Is There an Improved Method of El Niño Forecasting? –And– There’s No Consensus About ENSO in Paleo-Studies
AN IMPROVEMENT IN EL NIÑO FORECASTING?
A recently published paper by Ludescher et al (2013) Improved El Niño forecasting by cooperativity detection (paywalled) claims to be able to forecast an El Niño event more than a year in advance. The EurekAlert press release is here. And the preprint version of the paper is here. WattsUpWithThat provides an introductory post here.
Statistical and dynamical models are presently used to predict whether the upcoming winter with be impacted by an El Niño or a La Niña, or if it will be a neutral El Niño-Southern Oscillation (ENSO) season. Those models are hindered by a phenomenon called the spring predictability barrier. But this new method, which is another statistical model, is supposed to be able to hurdle that barrier and make predictions a year ahead based on measured variables from more than 200 locations around the Pacific. But it is limited to El Niño events—at present.
I hope the method presented in Ludescher et al (2013) works. But I do recall papers in the past that have made similar claims—never to be heard of again.
WHAT THE PRESS HAS TO SAY
In USA Today, Dan Vergano and Doyle Rice provide an overview of the findings of Ludescher et al (2013), but the reporters do have a number of errors and poor wordings in an early paragraph. They write:
El Niños strike every decade, driven by warm Pacific Ocean water piling on the West Coast and affecting weather worldwide, triggering floods, droughts and heat. The most recent El Niño ended in 2007.
The most recent official El Niño occurred during 2009/10, so they missed one. See Figure 1. It represents the sea surface temperature anomalies of the NINO3.4 region (5S-5N, 170W-120W), which is an east-central portion of the equator in the Pacific, starting in November 1981. Sea surface temperature anomalies in that region are a commonly used index for the frequency, strength and duration of El Niño and La Niña events. El Niño events are the large upward spikes and La Niñas are the downward ones. That monthly ENSO index actually reached El Niño conditions (NINO3.4 sea surface temperature anomalies warmer than 0.5 deg C) during the (boreal) summer of 2012, but they did not stay there long enough to have been classified as an official El Niño event, nor did they extend into the normal El Niño and La Niña (boreal) winter season. Nonetheless, the tropical Pacific did release extra heat into the atmosphere last year, but it was only for a short time—a couple of months.
(Figure 1 is from my most recent sea surface temperature update.)
And not all El Niños reach the west coast of South America. The ones that do are called East Pacific El Niños, and the ones that do not extend all the way to the coast of the Americas are called Central Pacific El Niño events or El Niño Modoki. (See the El Niño Modoki discussions here, here and here.)
And it’s true that El Niños strike every decade, but that could be misleading, because they strike multiple times a decade. In the 1950s, there were 4 El Niños, according to NOAA’s Oceanic NINO Index-based table of Historical El Nino/ La Nina episodes (1950-present). The 1960s had 3, including one that lasted through 2 ENSO seasons. The 1970s had 3. During the 1980s, the frequency dropped, there were only 2 El Niños, but one was extremely strong and the second lasted for about a year and half. The 1990s initially saw two moderate El Niños, followed by the colossal 1997/98 El Niño. Then the 2000s saw a string of 3 moderate El Niños and it ended with the formation of the 2009/10 El Niño. So the 1950s and the 2000s both had 4 El Niños; the 60s, 70s and 90s had 3 each; and the 1980s had 2.
For those who want to look at the decades before the 1950s, see the Oceanic Nino Index-like tables created from HADISST-based NINO3.4 sea surface temperature anomalies here. Keep in mind that observations are very sparse along the equatorial Pacific in the early part of the data (especially before the opening of the Panama Canal in 1914) and that Giese et al (2009) indicate that a few of the El Niño events before 1950 were likely stronger.
The USA Today article also includes quotes from Anthony Barnston, an ENSO researcher from Columbia University’s International Research Institute for Climate and Society (IRI), and from Michael Mann, paleoclimatologist, both of whom are skeptical of the results of Ludescher et al (2013).
Best of luck to Ludescher et al (2013). It would be nice to be able to predict ENSO events more than a few months in advance.
NO CONSENSUS ABOUT ENSO IN PALEO-STUDIES
Personally, paleoclimatological studies do not stand high on my list of believable climate studies. I prefer data-based studies. Which proxies paleo-studies use and the weighting of the proxies have significant impacts on the outcomes, and some of the methods and proxies used in paleoclimatological studies have been found to be questionable. So paleo-studies always seemed to me to be open to too much of the authors’ preconceptions. When paleo-studies add climate models, they require another leap of faith because climate models are imperfect representations of climate: make-believe models trying to simulate make-believe data.
The USA Today article by Dan Vergano and Doyle Rice also give honorable mention to Li et al (2013) El Niño modulations over the past seven centuries (Paywalled). Supplementary Information is here. Li et al (2013) has been getting a lot of press over the past few days. Anthony Watts has a quick post about it here. Vergano and Rice write about Li et al (2013):
A related study in the Nature Climate Change journal suggests that the chances of that next El Niño might be more likely because of the world’s warming climate. Led by Jinbao Lu [sic] of the University of Hong Kong, the study looked at 2,222 tree-ring growth records over the last seven centuries from the Pacific Rim. They find El Niño happened a lot more than normal over the last 50 years, just as temperatures worldwide rose due to global warming, suggesting a connection. Past studies have been more uncertain about climate effect’s [sic] on El Niño.
