ENSO Forecast Navigation

Download the Quicklook PDF

ENSO Forecasts

IRI ENSO Forecast

2016 May Quick Look

Published: May 19, 2016

A monthly summary of the status of El Niño, La Niña, and the Southern Oscillation, or ENSO, based on the NINO3.4 index (120-170W, 5S-5N)

Use the navigation menu on the right to navigate to the different forecast sections

During mid-May 2016 the positive tropical Pacific SST anomaly was quickly weakening, now indicating only a weak El Niño. The atmospheric variables continue to support the El Niño pattern, but at much reduced strength. This includes only a mildly weakened Walker circulation and excess rainfall in the central tropical Pacific, failing to extend eastward as it did in previous months. Most ENSO prediction models indicate a return to neutral by the end of May, with likely development of La Niña (of unknown strength) by fall.

Historically Speaking

    El Niño and La Niña events tend to develop during the period Apr-Jun and they
  • Tend to reach their maximum strength during Dec-Feb
  • Typically persist for 9-12 months, though occasionally persisting for up to 2 years
  • Typically recur every 2 to 7 years

Figure 1 is based on a consensus of CPC and IRI forecasters, in association with the official CPC/IRI ENSO Diagnostic Discussion

Figure 3 is purely objective, based on regression, using equally weighted model predictions from the plume

IRI ENSO Forecast

CPC/IRI ENSO Update

Published: May 12, 2016

El Niño/Southern Oscillation (ENSO) Diagnostic Discussion issued jointly by the Climate Prediction Center/NCEP/NWS and the International Research Institute for Climate and Society

ENSO Alert System Status: El Niño Advisory/La Niña Watch

Synopsis: La Niña is favored to develop during the Northern Hemisphere summer 2016, with about a 75% chance of La Nina during the fall and winter 2016-17.

During the past month, sea surface temperature (SST) anomalies decreased across the equatorial Pacific Ocean, with near-to-below average SSTs recently emerging in the eastern Pacific (Fig. 1). The latest Niño region indices also reflect this decline, with the steepest decreases occurring in the Niño-3 and Niño-1+2 regions (Fig. 2). The surface cooling was largely driven by the expansion of below-average subsurface temperatures, which extended to the surface in the eastern Pacific (Fig. 3 and Fig. 4). While oceanic anomalies are clearly trending toward ENSO-neutral, many atmospheric anomalies were still consistent with El Niño, such as the negative equatorial and traditional Southern Oscillation indices. Upper-level easterly winds persisted over the central and eastern Pacific, while low-level winds were near average. Enhanced convection continued over the central tropical Pacific and was suppressed north of Indonesia (Fig. 5). Collectively, these anomalies reflect a weakening El Niño and a trend toward ENSO-neutral conditions.

Most models predict the end of El Niño and a brief period of ENSO-neutral by early Northern Hemisphere summer (Fig. 6). The model consensus then calls for increasingly negative SST anomalies in the Niño 3.4 region as the summer and fall progress. However, there is clear uncertainty over the timing and intensity of a potential La Niña (3-month Niño-3.4 SST less than or equal to -0.5°C). The forecaster consensus favors La Niña onset during the summer, mainly weighting the dynamical models (such as NCEP CFSv2) and observed trends toward cooler-than-average conditions. Overall, La Niña is favored to develop during the Northern Hemisphere summer 2016, with about a 75% chance of La Nina during the fall and winter 2016-17 (click CPC/IRI consensus forecast for the chance of each outcome for each 3-month period).

This discussion is a consolidated effort of the National Oceanic and Atmospheric Administration (NOAA), NOAA’s National Weather Service, and their funded institutions. Oceanic and atmospheric conditions are updated weekly on the Climate Prediction Center web site (El Niño/La Niña Current Conditions and Expert Discussions). Forecasts are also updated monthly in the Forecast Forum section of CPC’s Climate Diagnostics Bulletin. Additional perspectives and analysis are also available in an ENSO blog.

