A recent paper in Reviews of Geophysics describes new scientific insights on the El Niño-Southern Oscillation from analysis of the 2015-2016 extreme El Niño
January 10, 2018 — The El Niño-Southern Oscillation (ENSO) cycle is a periodic change in the ocean-atmosphere system of the tropical Pacific Ocean but which has knock-on effects on weather around
the world. In an article recently published in Reviews in Geophysics, Santoso et al.  examined the 2015-2016 El Niño event which was particularly extreme. The editor asked one of the authors to explain the characteristics of that El Niño and what scientists learned from it.
Why was the El Niño event of 2015-2016 particularly significant?
The 2015-2016 event was a particularly strong El Niño, comparable to the extreme 1997-1998 El Niño which was dubbed “the climate event of the 20th century”. It developed rapidly and had spectacular climate impacts worldwide. It also followed on the heels of new research that found the frequency of extreme El Niños will likely double in the 21st century if greenhouse gas emissions continue unabated.
Another reason why the event was significant is that we previously had observed only two extreme El Niños with any sort of detail in our short instrumental record: the 1997-1998 and the 1982-1983 events. These two El Niños shared similar key characteristics with each other. Thus, observing another extreme event was very valuable for testing the current state of our understanding about extreme El Niños. Nature did not cooperate completely though, since it turned out that the 2015-2016 El Niño had some important difference from its extreme predecessors. It was a monster in its own special way.
How were the characteristics of that El Niño different from ones observed previously?
Tropical Pacific sea surface temperature (SST) anomalies (in other words, SST deviations from normal) are the most commonly used metric to compare El Niño events.
For the 2015-2016 El Niño, the SST anomaly peaked toward the central equatorial Pacific, whereas the 1982-1983 and 1997-1998 events peaked closer to the South American coast. Consequently, the 2015-2016 El Niño exhibited anomalously higher rainfall in the central Pacific than in the eastern Pacific, which is opposite to the 1982-1983 and 1997-1998 patterns.
In fact, SSTs and rainfall in the central Pacific broke records in 2015-2016. Also, even though the 2015-2016 eastern Pacific rainfall was not as high as during the 1982-1983 and 1997-1998 events, it was still substantially higher than in other garden-variety El Niño events. Another interesting feature is that the 2015-2016 event did not show a clear west to east migration of SST anomalies along the equator, which is a unique characteristic of the previous extreme El Niños.
What new insights did the 2015-2016 events give into the nature of extreme El Niño?
We generally classify ENSO events into Eastern-Pacific and Central-Pacific types (based on the peak region of the SST anomalies). According to this classification, the 1982-1983 and 1997-1998 events were strongly of the Eastern Pacific type. All La Niñas and moderate El Niños tend to be classified as Central Pacific events. The 2015-2016 El Niño was somewhat a mixture of the two. It was similar to the strong 1972-1973 El Niño but more intense in many ways.
This highlights that El Niños are not all made from the same cookie cutter—they exhibit a wide range of behaviours between the scorching intensity of the 1982-1983 and 1997-1998 events to weaker and more moderate events. The latest event also provided a glimpse into how background ocean warming trends, which have become more prominent since the last extreme El Niño in 1997-1998, can affect the development and impacts of El Niño.
What features determine whether an El Niño event should be classified as extreme?
All El Niño and La Niña events involve significant alterations in ocean and atmospheric circulation, and we track these changes using a variety of measures. Generally speaking, though, an extreme El Niño pushes the envelope of what we consider to be unusual. This means that westward blowing trade winds and the equatorial surface currents slow to a complete stop and even reverse. Also, rain bands normally located north and south of the equator merge on the equator and expand into the eastern Pacific.
Such disruptions involve changes in several other processes, such as poleward discharge towards higher latitudes of upper ocean heat content following the peak of El Niño. This discharge process cools the equatorial Pacific and preconditions the system for a La Niña. That is why a La Niña is more likely to happen following an extreme El Niño. The heat discharge process of the 1997-1998 extreme El Niño for example was strong enough to cause back-to-back extreme La Niñas as illustrated by this animation.
What are some of the remaining challenges of predicting, monitoring, and projecting future El Niño events?
Our knowledge of extreme El Niño events is very limited because we have had so few examples to study from the short observational record of the past 35 years when oceanic and atmospheric measurements were sufficiently abundant. We can reconstruct past records using paleo proxies from corals, tree rings, and lake sediments in El Niño affected regions. But while this approach provides useful information on the overall El Niño statistics in the distant past, there are still large uncertainties in the character of individual events.
We use climate models to guide the interpretation of paleo reconstructions and vice versa. And we use models to make seasonal forecasts and future projections. However, while the current generation of climate models can provide useful insights into the mechanisms behind future projections, they still need to be significantly improved to narrow uncertainties in future projected El Niño behaviour. It is critical therefore that we sustain our observing networks in the ocean and atmosphere as these are used to benchmark and initialize models for monitoring and for making future projections.
Are we likely to experience more extreme El Niño events in future?
Unquestionably. And be sure that nature will surprise us when they occur because our expectations are based on so few examples. We continue to study this compelling problem using that subset of climate models that are able to simulate extreme El Niño events like we have observed in the past. Our recent studies [Cai et al., 2015] using these models tell us that the frequency of extreme El Niño will double under business-as-usual greenhouse gas emission scenarios by the end of the 21st century. Even under the Paris Agreement scenario of limiting greenhouse gas warming to 1.5°C globally, the frequency is still projected to increase significantly and to continue to increase for a century after global warming stops [Wang et al., 2017]. On the other hand, we found the risk of extreme La Niña to increase under a business as usual scenario, but to taper off if the Paris Agreement goals can be met.