• Home
  • News
  • September Climate Briefing: Surprise La Niña or Ephemeral Cooling?

September Climate Briefing: Surprise La Niña or Ephemeral Cooling?

Authored by Elisabeth Gawthrop on

Read our ENSO Essentials & Impacts pages for more about El Niño and La Niña.

Tony Barnston provides an overview of the briefing

What’s New

Since last month’s briefing, sea-surface temperatures (SSTs) have continued to cool in the area of the central equatorial Pacific Ocean that define El Niño and La Niña events, called the Nino3.4 region, as well as to its east. The weekly SST anomalies in the last month have ranged from -0.6ºC to -0.2ºC; their average is on the borderline of a La Niña state, although conditions would have to persist for several months for a La Niña to be officially declared.

Atmospheric variables, including tradewinds (see tweet below) and convection patterns, show an increasingly possibility of La Niña, but are not strong enough to indicate certainty in an event developing. Sub-surface ocean temperatures, another potential signal of upcoming ENSO activity, are also somewhat indicative of La Niña. Barnston says it’s not clear what has caused these cooler temperatures, as they don’t seem to be stemming from some of the usual sources (e.g. upwelling waves and certain wind patterns).

The National Oceanic and Atmospheric Administration’s Climate Prediction Center’s ENSO alert system upgraded its status to a La Niña Watch.

ENSO Forecasts

To predict ENSO conditions, computers model the SSTs in the Nino3.4 region over the next several months. The plume graph below shows the outputs of these models, some of which use equations based on our physical understanding of the system (called dynamical models), and some of which use statistics, based on the long record of historical observations.

The SST anomalies predicted by the models in last month’s forecast were split fairly evenly below and above the 0.0ºC marker. This month, nearly all of the models fall in the negative range through the beginning of 2018. 

This graph shows forecasts made by dynamical and statistical models for SST in the Nino 3.4 region for nine overlapping 3-month periods. Note that the expected skills of the models, based on historical performance, are not equal to one another.
The IRI/CPC probabilistic ENSO forecast issued mid-September 2017. Note that bars indicate likelihood of El Niño occurring, not its potential strength. Unlike the official ENSO forecast issued at the beginning of each month, IRI and CPC issue this updated forecast based solely on model outputs. The official forecast, available at http://1.usa.gov/1j9gA8b, also incorporates human judgement.

The mean of the statistical models stays pretty close to 0º, and if the ENSO forecast were made only from those, we would still have a neutral forecast. But the dynamical models’ mean is cooler — dipping below the La Niña threshold of -0.5ºC through the end of 2017.  

Barnston says one reason the dynamical models may be showing a cooler prediction is because they’re picking up on new, cooler temperatures below the surface of the equatorial Pacific Ocean. Statistical models don’t “see” below the surface in the same way because they usually average over the most recent one month or even three months, so they are not sensitive to very recent rapid developments. More on these surprise, cooler anomalies from Barnston:

Based on the model outputs, La Niña conditions are the most likely ENSO outcome in the October – January timeframe, with odds topping out at just over 50%. Neutral conditions then take over again as the most likely for the rest of 2018 that is in the forecast period. 

ENSO in context: Resource page on climate variability

The official probabilistic forecast issued by CPC and IRI in early September indicates a similar overall outlook, but with La Niña odds reaching just over 60% and being the favored outcome for a bit longer into 2018. This early-September forecast uses human judgement in addition to model output, while the mid-month forecast relies solely on model output. More on the difference between these forecasts in this IRI Medium post.

IRI’s global seasonal forecasts

Each month, IRI issues seasonal climate forecasts for the entire globe. These forecasts take into account the latest model outputs and indicate which areas are more likely to see above- or below-normal temperatures and rainfall.

For the upcoming October-December season, odds are tipped in favor of above-normal rainfall across much of Indonesia, as well as northeastern Russia, Central America and northern South America. Some parts of East Africa show a slight favoring of above-average precipitation — likely due to the influence of the positive Indian Ocean Dipole.

The south-central and southeastern United States, northern Mexico and much of the Middle East show the strongest, most widespread probabilities for drier-than-normal conditions. Northern Ethiopia, Uruguay, and other scattered areas of South America also show some increased chance for drier-than-normal conditions. Seasons later in the year and into 2018 show even stronger climate signals in some areas. All forecast maps, including temperature in addition to precipitation, are available on our seasonal forecast page.

As of April of this year, these seasonal forecasts use a new methodology. The IRI probabilistic seasonal climate forecast product is now based on a re-calibration of model output from the NOAA’s North American Multi-Model Ensemble Project (NMME). The output from each NMME model is re-calibrated prior to multi-model ensembling to form reliable probability forecasts. The forecasts are now presented on a 1-degree latitude-longitude grid. More on this change on the seasonal forecast page, as well as in this Q&A with three of our climate scientists.

Learn more about El Niño and La Niña on our ENSO resources page, and sign up here to get notified when the next forecast is issued. In the meantime, check out #IRIforecast.