The next ENSO Diagnostics Discussion is scheduled for 9 Jun 2016. To receive an e-mail notification when the monthly ENSO Diagnostic Discussions are released, please send an e-mail message to: ncep.list.enso-update@noaa.gov.

Climate Prediction Center
National Centers for Environmental Prediction
NOAA/National Weather Service
College Park, MD 20740

CPC/IRI Early-Month Official ENSO Forecast Probabilities

Season La Niña Neutral El Niño
AMJ 2016 3% 35% 62%
MJJ 2016 26% 58% 16%
JJA 2016 52% 42% 6%
JAS 2016 65% 31% 4%
ASO 2016 71% 26% 3%
SON 2016 75% 22% 3%
OND 2016 76% 21% 3%
NDJ 2016 76% 21% 3%
DJF 2017 76% 21% 3%

IRI ENSO Forecast

IRI Technical ENSO Update

Published: May 19, 2016

Note: The SST anomalies cited below refer to the OISSTv2 SST data set, and not ERSSTv4. OISSTv2 is often used for real-time analysis and model initialization, while ERSSTv4 is used for retrospective official ENSO diagnosis because it is more homogeneous over time, allowing for more accurate comparisons among ENSO events that are years apart. During ENSO events, OISSTv2 usually shows stronger anomalies than ERSSTv4, and during very strong events the two datasets may differ by as much as 0.5 C.

Recent and Current Conditions

Now in the very tail end of the very strong 2015-16 El Niño, the latest weekly SST anomaly in the NINO3.4 region is barely above 0.5 C during mid-May. For April 2016 the average NINO3.4 SST anomaly was 1.09 C, indicative of moderate El Niño conditions, and for Feb-Apr it was 1.72 C, in the strong El Niño range. The IRI’s definition of El Niño, like NOAA/Climate Prediction Center’s, requires that the SST anomaly in the Nino3.4 region (5S-5N; 170W-120W) exceed 0.5 C. Similarly, for La Niña, the anomaly must be -0.5 C or less. The climatological probabilities for La Niña, neutral, and El Niño conditions vary seasonally, and are shown in a table at the bottom of this page for each 3-month season. The most recent weekly SST anomaly in the Nino3.4 region was 0.6 C, in the category of weak El Niño. Accompanying this SST has been a continuing but weakening El Niño-like atmospheric pattern, including easterly upper-level wind anomalies (but nearly no low-level wind anomalies) and positive anomalies of convection limited to near the dateline. The Southern Oscillation Index (SOI) and the equatorial SOI have also been negative, but less strongly so than earlier in the year, indicating weakening El Niño conditions.

Expected Conditions

What is the outlook for the ENSO status going forward? The most recent official diagnosis and outlook was issued one week ago in the NOAA/Climate Prediction Center ENSO Diagnostic Discussion, produced jointly by CPC and IRI; it called for this strong El Niño to return to neutral by late spring or early summer 2016, with La Niña development quite possible during summer and around 70-75% likely for the fall and winter 2016-17. The latest set of model ENSO predictions, from mid-May, now available in the IRI/CPC ENSO prediction plume, is discussed below. Currently, while the most recent Nino3.4 SST anomalies are still in the weak El Niño category, subsurface temperature anomalies across the eastern equatorial Pacific are now below average in the central and eastern tropical Pacific. With these below-average subsurface temperatures, the SST is poised to fall into the neutral range for June and early July, and then likely below average in the months of late summer through the remainder of the year.  In the atmosphere, the basin-wide sea level pressure anomaly pattern (e.g. the SOI) has been at moderate El Niño levels, with some fairly large week-to-week variations.  Anomalous convection (as measured by OLR) has been above average near the dateline, but more weakly than seen in the previous months, and now lacking the portion extending to the east of the dateline. Together, the oceanic and atmospheric features reflect continuing but much weakening El Niño conditions for late April through mid-May. By early June, the Niño3.4 anomaly will likely be below 0.5 C, in the ENSO-neutral category, and during July it could decrease to below -0.5 C, in the weak La Niño range.

