoctoral researcher at the Cooperative Institute for Research in Environmental Sciences who’s published several recent papers on this two-stage concept.
In a subsequent study, drawing on a wide range of simulations carried out for the most recent Intergovernmental Panel on Climate Change assessment, Heede and Alexey Federov of Yale University found some but not all models depicting this two-phase response after an abrupt injection of carbon dioxide.
In one set of experiments with models that fully couple the atmosphere and ocean, Heede and colleagues found the observed La Niña-like state emerging and persisting for 20 to 100 years, depending on how quickly carbon dioxide is added. Somewhere between 50 and 100 years, the subtropical ocean circulation weakens and warms enough to cut down on cold upwelling, pushing the system toward El Niño-like conditions.
Learning from false alarms
The difficulty in modeling ENSO exists not only in century-scale climate simulations but also in the yearly-scale models that are run regularly to assess where ENSO is headed right now. This adds a hefty dose of challenge to the monthly ENSO outlooks issued by L’Heureux and colleagues at NOAA and the International Research Institute for Climate and Society.
Figure 6. The official Climate Prediction Center forecast for ENSO released on April 13, 2023, shows a very high likelihood (more than 80%) of El Niño conditions during the second half of 2022. (Image credit: NOAA/IRI)
There are good reasons why ENSO is so difficult to capture in both seasonal and multi-decadal modeling. For just one example, consider the thermocline, the oceanic boundary that separates warmer near-surface water from cooler deep water. Across the tropical Pacific, the thermocline depth normally varies from about 450 feet in the west to only about 50 feet in the east. That’s paper-thin when compared to the 12,000 miles of distance from west to east along the Pacific equator, and thus difficult for global climate models to capture with precision.
When a NOAA forecast of ENSO does “bust,” it’s usually because of a false alarm from the models, and it usually involves El Niño. As documented by Columbia’s Michael Tippett, the North American Multi-Model Ensemble projected that the eastern Pacific would warm between May and August in 2011, 2013, 2014, and 2017. All of these model forecasts were wrong. Only one of those four years, 2014, saw the Pacific make it to the El Niño threshold. That didn’t happen until autumn 2014, though, and the event was so borderline that it may end up being reclassified as neutral.
L’Heureux led a 2022 study on how the cooling trends in the eastern Pacific may be complicating ENSO forecasts; she’s also presenting a paper on the topic at the European Geophysical Union’s annual meeting in late April.
When climate-change guidance has to change course
An even bigger challenge than modeling ENSO may be communicating to stakeholders how they can prepare for ENSO-related trends that could be starkly different from what’s been long expected.
“East Africa has been struggling with drought for most of the 21st century, including the last few years,” Seager noted. “The models predict that that region should be getting wetter, and it hasn’t. The fact that it hasn’t is largely consistent with what the equatorial Pacific Ocean is doing.
“Food security in East Africa is a perennial problem. The implications are just enormous.”
Figure 7. Somali refugees return from collecting water at the edge of the Dagahaley refugee camp, which makes up part of the giant Dadaab refugee settlement, on July 22, 2011, in Dadaab, Kenya. Five consecutive biannual wet seasons failed during the three years of La Niña conditions from 2020 into early 2023, leading to the worst drought to affect the Horn of Africa in six decades. The drought, together with the ongoing civil war in Somalia, has resulted in an estimated 12 million people whose lives are threatened. (Image credit: Oli Scarff/Getty Images)
Isla Simpson, a colleague of Deser’s at the National Center for Atmospheric Research, is now co-organizing a workshop to examine where observational trends differ from climate models, including ENSO and other areas.
“We’re only really starting to be at the point where we’re able to observe these trends over a long enough period to know if the models are actually wrong,” Simpson said.
The U.S. Southwest will experience the landscape-parching effects of hotter weather no matter what ENSO does. But any continued tilt toward La Niña-like conditions would also tend to favor the prolonged deficits in rainfall and mountain snowfall that have plagued the region since 2000. Even the bountifully wet winter of 2022-23 is unlikely to turn the long-term tide. “The combined impact of warmer temps and potentially lower precipitation is not good news,” Simpson notes.
As summarized by Heede: “Climate scientists love studying what happens in the tropical Pacific, but my point of view is that no one’s living there. We have to make the link from what happens in the tropical Pacific to what happens over land.”
Jeff Masters contributed to this post. For more background on this topic, including a handy reference list, see the NOAA ENSO Blog post at climate.gov, “How the pattern of trends across the tropical Pacific Ocean is critical for understanding the future climate” (Jan. 26, 2023).