From the Potsdam Institute for Climate Impact Research (PIK) where breakthroughs happen even before they have a proven track record. It appears he was able to predict one event last year in 2012 from the method run in 2011, but just one success does not a breakthrough make, especially when we are dealing with a chaotic system and the method depends on “inspecting emerging teleconnections”. Only time will tell if his idea has any skill. Hansen has already tried and failed on predicting El Niño. -Anthony
Breakthrough in El Nino forecasting
In order to extend forecasting from six months to one year or even more, scientists have now proposed a novel approach based on advanced connectivity analysis applied to the climate system. The scheme builds on high-quality data of air temperatures and clearly outperforms existing methods. The study will be published this week in the Proceedings of the National Academy of Sciences.
“Enhancing the preparedness of people in the affected regions by providing more early-warning time is key to avoiding some of the worst effects of El Niño,” says Hans Joachim Schellnhuber, director of the Potsdam Institute for Climate Impact Research and co-author of the study by Josef Ludescher et al (Justus-Liebig Universität Giessen). The new approach employs network analysis which is a cutting-edge methodology at the crossroads of physics and mathematics. Data from more than 200 measurement points in the Pacific, available from the 1950s on, were crucial for studying the interactions between distant sites that cooperate in bringing about the warming.
Extending the forecasting time but also enhancing the reliability
According to Schellnhuber a new algorithm was developed and tested which does not only extend the forecasting time but also enhances the reliability. In fact, the novel method correctly predicted the absence of an El Niño-event in the last year. This forecast was made in 2011 already, whereas conventional approaches kept on predicting a significant warming far into 2012.
El Niño is part of a more general oscillation of the Pacific ocean-atmosphere system called ENSO, which also embraces anomalous cold episodes dubbed La Niña which can inflict severe damages as well. The present study focuses on the warming events only. However, an El Niño-year is followed by a La Niña-year, as a rough rule.
Climate change: a factor for ENSO changes?
“It is still unclear to which extent global warming caused by humankind’s emissions of greenhouse gases will influence the ENSO pattern,” says Schellnhuber. “Yet the latter is often counted among the so-called tipping elements in the Earth system, meaning that at some level of climate change it might experience a relatively abrupt transformation.” Certain data from the Earth’s past suggest that higher mean global temperatures could increase the amplitude of the oscillation, so correct forecasting would become even more important.
Article: Ludescher, J., Gozolchiani, A., Bogachev, M.I., Bunde, A., Havlin, S., Schellnhuber, H.J. (2013): Improved El Niño forecasting by cooperativity detection. Proceedings of the National Academy of Sciences (early online edition) [DOI:10.1073/pnas.1309353110]
Weblink to the article once it is published: http://www.pnas.org/cgi/doi/10.1073/pnas.1309353110
Improved El Niño forecasting by cooperativity detection
-
Contributed by Hans Joachim Schellnhuber, May 30, 2013 (sent for review March 12, 2013)
Abstract
Although anomalous episodic warming of the eastern equatorial Pacific, dubbed El Niño by Peruvian fishermen, has major (and occasionally devastating) impacts around the globe, robust forecasting is still limited to about 6 mo ahead. A significant extension of the prewarning time would be instrumental for avoiding some of the worst damages such as harvest failures in developing countries. Here we introduce a unique avenue toward El Niño prediction based on network methods, inspecting emerging teleconnections. Our approach starts from the evidence that a large-scale cooperative mode—linking the El Niño basin (equatorial Pacific corridor) and the rest of the ocean—builds up in the calendar year before the warming event. On this basis, we can develop an efficient 12-mo forecasting scheme, i.e., achieve some doubling of the early-warning period. Our method is based on high-quality observational data available since 1950 and yields hit rates above 0.5, whereas false-alarm rates are below 0.1.
Preprint here: http://arxiv.org/ftp/arxiv/papers/1304/1304.8039.pdf (h/t to Dr. Leif Svalgaard)
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
preprint here http://arxiv.org/ftp/arxiv/papers/1304/1304.8039.pdf
I’ve heard that Kreskin is already capable of such predictions.
I compiled this nearly two years ago
http://www.vukcevic.talktalk.net/ENSO.htm
will have to update sometime.
“Breakthrough in El Nino forecasting”
Let me fix that headline for you . . .
“Breakthrough in El Nino forecasting claims”
They’ve optimized a statistical algorithm. No physical modeling. Might actually work; a simple dumb classifier that doesn’t pretend to know anything about global warming.
t’s a pity that the paper is besmirched by the name of Mr. Great Transformation himself.