As of mid-may, 19% of the dynamical or statistical models models predicts El Niño SST conditions, 62% predicts neutral conditions, and 19% predicts La Niña conditions for the initial May-Jul 2016 season. At lead times of 3 or more months into the future, statistical and dynamical models that incorporate information about the ocean’s observed subsurface thermal structure generally exhibit higher predictive skill than those that do not. For the Aug-Oct 2016 season, among models that do use subsurface temperature information, 33% predicts ENSO-neutral conditions, 62% predicts La Niña conditions, and 5% predicts El Niño conditions. For all model types, the probabilities for El Niño are 19% for May-Jul, and 5% or less for all of the seasons from Jul-Sep 2015 through Jan-Mar 2017.  Chances for La Niña conditions rise to 50% for Jul-Sep and Aug-Oct, and to at least 60% for Oct-Dec and Nov-Jan. Chances for ENSO-neutral are below 50% for most of the forecast periods following May-Jul, and reach a minimum in the 30-35% range during Oct-Dec and Nov-Jan 2016-17.

Note  – Only models that produce a new ENSO prediction every month are included in the above statement.

Caution is advised in interpreting the distribution of model predictions as the actual probabilities. At longer leads, the skill of the models degrades, and skill uncertainty must be convolved with the uncertainties from initial conditions and differing model physics, leading to more climatological probabilities in the long-lead ENSO Outlook than might be suggested by the suite of models. Furthermore, the expected skill of one model versus another has not been established using uniform validation procedures, which may cause a difference in the true probability distribution from that taken verbatim from the raw model predictions.

An alternative way to assess the probabilities of the three possible ENSO conditions is more quantitatively precise and less vulnerable to sampling errors than the categorical tallying method used above. This alternative method uses the mean of the predictions of all models on the plume, equally weighted, and constructs a standard error function centered on that mean. The standard error is Gaussian in shape, and has its width determined by an estimate of overall expected model skill for the season of the year and the lead time. Higher skill results in a relatively narrower error distribution, while low skill results in an error distribution with width approaching that of the historical observed distribution. This method shows probabilities for La Niña at 43% for Jun-Aug 2016, rising to near 60% from Aug-Oct through Jan-Mar 2017. Model probabilities for neutral ENSO conditions are 91% for May-Jul, falling to below 50% from Jul-Sep through Jan-Mar 2017, with lowest values near 30% from Oct-Dec to Dec-Feb 2016-17. Probabilities for El Niño are 10% or lower throughout all forecast periods.  A plot of the probabilities generated from this most recent IRI/CPC ENSO prediction plume using the multi-model mean and the Gaussian standard error method summarizes the model consensus out to about 10 months into the future. The same cautions mentioned above for the distributional count of model predictions apply to this Gaussian standard error method of inferring probabilities, due to differing model biases and skills. In particular, this approach considers only the mean of the predictions, and not the total range across the models, nor the ensemble range within individual models.

In summary, the probabilities derived from the models on the IRI/CPC plume describe, on average, increasing likelihood for La Niña conditions from the initially low levels of the initial May-Jul 2016 season. La Nina development is slightly more likely than not by late northern summer 2016, and most likely later in 2016: that is, at least 60% likely. A caution regarding this latest set of model-based ENSO plume predictions, is that factors such as known specific model biases and recent changes that the models may have missed will be taken into account in the next official outlook to be generated and issued in early October by CPC and IRI, which will include some human judgement in combination with the model guidance.

Climatological Probabilities

Season La Niña Neutral El Niño
DJF 36% 30% 34%
JFM 34% 38% 28%
FMA 28% 49% 23%
MAM 23% 56% 21%
AMJ 21% 58% 21%
MJJ 21% 56% 23%
JJA 23% 54% 23%
JAS 25% 51% 24%
ASO 26% 47% 27%
SON 29% 39% 32%
OND 32% 33% 35%
NDJ 35% 29% 36%

 

