Guilyardi et al (2009) “Understanding El Niño in Ocean-Atmosphere General Circulation Models: progress and challenges”
Guest post by Bob Tisdale
The preliminary Reynolds OI.v2 sea surface temperature data for July 2012 won’t be available until next Monday July 30th, and there haven’t been any changes in weekly NINO3.4 region or global sea surface temperature anomalies since the update last week that are worth a separate post. So I’ve decided to give readers a heads-up about a paper I’m using as a primary reference for Chapter 5.8 Scientific Studies of the IPCC’s Climate Models Reveal How Poorly the Models Simulate ENSO Processes of my upcoming book Who Turned On The Heat?: El Niño-Southern Oscillation, the Unsuspected Global Warming Culprit.
The paper is Guilyardi et al (2009) Understanding El Niño in Ocean-Atmosphere General Circulation Models: progress and challenges. It is basically an overview of the many problems experienced by climate models in their attempts to simulate ENSO, and as such they cite more than 100 other papers. Many of the cited papers discuss individual climate model problems, while others look at many models, while still other papers are detailed discussions of ENSO processes. If you’re not extremely familiar with ENSO, some of the topics may be overly technical. On the other hand, some of the problems discussed may even seem nitpicky, but Guilyardi et al do explain why these model errors are, in fact, problems.
The “sidebar” discussions, highlighted in burnt orange, are about how
wellpoorly climate models capture the mean state and annual cycle of the tropical Pacific. The models have such severe problems with their portrayal of the mean state and annual cycle that you’re left asking yourself one question: With all of those problems, how could the modelers think that the El Niño and La Niña events they place atop the mean state have any chance at capturing reality?
During their discussion of the model portrayal of El Niño and La Niña events, Guilyardi et al (2009) use the term “skewness”. With skewness, they’re referring to the fact that since 1976, El Niño events have been stronger than La Niña events. ENSO has been skewed toward the El Niño phase. But as Guilyardi et al (2009) discuss, they’re not in the climate models. Modeled La Niña events tend to be the same strength as El Niño events. Let’s take that further. If the models also produce the same number of La Niña events as El Niño events, then modeled ENSO would become a nonentity. Is this the reason why models show the equatorial Pacific warming nearly as fast as, or faster than, the mid latitudes, depending on model vintage? (See Figure 1, which is Figure 5-37 from my upcoming book.) When in reality, the equatorial Pacific has cooled over the past 30 years, while the mid latitudes have warmed. This indicates to me that in the real world, ENSO-related processes are distributing warm water from the tropics toward the mid latitudes in the Pacific—something that’s not captured by the models. And that’s a big problem.
In the upcoming book Who Turned On The Heat?: El Niño-Southern Oscillation, the Unsuspected Global Warming Culprit, I discuss and illustrate the major problems with climate models presented by Guilyardi et al to make them easier to understand. But there is a sentence in that paper that needs no further explanation. They write:
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).
Do you think that sentence will make it to the IPCC’s 5th Assessment Report? It should.