What does Tisdale have to say?
Hansen failed at something? Oh I get it, it has a short time horizon and can be fact checked easily. This is not the stuff fat retirements and expensive solar arrays are bought with.
I think I might do rather better with my tarot cards. Not that I believe that pieces of pasteboard can foretell the future but many do. I collect them because some are beautiful works of art especially those based on B P Grimaud.
So cross my palm with silver and we shall see,
Kindest Regards
OK, he’s got a prediction method. How about using it to actually make a prediction? We are now past the difficult boreal spring period, so he should be able to make the prediction covering the year between now and July 1, 2014.
Since Figure 2 ends at about Jan 1, 2011, and the level of the index is below the threshold, Schnellnhuber is entitled to state that his model correctly predicted no El Nino in 2011. But enough of these retrodictions, how about putting the model on the line?
Our method is based on high-quality observational data available since 1950 and yields hit rates above 0.5, whereas false-alarm rates are below 0.1.
If it works this well in the future, it will miss almost half of them.
This looks to me like a modest step forward in statistical modeling of an important phenomenon.
So what does the method say about late 2013/early 2014 El niño chances?
so, a slightly biased coin. ?
if it lands heads.. El Nino…
tails.. no idea.
crosspatch says: “What does Tisdale have to say?”
In summary, Ludescher et al (2013) presented another statistical model:
http://bobtisdale.wordpress.com/2013/07/02/el-nino-in-the-news/
I hope it works.
I think this kind of thing is the future of climate modelling. The reason being, even semi-reliable weather forecasts a year to a decade into the future are worth very large amounts of money because they allow many things to be planned in a way they can’t be at the moment, starting with agriculture.
To date, untold billions have gone into the physics based climate models, which have been an utter failure. Far larger amounts of money have been wasted on activities based on the outputs of these models and businesses and politicians are realizing this.
Once you have a statistical model that seems to work, you look for the underlying physical mechanisms. Essentially, back to the drawing board on how the climate actually works.
Bob Tisdale, you are generously hopeful that the new statistical model will prove to be useful in your excellent-as-usual commentary here:
http://bobtisdale.wordpress.com/2013/07/02/el-nino-in-the-news/ .
I have a couple of (very UN-educated) questions (don’t bother with them if they are not worthwhile — or, just use them as mere launching pads to present other, more helpful information!):
1) A. Given the varying intensities of El Nino’s, is predicting (if they ever get to that level of ability!) the mere event all that useful to anyone, farmers, for instance? (or are they predicting intensity, too?)
B. Is there a minimum likely land temperature-raising effect so that the mere event of an El Nino is significant?
2) Are there other, countervailing, natural events that could easily cancel out a predicted (if they ever get there) El Nino’s land temperature-raising effect (thus, making a prediction largely a waste of time)?
Thanks, also, to Leif Svalgaard and to A-th-y for drawing this article to our attention.
El Nino is poorly defined–Do we trust the sea buoys? What about the exact dates? What about the averages? Do we only go with area 3.4?
This seems to be a game where everyone can win, especially if more funding is at stake….
Re: my questions at 5:17PM:
1) Please don’t let my addressing my post to Bob Tisdale prevent anyone who would be willing, if you would be so kind, to answer my questions. Thanks!
2) I sloppily used the term “land temperatures” to refer to what should correctly be called: “land surface air temperatures” in my post.
Janice, I will answer your questions. (Boy, will this draw comments….)
1) A. Given the varying intensities of El Nino’s, is predicting (if they ever get to that level of ability!) the mere event all that useful to anyone, farmers, for instance? (or are they predicting intensity, too?)
We know enough about El Nino’s and La Nina’s to certainly advise farmer in continental US.
B. Is there a minimum likely land temperature-raising effect so that the mere event of an El Nino is significant?
I don’t understand the question. 🙂
2) Are there other, countervailing, natural events that could easily cancel out a predicted (if they ever get there) El Nino’s land temperature-raising effect (thus, making a prediction largely a waste of time)?
Sure, anything can happen, but El Nino/La Nina is a significant phenomenon–do not get in the way….
I think forecasting the ENSO 1 year out is just going to be a 50:50 proposition in any event which is more-or-less what the authors are saying.
I’ve made some predictions going out 1 year (6 different times over 6 different years now) and about 50:50 is all I have gotten to (just saying that things change and the system is a self-reinforcing oscillation that can push out and accelerate in a certain direction and then pull back decelerate in the other direction which couldn’t have been foreseen).
