Learn how plants respond to higher atmospheric CO2 concentrations

How does rising atmospheric CO2 affect marine organisms?

Click to locate material archived on our website by topic

Predictive Skill: Guess How Many Climate Models Passed the El Niņo Test?
Volume 3, Number 22: 13 September 2000

We are a society obsessed with tests, and rightly so.  From crash-test dummies to medical doctors, we want to have confidence that those to whom we entrust our futures are indeed trustworthy; and the way we as a society make this appraisal is via standardized tests designed and administered by experts in whom we place our trust.

In an eye-opening article in the September 2000 issue of the Bulletin of the American Meteorological Society, Christopher W. Landsea and John A. Knaff apply this time-honored tradition of testing to the climate models we are asked to trust enough to totally restructure national and world energy policies in a way that will dramatically affect practically everyone on earth ? and not necessarily for the better!  Specifically, these NOAA scientists employ a simple statistical tool to evaluate the skill of twelve state-of-the-art climate models in real-time predictions of the development of the 1997-98 El Niņo.

Their findings?  In the words of the authors, "the current answer to the question posed in this article's title [How much skill was there in forecasting the very strong 1997-98 El Niņo?] is that there was essentially no skill in forecasting the very strong 1997-98 El Niņo at lead times ranging from 0 to 8 months."  Indeed, they say, "there were no models ? able to anticipate even one-half of the actual amplitude of the El Niņo's peak at medium range (6-11 months) lead."  Also, "since no models were able to provide useful predictions at the medium and long ranges, there were no models that provided both useful and skillful forecasts for the entirety of the 1997-98 El Niņo."  [Authors' italics].

These observations raise an interesting question.  If a climate model is totally lacking in skill at predicting the largest climatic phenomenon of our day, which raises havoc all around the world, can it be any good at predicting the climatic phenomenon that people anticipate could negatively impact the planet in the near future, i.e., CO2-induced global warming?  Almost anything, of course, is possible; but for a model to fail this well-defined and objective real-world test certainly does not engender confidence that it would do any better at predicting future long-term warming or cooling.  Hence, it would logically follow that if a group of people were intent on convincing the world that anthropogenic CO2 emissions are bad for the planet, in support of some other agenda, there would be a great temptation to hide the fact that the best climate models in existence failed this important test.

So how would one go about hiding this fact?  Perhaps the most effective way would be to claim that just the opposite was true, i.e., that the models performed admirably in predicting the development of the 1997-98 El Niņo.  And reading between the lines of the Landsea and Knaff paper, that is exactly what appears to have happened.

The authors state, for example, that their results "may be surprising given the general perception [our italics] that seasonal El Niņo forecasts from dynamical models have been quite successful and may even be considered a problem solved."  In this regard, they cite the Science report of Kerr (1998), entitled "Models win big in forecasting El Niņo," which, they claim, was based on an "unrefereed and incomplete analysis," noting further that when the study was finally completed and published, with the results demonstrating that the models "did not 'win big' after all," Science was silent.

"Also disturbing," the authors state, "is that others are using the supposed success in dynamical El Niņo forecasting to support other agendas," citing as an example the American Geophysical Union's Position Statement on Climate Change and Greenhouse Gases, which suggests that confidence in the use of models to predict anthropogenic global warming is enhanced by their ability to predict the El Niņo-Southern Oscillation (ENSO) phenomenon.

"The bottom line," say Landsea and Knaff, "is that the successes in ENSO forecasting have been overstated (sometimes drastically) and misapplied in other arenas."  Indeed, the results of their study, they say, should engender even "less confidence in anthropogenic global warming studies because of the lack of skill in predicting El Niņo."

It is unfortunate that the stakes in the global warming debate have grown so high that problems such as these are beginning to permeate the science.  For that, we suppose, we have politics to thank.  An important implication for all to ponder is the difficult task it has become to discern the truth of many matters.  Even statements that sound like recitations of established fact are sometimes anything but fact.  And if a little knowledge is a dangerous thing, think of the disaster that can come from misinformation on a topic of major concern.

Dr. Craig D. Idso
Dr. Keith E. Idso
Vice President

Kerr, R.A.  1998.  Models win big in forecasting El Niņo.  Science 280: 522-523.

Landsea, C.W. and Knaff, J.A.  2000.  How much skill was there in forecasting the very strong 1997-98 El Niņo?  Bulletin of the American Meteorological Society 81: 2107-2119.