It appears Vergano and Rice may even have misspelled Dr. Jinbao Li’s name.
That aside, if “Past studies have been more uncertain about climate effect’s [sic] on El Niño” but Li et al (2013) are more certain, that strongly suggests that there is no paleo study-based consensus about ENSO—not that Li et al (2013) are correct.
The Li et al (2013) abstract includes:
Our data indicate that ENSO activity in the late twentieth century was anomalously high over the past seven centuries, suggestive of a response to continuing global warming.
The word “continuing” is a nice touch.
Ray & Giese (2012) Historical changes in El Niño and La Niña characteristics in an ocean reanalysis came to a different conclusion. Their abstract ends:
Overall, there is no evidence that there are changes in the strength, frequency, duration, location or direction of propagation of El Niño and La Niña anomalies caused by global warming during the period from 1871 to 2008.
Also, Vergano and Rice interviewed Michael Mann about Ludescher et al (2013), but they failed to mention that Mann, a paleoclimatologist, was part of another very detailed paleoclimatological study of El Niño, the results of which were published at the end of last year.
That 2-part paper was Emile-Geay et al (2012) “Estimating Central Equatorial Pacific SST variability over the Past Millennium”. The abstract of Part 1 is here and the paper is here. The abstract of Part 2 is here and the paper is here.
The conclusions of Emile-Geay et al (2012) begin on page 20 of the second part of the paper linked above. They did not come to the same conclusion as Li et al (2013). The last 2 sentences of the Emile-Geay et al (2012) conclusions read (my boldface):
Progress on this question is most pressing: whether or not the tropical Pacific climate responded consistently to natural radiative forcing in the recent past bears directly on the question of its real-world sensitivity to anthropogenic climate change. Therefore, this knowledge could directly impact our ability to predict global and regional climate variability over the coming century.
My favorite conclusion of Emile-Geay et al (2012), though, is the first one. In it, they blame the sparseness of the sea surface temperature data in the NINO3.4 region since the 1850s for much of the uncertainty in the paleoclimatological results. While they propose methods for improving the sea surface temperature data, any reconstruction of sea surface temperature of the central equatorial Pacific is still speculation.
And then there’s Ault et al (2013) “Characterizing decadal to centennial variability in the equatorial Pacific during the last millennium”. Preprint version is here. Their abstract reads (my boldface):
The magnitude of sea surface temperature variability in the NINO3.4 region of the equatorial Pacific on decadal and longer timescales is assessed in observational data, state-of-the-art (CMIP5) climate model simulations, and a new ensemble of paleoclimate reconstructions. On decadal to multidecadal timescales, variability in these records is consistent with the null hypothesis that it arises from “multivariate red noise” (a multivariate Ornstein-Uhlenbeck process) generated from a linear inverse model of tropical ocean-atmosphere dynamics. On centennial and longer timescales, both a last millennium simulation performed using the Community Climate System Model 4 (CCSM4) and the paleoclimate reconstructions have variability that is significantly stronger than the null hypothesis. However, the time series of the model and the reconstruction do not agree with each other. In the model, variability primarily reflects a thermodynamic response to reconstructed solar and volcanic activity, whereas in the reconstruction, variability arises from either internal climate processes, forced responses that differ from those in CCSM4, or non-climatic proxy processes that are not yet understood. These findings imply that the response of the tropical Pacific to future forcings may be even more uncertain than portrayed by state-of-the-art models because there are potentially important sources of century-scale variability that these models do not simulate.
It’s no surprise that climate models can’t simulate ENSO variability on century time scales. Climate models can’t simulate the basic processes of ENSO. Why would researchers think they’d perform well over long terms? I’ve linked the following study many times in the past and I’ll link it here again. It’s Guilyardi et al (2009) Understanding El Niño in Ocean-Atmosphere General Circulation Models: progress and challenges. It discusses decadal, annual and seasonal model failings when trying to simulate ENSO. It includes a statement that’s very similar to the quote I provided above from Emile-Geay et al (2012). Guilyardi et al (2009) wrote:
Because ENSO is the dominant mode of climate variability at interannual time scales, the lack of consistency in the model predictions of the response of ENSO to global warming currently limits our confidence in using these predictions to address adaptive societal concerns, such as regional impacts or extremes (Joseph and Nigam 2006; Power et al. 2006).
DID GLOBAL WARMING CAUSE THE EL NIÑOS OR DID EL NIÑOS CAUSE GLOBAL WARMING?
Numerous datasets indicate that El Niño events are fueled naturally. Additionally, satellite-era sea surface temperature records indicate that El Niño events are responsible for the warming of sea surface temperatures over the past 31 years, not vice versa as Li et al (2013) have suggested. If this topic is new to you, refer to my illustrated essay “The Manmade Global Warming Challenge” [42MB].