IRI ENSO Forecast

IRI/CPC Model-Based Probabilistic ENSO Forecast

Published: May 19, 2016



IRI/CPC Mid-Month Model-Based ENSO Forecast Probabilities

Season La Niña Neutral El Niño
MJJ 2016 6% 91% 3%
JJA 2016 43% 54% 3%
JAS 2016 55% 41% 4%
ASO 2016 58% 36% 6%
SON 2016 58% 34% 8%
OND 2016 58% 32% 10%
NDJ 2016 60% 31% 9%
DJF 2016 61% 32% 7%
JFM 2016 58% 36% 6%

IRI ENSO Forecast

CPC/IRI Official Probabilistic ENSO Forecast

Published: May 12, 2016



CPC/IRI Early-Month Official ENSO Forecast Probabilities

Season La Niña Neutral El Niño
AMJ 2016 3% 35% 62%
MJJ 2016 26% 58% 16%
JJA 2016 52% 42% 6%
JAS 2016 65% 31% 4%
ASO 2016 71% 26% 3%
SON 2016 75% 22% 3%
OND 2016 76% 21% 3%
NDJ 2016 76% 21% 3%
DJF 2017 76% 21% 3%

IRI ENSO Forecast

IRI/CPC ENSO Predictions Plume

Published: May 19, 2016

Note on interpreting model forecasts

The following graph and table show 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 skills also generally decrease as the lead time increases. Thirdly, forecasts made at some times of the year generally have higher skill than forecasts made at other times of the year--namely, they are better when made between June and December than when they are made between February and May. Differences among the forecasts of the models reflect both differences in model design, and actual uncertainty in the forecast of the possible future SST scenario.


Because of occasional data corrections and late model runs following the time of ENSO produce issuance, the data shown in the ENSO forecast table and the ENSO plume graph may not always match. The best source of the ENSO forecast data is http://iri.columbia.edu/~forecast/ensofcst/Data/ensofcst_ALLtoMMYY where MM is the month number and YY is the year.


Seasons (2016-2017)
Model MJJ JJA JAS ASO SON OND NDJ DJF JFM
Dynamical models
NASA GMAO model -1.1 -1.8 -1.9 -1.7 -1.5 -1.4 -1.2
NCEP CFS version 2 -0.3 -0.9 -1.1 -1.2 -1.2 -1.3 -1.4 -1.4
Japan Met. Agency model -0.4 -0.8 -1.1 -1.2 -1.3
Scripps Inst. HCM -0.6 -1.2 -1.7 -2 -2.2 -2.3 -2.3 -2.3 -2.2
Lamont-Doherty model 0.8 0.7 0.6 0.7 0.7 0.5 0.2 0 0
POAMA (Austr) model -0.8 -1.1 -0.9 -0.8 -0.7 -0.7 -0.7
ECMWF model -0.5 -0.7 -0.6 -0.5 -0.5
UKMO model -0.5 -1.1 -1.3 -1.4
 KMA SNU model 0.5 0.2 -0.1 -0.3 -0.5 -0.6 -0.8 -1 0
IOCAS (China) Intermed. Coupled model 0.3 0.2 0.1 -0.1 -0.2 -0.4 -0.5 -0.5 -0.5
COLA CCSM4 model -0.5 -1 -1.3 -1.4 -1.6 -1.7 -1.7 -1.6 -1.4
MÉTÉO FRANCE model -0.4 -0.8 -1.1 -1.1 -1.1
Japan Frontier Coupled model 0 -0.4 -0.6 -0.6 -0.7 -0.8 -1 -1.1 -1
CSIR-IRI 3-model MME 0.1 -0.1 -0.4 -0.5 -0.7 -0.9
GFDL CM2.1 Coupled Climate model -1.2 -1.7 -1.5 -1.2 -0.8 -0.6 -0.5 -0.5 -0.4
Canadian Coupled Fcst Sys -0.3 -0.8 -1.1 -1.2 -1.3 -1.4 -1.4 -1.3 -1.1
GFDL CM2.5 FLOR Coupled Climate model -0.6 -1.1 -1.2 -1.1 -1 -1 -1 -1 -0.9
Average, dynamical models -0.4 -0.8 -0.9 -0.9 -0.9 -1 -1 -1.1 -0.9
Statistical models
NCEP/CPC Markov model 0.3 0.2 0.1 -0.1 -0.1 -0.2 -0.2 -0.2 -0.2
NOAA/CDC Linear Inverse 0.5 0.2 0 -0.3 -0.3 -0.4 -0.5 -0.5 -0.5
NCEP/CPC Constructed Analog 0.2 -0.2 -0.3 -0.4 -0.5 -0.5 -0.5 -0.5 -0.4
NCEP/CPC Can Cor Anal 0.7 0.3 0 -0.2 -0.4 -0.5 -0.7 -0.9 -0.9
Landsea/Knaff CLIPER 0.4 0.3 0.1 -0.1 -0.1 -0.1 -0.1 -0.1 -0.2
Univ. BC Neural Network 0.3 -0.1 -0.2 -0.3 -0.4 -0.4 -0.3 -0.3 -0.3
FSU Regression 0.2 -0.2 -0.3 -0.5 -0.6 -0.7 -0.8 -0.8 -0.7
TCD – UCLA 0.7 0.5 0.4 0.3 0.2 0.2 0.2 0.1 0.1
UNB/CWC Nonlinear PCA 0.8 0.6 0.3 0.1 -0.1 -0.2 -0.2 -0.2 -0.1
Average, statistical models 0.5 0.2 0 -0.2 -0.3 -0.3 -0.4 -0.4 -0.3
Average, all models -0.1 -0.4 -0.6 -0.7 -0.7 -0.7 -0.7 -0.7 -0.6

