How Strong Was That El Niño or La Niña? – No One Knows For Sure

We recently discussed and illustrated how the differences between sea surface temperature datasets prevented us from knowing which of the recent strong El Niño events (the 1982/83, 1997/98 or 2015/16 El Niños) was actually strongest.  See the post here. That post, of course, was intended to counter all of the nonsense from alarmist bloggers and the mainstream media about the current El Niño being the strongest ever.

In this post, we’re going to illustrate the differences between the monthly long-term (1870 to present) NINO3.4 region sea surface temperature anomalies from 5 datasets to further show that the differences grow considerably as we travel back in time.

The post concludes with a recent comment at NOAA’s ENSO blog about the uncertainties of NINO3.4 sea surface temperature data from a well-known and well-respected ENSO researcher. My thanks to Larry Kummer, Editor of the FabiusMaximus blog, for calling my attention to it on the thread here at WattsUpWithThat.


The sea surface temperature anomaly data for the NINO3.4 region are a commonly used index for the strength, frequency and duration of El Niño and La Niña events. The NINO3.4 region covers a large area of the equatorial Pacific.  See Figure 1.  It is bordered by the coordinates of 5S-5N, 170W-120W.  It covers an area that’s roughly 6.2 million square kilometers (or 2.4 million square miles).  As references, Australia covers a surface area of about 7.7 million km^2 and the contiguous United States covers about 8.0 million km^2.  So the NINO3.4 region is not small.

Figure 1

Figure 1

Since the early 1990s, NOAA has had moored buoys in place that sample sea surface temperature and a number of other metrics in the tropical Pacific. (I believe there are 20 buoys in the NINO3.4 region.) Prior to then, the number and location of sea surface temperature samples depended on ship traffic.  Drifting buoys (not ARGO floats) have also sampled sea surface temperatures in the NINO3.4 region since the early 2000s, but the number of samples and their locations depend on whether the drifters have wandered into the NINO3.4 region. The number of monthly in situ temperature samples for the NINO3.4 region that are available from the International Comprehensive Ocean-Atmosphere Data Set (ICOADS) are shown in Figure 2.

Figure 2

Figure 2

The large dip and rebound in 2013 was due to temporary budget cuts by NOAA for TAO project maintenance, if memory serves.

Two long-term reconstructions also include satellite-based data, in addition to the in situ data from various types of buckets, from ship inlets and from moored and drifting buoys.

For their sea surface temperature reconstructions, data suppliers like NOAA and the UK Met Office then make numerous adjustments to the source sea surface temperature data from ICOADS to account for the different types of measuring devices.


Figure 3 compares the sea surface temperature anomalies for the NINO3.4 region from 5 different reconstructions, for three timeframes:

Note: I used the Cowtan and Way data because the HADSST3 data are not infilled, and there are numerous gaps in the HADSST3 NINO3.4 region data due to the poor sampling in that region, especially during the world wars. Also for you consideration, for the period of Jan 1950 to October 2015, the correlation coefficient of the HADSST3 and the Cowtan and Way NINO3.4 sea surface temperature anomalies is 0.99. [End note.]

The top comparison runs from January 1870 (the start year of the HADISST data) to October 2015 (the last month of HADISST data as of this writing).  The anomalies are referenced to the full term of the data (1870 to 2014) so not to skew the later results. The center comparison starts in January 1950 and the bottom comparison starts in 1975.

Figure 3

Figure 3

Because of the overlaps, it’s difficult to see the differences between the reconstructions of NINO3.4 region sea surface temperature anomalies. So I prepared Animation 1, which presents the individual products.  As shown, there can be noticeable differences in the strengths of El Niño and La Niña events.  Also, the ERSST.v3b data are missing a few events prior to the mid-1880s.

