by Rud Istvan
In 2001, MIT’s Professor Richard Lindzen and colleagues published a controversial paper titled “Does the Earth have an Adaptive Infrared Iris?”  If there were a tropical adaptive infrared iris, then Earth’s sensitivity to GHGs would be much less than the IPCC had supposed.
The essence of Lindzen’s adaptive iris was simple and logical. He began by noting that satellite observations of upper troposphere humidity show sharply delineated high and low humidity regions, especially in the tropics (his Fig. 1)
He then observed that low humidity regions were cirrus ‘deficient’, while high humidity regions were cirrus ‘surplus’. (Cirrus clouds have an inordinate impact on ‘greenhouse effects’. These high thin ice clouds have low ‘short wave’ albedo so are nearly transparent to incoming sunlight, while high albedo and nearly ‘opaque’ to outgoing long wave (infrared) radiation–simply because cirrus clouds are comprised of ice crystals. Cirrus indisputably net warms.)
Lindzen’s hypothesis was that high altitude tropical cirrus results from remnant moisture ‘detrained’ from towering cumulonimbus clouds (convection cell thunderstorms). Depending on a lot of complicated cloud microphysics, a warmer sea surface could produce more massive convection cells containing more moisture, which would precipitate more rainfall, leaving less upper troposphere residual moisture ‘detraining’ to form warming cirrus. So the proportions of upper troposphere “dry” and ‘moist’ regions would counter-intuitively change as SST warmed in favor of ‘dry’–and less cirrus. Hence there would be an adaptive infrared iris where warmer sea surfaces would produce via intermediate tropospheric convection and precipitation processes less cirrus, and therefore a net negative tropical cloud feedback. (This oversimplified description suffices for this post.)
A new paper modeling Lindzen’s possible iris effect has just been published by Mauritzen and Stephens. (Paywalled, but the Supplementary Information is available.) The abstract reads (emphasis added):
Equilibrium climate sensitivity to a doubling of CO2 falls between 2.0 and 4.6 K in current climate models, and they suggest a weak increase in global mean precipitation. Inferences from the observational record, however, place climate sensitivity near the lower end of this range and indicate that models underestimate some of the changes in the hydrological cycle. These discrepancies raise the possibility that important feedbacks are missing from the models. A controversial hypothesis suggests that the dry and clear regions of the tropical atmosphere expand in a warming climate and thereby allow more infrared radiation to escape to space. This so-called iris effect could constitute a negative feedback that is not included in climate models. We find that inclusion of such an effect in a climate model moves the simulated responses of both temperature and the hydrological cycle to rising atmospheric greenhouse gas concentrations closer to observations. Alternative suggestions for shortcomings of models — such as aerosol cooling, volcanic eruptions or insufficient ocean heat uptake — may explain a slow observed transient warming relative to models, but not the observed enhancement of the hydrological cycle. We propose that, if precipitating convective clouds are more likely to cluster into larger clouds as temperatures rise, this process could constitute a plausible physical mechanism for an iris effect.
This important paper got recent attention at BishopHill, and at Climate Audit in part 3 of Nic Lewis’ explanation of his Ringberg climate sensitivity presentation. Nic’s post also provides additional evidence and references for points 2 and 3 in the following paragraph.
Important discrepancies exist between CMIP3/5 results and observations, reviewed in depth (with many footnote references) in essays Humidity is still Wet, Cloudy Clouds, and Sensitive Uncertainty in ebook Blowing Smoke:
- There is no observational support for the modeled upper troposphere hotspot. (See Prof. Christy’s 2014 APS testimony posted previously at CE.)
- Modeled tropical precipitation is significantly less than observed.
- Modeled cloud feed back is positively overstated (too little modeled tropical cloud fraction and optical depth)
- Modeled TCR and ECS are about 2x observational calculations.
The new paper by Mauritzen and Stevens is hardly conclusive, since based on a single AOGCM model, ECHAM6. But it is highly suggestive that the adaptive iris exists. Model ‘experiments’ tested three possible iris ‘strengths’ (this post illustrates ‘medium’, Ie=0.5). The model ‘experiments’ directionally, but not fully, resolve all four model/observation discrepancies listed in the previous paragraph. For example, the IRIS modified model reduces ECHAM6 sensitivity from 2.8 to 2.2, but not to ~1.7 computed by Lewis and Curry (2014) using the observational energy budgets in IPCC AR5.
1. The unobserved tropical troposphere model hotspot is much reduced
2. Tropical precipitation increases toward observed levels.
All four factors suggest one reason CMIP5 models are running hot is that they lack a negative adaptive iris feedback.
“All models are wrong, but some are useful.”  The new IRIS version of ECHAM6 appears to ‘usefully’ help reconcile several important modeled to observed discrepancies. Fresh evidence that climate science is not settled.
 BAMS 82: 417-432 (2001).
 Mauritsen and Stevens, Missing Iris effect as a possible cause of muted hydrological change and high climate sensitivity, Nature Geoscience, doi10:1038/ngeo2414.
 Box and Draper, Empirical Model Building and Response Surfaces, ISBN 0-471-81033-3 (1987), page 424.
JC comment: There is a companion post Observational support for Lindzen’s Iris Hypothesis, which addresses theoretical and observational issues of the iris hypothesis.