May 2015 ENSO Forecast: Will this El Niño be an overachiever, or peaked-in-high-school?

May 14, 2015

There’s a 90% chance that the current El Niño will continue through the summer, and forecasters estimate the chance that it will continue through the end of 2015 at greater than 80%. This is a pretty confident forecast. What are the forecasters looking at that gives them such confidence?

Warm and getting warmer

Sea surface temperatures in the equatorial Pacific remained substantially above average during April. Also, there is still a lot of warmer-than-average water below the surface in the upper 300 meters of the ocean, helping to ensure that the above-average sea surface temperatures will continue for at least the next few months.

SST anomalies May 06 2015

Sea surface temperature departures from average in degrees Celsius during the week of May 6, 2015. Figure from CPC.

The atmospheric response to the warmer waters in the equatorial Pacific that we began to see in February strengthened through March and April: the Trade winds are more westerly than average, upper-level winds are weaker than average, and more rain was present in the central equatorial Pacific. These are all signs of the weakened Walker Circulation, present during El Niño events, and act to reinforce El Niño events.

This consistent atmospheric coupling is a change from the pattern we saw throughout 2014, when conditions changed from week-to-week. With a climate phenomenon like the El Niño/Southern Oscillation (ENSO), we expect to see more of a persistent pattern. (This doesn’t mean that the El Niño atmospheric conditions are always present every week, but that they are on average over the season.)

Conditions look good, but what’s behind that confident forecast?

Nearly all computer model forecasts predict a continuation of the warmer-than-average sea surface temperatures in the Niño3.4 region through the end of 2015. There is a lot of agreement among the models that the SSTs will remain warm, and most models are forecasting SSTs more than 0.5°C above average through the October-December period. This agreement in forecasts, in combination with an ongoing consistent atmospheric response, is a large contributor to the high probability that this El Niño will continue through the fall.


NMME SST forecast

Probabilistic sea surface temperature forecast from the North American Multi-Model Ensemble, made early May 2015, for the October-November-December 2015 average. Forecast is expressed as percent likelihood for each of three categories: Above normal, below normal, and near normal. Figure from CPC.


As far as how strong this El Niño will ultimately be, it’s difficult to say. There is a fair amount of variation in the forecasts for the Niño3.4 region. The statistical models, which predict how current conditions are likely to change by applying statistics to historical conditions, are generally on the lower side of the forecast envelope, around 0.5°C – 1°C, the range of a “weak” event.

Many of the dynamical models, which use physical equations to predict how current conditions will evolve, have Niño3.4 Index forecasts of warmer than 1.5°C, our threshold for “strong” events. Based on past model runs, the bias-corrected NMME dynamical model forecasts made in early May for the October-December period have a typical error of about 0.6°C, meaning, according to this set of models, there’s a 2-in-3 chance that the average temperature in the Niño3.4 region will be in the range of 2.2°C +/- 0.6°C—i.e., between 1.6°C and 2.8°C. *

As Michelle discussed in her last post, models – both dynamical and statistical – tend to have a harder time making successful forecasts during the spring as well. Also, El Niño events typically peak in the early winter, which is still six months away. These factors combine to make it difficult to predict the peak strength of this El Niño. It’s likely that we’ll have a clearer picture of the potential strength in the next month or so. For reference, the potential strength of the strong 1997-1998 El Niño didn’t become apparent, and wasn’t formally mentioned by CPC, until July of 1997  (Barnston et al. 1999).

A special snowflake

Speaking of typical events, though – this is not one of them. As you can see below, it is unusual for sea surface temperatures in the Niño3.4 region to start off warm in the winter and then continue to be warm through the spring and summer. In the 60-year record, only one El Niño event, in 1986-1987, had similar behavior. The evolution and strength of this event might be a little easier to predict if it were starting at a more typical time of year.

historical El Nino

Sixteen El Nino events, 1952 - present, with 2014-2015 in red. Average is shown in black. 1997-98 and 1986-87 are yellow and blue, respectively. Figure by from CPC data.


OK, but what’s going to happen??

The most substantial US temperature and rain impacts from El Niño occur during winter. Right now, it’s too early to forecast with much confidence the effect this El Niño may have on the US next winter. (Although the good chance that this El Niño will last into winter does tilt the odds towards the expected temperature and precipitation impacts.)

The Australian Bureau of Meteorology has just raised their event tracker to “El Niño” status, the equivalent of issuing an El Niño Advisory for us. (They have slightly different thresholds for declaring the onset of El Niño.) El Niño events are linked to increased drought and heat waves in Australia, especially during their winter (our summer), so they monitor its evolution closely. If you’re interested in the Australian perspective on ENSO, check out this great post authored by Andrew Watkins of the BOM. 

Also, since it’s very likely that this event will continue through the summer, we may see some effects on the tropical cyclone seasons. The western Pacific tropical cyclone season is off to a roaring start, with seven named storms so far (the average is two!), which is likely linked to the warm Pacific waters. Keep your eye out for the NOAA hurricane forecast (issued this year on May 27) for potential effects of an ongoing El Niño.

Barnston, A. G., M. H. Glantz, and Y. He, 1999: Predictive skill of statistical and dynamical climate models in SST forecasts during the 1997/98 El Nino episode and the 1998 La Nina onset. Bull. Amer. Meteor. Soc., 80, 217-243.

*This is based on the Root-Mean-Square Error (RMSE), calculated from 32 years of past forecasts made in early May for the October-December average from the North American Multi-Model Ensemble