I think 3 months to 6 months out can be alot more accurate. Mainly because the sub-surface equatorial ocean drives the ENSO on a short-term basis. Right now, the water that is surfacing in the circulation and at the starting pointof the ENSO at the Galapagos Islands, is very cold and this will provide a short-term cooling of the ENSO regions. But coming right behind it, is alot of warmer water which will cut-off any La Nina development. After that, no El Nino can form either because there is just so much cooler-than-normal water in the eastern Pacific.
Official call, short-term cooling up to the end of August, then slight warming till February , then neutral again. Sounds like covering all the bases but that is what the pattern says now.
http://www.cpc.ncep.noaa.gov/products/GODAS/pent_gif/xz/movie.temp.0n.gif
http://www.cpc.ncep.noaa.gov/products/GODAS/pent_gif/xy/movie.h300.gif
And to add to Janice Moore’s request, she would welcome answers from other persons so that she gets additional input to the one I provided here:
http://bobtisdale.wordpress.com/2013/07/02/el-nino-in-the-news/#comment-12106
which was:
Reasonably accurate predictions would always be helpful to farmers and others whose production is influenced by weather patterns. But predictions are based on tendencies. During El Ninos, the jet streams tend to portray one spatial pattern, and during La Ninas, the jet streams tend to be in another location. I believe (don’t know for sure) that the strength of the ENSO events also dictates the strength of the jet stream’s physical pattern and if other factors could influence the pattern.
Not sure if that helped or if I was stating the obvious.
Bill Illis: Reasonable prediction.
From that can we assume because the tropical Pacific isn’t creating strong El Nino or La Nina jet stream patterns, we’ll continue to see flukey weather?
Luckily, it’s been cool and gloomy where I am–just like I like it.
1. What causes the Pacific hotspot before an El Nino?
2. At a given site, what is the correlation between Tmax taken on adjacent days? How does this compare with correlations taken hundreds of km apart?
3. GIGO?
If I was a farmer, I would NOT rely on a 1 year out forecast for the ENSO for any planting decisions. I mean none unless you want to guess wrong half the time.
I would look at everything I could in the 3 to 6 months out timeframe if that matters to your region.
If you are in Australia or Indonesia or southern China, a La Nina forecast means that you should plant crops that that benefit from extra rainfall. If an El Nino is forecast, plant the drought-tolerant crops. And this is very important because precipitation can be +/- 50% in this region depending on the ENSO.
If you are in the Horn of Africa, an El Nino means you will get more rain and even floods. A La Nina means severe drought. This impact is not that well recognized by some but the Horn of Africa is one of the most impacted regions from ENSO.
If you are in the south and south-east US and also California, La Nina provides drought and high temperatures. An El Nino provides more more rain and less extreme temperatures.
If you are in the North American north-west, like Australia and Indonesia, you are most affected by the ENSO of anyone. El Nino can mean warm winters but potentially severe drought in the summer. La Nina brings cold winters but more rain in the spring and early summer. In a significant El Nino, plant early and plant drought-tolerant crops. La Nina, plant later and put in the high-cash crops that grow better in more rain.
If you are in northern South America, El Nino brings less rain and a La Nina brings more rain.
If you are in the rest of the planet, the ENSO is not that important. It doesn’t change temperature or rainfall patterns that much.
On the other hand, the higher CO2 levels in the atmosphere means that almost every crop will grow 50% better than it did 75 years ago. Irrigation is going to be needed less and less as levels rise even further. But then, C3 weeds will start to become more of a problem and more weed reduction strategies will be required in corn and suger cane crops particularly but also for all crops.
“According to Schellnhuber a new algorithm was developed and tested which does not only extend the forecasting time but also enhances the reliability. The new approach employs network analysis which is a cutting-edge methodology at the crossroads of physics and mathematics. On this basis, we can develop an efficient 12-mo forecasting scheme, i.e., achieve some doubling of the early-warning period.”
Congratulations Herr Schellnhuber, you just earned a “Put up or shut up” moment here on WUWT. Use your “cutting-edge methodology at the crossroads of physics and mathematics” new algorithm to provide us with your prediction / projection of what ENSO will do this coming year. Give us a prediction today (first week of July 2013) that we can evaluate in late summer of 2014.
Enlighten us with your WIzardry, Herr Schellnhuber. Take your best shot and give us your guess.
The smart money is on Bob Tisdale.