Discussion of Current Forecasts

All of the set of dynamical and statistical model predictions issued during late April and early May 2016 predict weakening El Niño SST conditions during May, becoming neutral conditions by late May or early June.  El Niño probabilities are 10% or less between May-July and Jan-Mar 2017.   In the most recent week, the SST anomaly in the Nino3.4 region was 0.6 C, reflecting weak El Niño conditions in this weekly time scale, and 1.09 C for the month of April, still at a low-moderate level.  Many of the atmospheric variables continue to reflect El Niño, but less strongly than in previous months. This includes upper level wind anomalies, the Southern Oscillation Index and the pattern of anomalous convection. Based on the multi-model mean predictions, and the expected skill of the models by start time and lead time, the probabilities (X100) for La Niña, neutral and El Niño conditions (using -0.5C and 0.5C thresholds) over the coming 9 seasons are:

IRI/CPC Mid-Month Model-Based ENSO Forecast Probabilities

Season La Niña Neutral El Niño
MJJ 2016 6% 91% 3%
JJA 2016 43% 54% 3%
JAS 2016 55% 41% 4%
ASO 2016 58% 36% 6%
SON 2016 58% 34% 8%
OND 2016 58% 32% 10%
NDJ 2016 60% 31% 9%
DJF 2016 61% 32% 7%
JFM 2016 58% 36% 6%

Summary of forecasts issued over last 22 months

The following plots show the model forecasts issued not only from the current month (as in the plot above), but also from the 21 months previous to this month. The observations are also shown up to the most recently completed 3-month period. The plots allow comparison of plumes from the previous start times, or examination of the forecast behavior of a given model over time. The first plot shows forecasts for dynamical models, the second for statistical models, and the third for all models. For less difficult readability, forecasts are shown to a maximum of only the first five lead times. Below the third plot, we provide a mechanism for highlighting the forecasts of one model at a time against a background of more lightly colored lines for all other models.


Notes on the data 

Only models producing forecasts on a monthly basis are included. This means that some models whose forecasts appear in the Experimental Long-Lead Forecast Bulletin (produced by COLA) do not appear in the table.