Animation 1

Animation 1


How large are those differences?   Figure 4 presents the differences between the monthly minimum and maximum NINO3.4 sea surface temperature anomalies for those five datasets.  That is, I had the spreadsheet determine the monthly maximum and minimum values; then I subtracted the minimums from the maximums.  In looking at the differences between datasets, consider that, according to NOAA, the threshold for a weak El Niño is a NINO3.4 sea surface temperature anomaly of +0.5 deg C, for a moderate El Niño the threshold is +1.0 deg C and for a strong El Niño it’s +1.5 deg C.  The thresholds are the reverse for weak, moderate and strong La Niñas.  See the footnotes in the NOAA ENSO Blog post here. Also, please note that the y-axis in the top graph of Figure 4 is different than the other two.

Figure 4 NINO3.4 Spread

Figure 4

Note: We’ve already discussed how NOAA’s transition from ERSST.v3b to ERSST.v4 changed which seasons are considered El Niño and La Niña events in their Oceanic NINO Index. See the post Weak El Niños and La Niñas Come and Go from NOAA’s Oceanic NINO Index (ONI) with Each SST Dataset Revision.  That is, even updates to sea surface temperature datasets can change what NOAA considers to be a weak El Niño or La Niña season. Most notably, with the ERSST.v3b data, the 2014/15 season registered as a weak El Niño, but with the ERSST.v4 data, the 2014/15 season became ENSO neutral. [End note.]

The 2.5 deg C spike that peaks in boreal winter 1877/78 was caused by the ERSST.v3b data not including the El Niño that happened then. Refer again to Animation 1.

On the other hand, the 3.0 deg C spike in April 1919 is caused by something that’s odd.  Figure 5 includes the evolutions of the 1918/19 El Niño from the five datasets.  There’s a curious one-month drop-off in the new NOAA ERSST.v4 “pause buster” sea surface temperature data in April 1919.

Figure 5

Figure 5

Makes one wonder how many other anomalous downward spikes appear in the early ERSST.v4 data throughout the global oceans.  (No, I’m not going to search for them. You can if you like.)


As noted in the opening of this post, Larry Kummer, Editor of the FabiusMaximus blog, called my attention to a comment at the NOAA ENSO Blog about the uncertainties of NINO3.4 sea surface temperature anomaly data.  The comment appears on the thread of the NOAA post December El Niño update: phenomenal cosmic powers! by Emily Becker, and the Tue, 2015-12-15 18:24 comment that follows was written by Anthony Barnston, a well-known ENSO researcher from the International Research Institute for Climate and Society (IRI). The two sea surface temperature datasets Anthony Barnston refers to are the NOAA “pause-buster” ERSST.v4 (in-situ only data) and the NOAA Optimum Interpolation SST data version 2 (a.k.a. Reynolds OI.v2), which is satellite-enhanced data. He writes in response to a question about data accuracy (my boldface):

The accuracy for a single SST-measuring thermometer is on the order of 0.1C. It may give a read-out to hundredths, but the last digit would be wobbling around as different water touched the sensor (on a ship, on on a buoy, for example). It could be recorded either to the nearest tenth or the nearest hundredth. But that’s for one thermometer. We’re trying to measure the Nino3.4 region, which extends over an enormous area. There are vast portions of that area where no measurements are taken directly (called in-situ). The uncertainty comes about because of these holes in coverage. Satellite measurements help tremendously with this problem. But they are not as reliable as in-situ measurements, because they are indirect (remote sensed) measurements. We’ve come a long way with them, but there are still biases that vary in space and from one day to another, and are partially unpredictable. These can cause errors of over a full degree in some cases. We hope that these errors cancel one another out, but it’s not always the case, because they are sometimes non-random, and large areas have the same direction of error (no cancellation). Because of this problem of having large portions of the Nino3.4 area not measured directly, and relying on very helpful but far-from-perfect satellite measurements, the SST in the Nino3.4 region has a typical uncertainty of 0.3C or even more sometimes. That’s part of why the ERSSv4 and the OISSTv2 SST data sets, the two most commonly used ones in this country, can disagree by several tenths of a degree. So, while the accuracy of a single thermometer may be a tenth or a hundredth of a degree, the accuracy of our estimates of the entire Nino3.4 region is only about plus or minus 0.3C. Sorry about this big disapointment. It bothers me also. We need thousands of ships evenly spaced across the region to get a truly accurate reading. It’s not worth the money, so it isn’t going to happen. If we improved satellite measurement technology, that could be the key.