The SST anomaly forecasts are for the 3-month periods shown, and are for the Nino 3.4 region (120-170W, 5N-5S). Often, the anomalies are provided directly in a graph or a table by the respective forecasting centers for the Nino 3.4 region. In some cases, however, they are given for 1-month periods, for 3-month periods that skip some of the periods in the above table, and/or only for a region (or regions) other than Nino 3.4. In these cases, the following means are used to obtain the needed anomalies for the table:

  • Temporal averaging
  • Linear temporal interpolation
  • Visual averaging of values on a contoured map

The anomalies shown are those with respect to the base period used to define the normals, which vary among the groups producing model forecasts. They have not been adjusted to anomalies with respect to a common base period. Discrepancies among the climatological SST resulting from differing base periods may be as high as a quarter of a degree C in the worst cases. Forecasters are encouraged to use the standard 1971-2000 period as the base period, or a period not very different from it.

IRI ENSO Forecast

IRI/CPC Experimental ENSO Prediction Plumes Based on the North American Multi-model Ensemble (NMME)

Published: May 19, 2016


The three plots on this page show predictions of seasonal (3-month average) sea surface temperature (SST) anomaly in the Niño3.4 region in the east-central tropical Pacific (5°N-5°S, 120°-170°W), covering the nine overlapping seasons beginning with the current month. The predictions are based on most of the models making up the set belonging to the NMME.

See below for detailed descriptions of the plots.


The first plot (Figure 1) shows the ensemble mean predictions of each of the individual models, and also the average of the individual model predictions (the NMME). Here, the NMME average is not weighted by the number of ensemble members in the individual models. This plot is intended to provide some idea of the disagreement among the individual models. Corrections for systematic biases are not done.

Predictions of ENSO are probabilistic. The ensemble mean prediction it is only a best single guess. On either side of that prediction, there is a substantial uncertainty distribution, or error tolerance. The second plot (Figure 2) shows the estimated probability distribution of the predictions, showing a set of percentiles within that distribution for each lead time. The distribution is modeled as a normal (Gaussian) distribution, so that the NMME forecast represents the center, or 50 percentile, in the distribution. On either side, other percentile values are shown symmetrically, ranging from 1 to 99 and including some intermediate percentiles (5 and 95, 15 and 85, and 25 and 75). The plot enables a user to estimate the probability of the Niño3.4 SST anomaly to be greater or less than some critical value, or within some interval. If, for example, the 85 percentile falls at 1.8° C above average, the probability of the SST exceeding 1.8° C can be estimated at 15%. Probabilities for exceeding or not exceeding values not exactly on percentile line can be roughly interpolated by eye. The overall width of the probability distribution is derived from the historical skill of the hindcasts of the NMME models, from 1982 to present, for the specific forecast start time and lead time. This method of defining the probability distribution represents one of two general approaches, the other approach being a direct counting of ensemble members within each of the percentile bands. This second approach assumes that the ensemble spreads of the models are true representations of the uncertainty. One reason the second approach is not used here is that the ensemble spreads may have biases in representing the real world uncertainty. Individual model spreads have often been found to be somewhat narrower than they should be, although in multi-model ensembles this tendency has been shown to be milder or even eliminated. Another reason the ensemble member counting approach is not used here is that there are not enough ensemble members in the NMME to produce a smooth probability distribution, particularly for the relatively detailed percentile bands presented here.

The third plot (Figure 3), sometimes called a spaghetti diagram, shows synthetically generated prediction scenarios that are equally likely. Here, 100 scenarios are shown; any number can be generated for such a diagram. Each scenario is produced using a random number generator, combined with knowledge of the NMME forecast and its uncertainty, as well as the amount of persistence of anomalies. The degree of persistence of anomalies is based on the correlation of prediction errors from one lead time to another. In other words, the individual lines are designed to show the correct amount of persistence as expected in nature, rather than jumping around more randomly from one lead time to the next. Sometimes the “spaghetti density” may appear asymmetric about the NMME forecast or outside of the 85 and 15 percentile lines. This is purely sampling variability, and would not occur if many thousands of such lines were plotted. But with that many lines, most of the plot would be too crowded to get a sense of the behavior of the lines near the center of the distribution. The main purpose of the diagram is to serve users who want to assess realistic individual scenarios of ENSO behavior rather than statistical summaries of the forecast like the percentiles shown in the second plot.