I think it would be safe to assume that the +/- 0.3 Deg C accuracies noted are for recent decades, since the early 1990s, when the TAO project buoys were in place. Before then, the sampling was much poorer and the differences between datasets are considerably larger.


Not only do the differences between sea surface temperature datasets and the uncertainties of the data prevent us from knowing the strengths of El Niño or La Niña events, those differences and uncertainties also prevent us from knowing if El Niño or La Niña events have taken place and how long they lasted.


The sea surface temperature data, the Cowtan and Way data and the number of ICOADS sea surface temperature samples are available from the KNMI Climate Explorer.

About Bob Tisdale

Research interest: the long-term aftereffects of El Niño and La Nina events on global sea surface temperature and ocean heat content. Author of the ebook Who Turned on the Heat? and regular contributor at WattsUpWithThat.
This entry was posted in El Nino-La Nina Processes, ENSO Update. Bookmark the permalink.

8 Responses to How Strong Was That El Niño or La Niña? – No One Knows For Sure

  1. Hifast says:

    Reblogged this on Climate Collections and commented:
    Bob Tisdale again provides a supreme treatment of SSTs and ENSO historical data.

  2. Thanks, Bob. Very good explanation on the largest heat-transfer phenomenon on Earth (our oceans planet).
    ENSO can be large or small, or not there. ENSO seems to be tripolar and bi-sexual.
    No wonder it is almost impossible to predict its future; we cannot even tell its history.
    I think the best one-number index that has been developed is the ESRL-PSD Multivariate ENSO Index (Klaus Wolter, NOAA), at

  3. Pingback: How much damage has this El Niño done? | Fabius Maximus website

  4. Pingback: ENSO-Update Februar 2016: El Niño geht – La Niña kommt – wobleibtdieglobaleerwaermung

  5. johnmarshall says:

    A geologist,James Kamis, has come up with a theory that strong pulses of hot water from the sea bed circulation reinforce the El Nino warming effect. He claims that earthquakes correlate well with the ENSO cycles. This was posted on “notalotofpeopleknowthat” web site and may be worth a look. See what you think Bob.

  6. Bob Tisdale says:

    Thanks, John. But color me skeptical. Earthquakes are not an indicator of subsurface volcanic or hydrothermal vent activity.


  7. johnmarshall says:

    OK Bob but, earthquake activity is a strong indicator of an impending eruption from any volcano, also ocean vents, mainly on the oceanic ridge system, have associated earthquakes to R4 and below. Whether flow rates increase during these quakes is a matter of debate but to my mind a quake would increase flow rates. This super heated water, at around 450C, mixes with the surrounding water at 2C. The ratio of hot to cold water is very much on the cold side so I too am sceptic of this theory despite it comming from a fellow geologist.

  8. Bob Tisdale says:

    John, also see my comment at notalotofpeopleknowthat

    It reads:
    I only got as far as Figure 1. No need to go beyond that.

    Figure 1 shows an area in the eastern tropical Pacific outlined as “El Nino ‘Warming’” when in reality there are La Nina to ENSO neutral conditions in the equatorial Pacific. Surface temperature anomalies in the NINO1+2 region do not define an El Nino. So the author needs to better understand the definition of an El Nino.

    Second, subsurface temperatures are monitored along the equator in the Pacific by the NOAA TAO Project, and there is no evidence that the warm surface water outlined in Figure 1 originated in the western equatorial Pacific.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s