A couple of weeks ago, the Cato Institute ran a major advertisement in several leading newspapers across the country, intending to counter some seemingly alarming statements about climate change that were coming from the Obama Administration—primarily statements concerning the urgency of action and the certitude of the science behind the perceived crisis.
The ad campaign quickly drew criticism in the blogosphere from folks who share President Obama’s sense of urgency and are also leaders in raising climate alarm. The loudest among these critics was perhaps Dr. Joe Romm, who runs the blog Climate Progress. Romm’s criticism was quite vehement, but it was also quite wrong.
The Cato ad laid out the case against the certainty of drastic climate change occurring either now or in the foreseeable future. The ad began:
“We, the undersigned scientists, maintain that the case for alarm regarding climate change is grossly overstated.”
And it proceeded to explain why. First:
“Surface temperature changes over the past century have been episodic and modest and there has been no net global warming for over a decade now.”
This statement was designed to show that the degree of global temperature rise is variable in time and that during the past decade or so the rate of warming has been exceedingly slow. This was presented as a counter to the oft-made alarmist claim that the rate of global temperature rise is accelerating.
The Cato ad then declared:
“After controlling for population growth and property values, there has been no increase in damages from severe weather-related events.”
This statement was a direct refutation to alarmist claims that disasters around the world are clearly on the increase because of anthropogenic climate influences. While it is true that weather-related damage costs are on the increase, so too is human population and wealth. Without properly accounting for demographic changes, it is impossible to attribute the rise in disaster losses to climate variability, much less to anthropogenic climate change, and no scientific study has convincingly done so.
Next in the ad came this:
“The computer models forecasting rapid temperature change abjectly fail to explain recent climate behavior.”
The point of this statement was that if the existing suite of climate models is unable to capture the range of natural variability at the broadest spatial scales (for which there is growing evidence), then we should maintain a healthy skepticism (at the very least) as to the reliability of their projections of the future climate.
The Cato ad closed with:
“Mr. President, your characterization of the scientific facts regarding climate change and the degree of certainty informing the scientific debate is simply incorrect.”
Simple and straightforward.
The criticism that the ad generated largely surrounded the scientific studies that were cited to back up its claims. The number of citations included in the actual ad was limited, perhaps too much so, but they were not inappropriate. Romm and others contend otherwise.
Romm’s particular point of attack was the citation of a very recent study published in the journal Geophysical Research Letters by researchers Kyle Swanson and Anastasios Tsonis from the University of Wisconsin-Milwaukee. The Swanson & Tsonis paper was entitled “Has the climate recently shifted?” and concluded that it was quite likely that the climate went through a state-change around the 2001/2002 time period, shifting from a period of rapidly rising temperatures that marked the 1980s and 1990s, to a period of relatively small temperature changes that has thus far marked the 21st century. After identifying other such climate state-changes during the past century, Swanson & Tsonis suggested that it was quite possible that the current period marked by the depressed rate of global temperature rise could continue for another decade or so into the future (see here for more about the Swanson &Tsonis paper). The Cato ad cited Swanson & Tsonis (along with a citation of the Hadley Center global temperature data set) in support of the statement “Surface temperature changes over the past century have been episodic and modest and there has been no net global warming for over a decade now”—true to the findings of Swanson & Tsonis (especially when accompanied by the global temperature record).
Romm was incensed, proclaiming that “a key study Cato uses to argue we may see much less warming than the models predict comes to exactly the opposite conclusion.” The study he was referring to was the same Swanson & Tsonis But the Cato ad did not use it to directly argue that the climate models were predicting too much warming—just that there had been little temperature rise during the past decade and that the global temperature rose in fits and starts rather than smoothly and continuously. Granted, this was one piece of the “climate change is not alarming” puzzle, but far from being its sole foundation.
Romm’s issue with the Swanson & Tsonis cite was that in the “Conclusions” section of their journal article—a section that allows for speculation from the authors—they wrote:
Finally, it is vital to note that there is no comfort to be gained by having a climate with a significant degree of internal variability, even if it results in a near-term cessation of global warming. It is straightforward to argue that a climate with significant internal variability is a climate that is very sensitive to applied anthropogenic radiative anomalies. If the role of internal variability in the climate system is as large as this analysis would seem to suggest, warming over the 21st century may well be larger than that predicted by the current generation of models, given the propensity of those models to underestimate climate internal variability.
This led Romm to the interpretation that the results of Swanson & Tsonis pointed to a future that was going to warm even faster than climate models were currently projecting. In his view, Swanson & Tsonis was a reason to ring the climate alarm bells louder, and he was appalled that they found their way into Cato’s ad calling for a muting of the alarm.
But Romm’s conclusions about the implications within Swanson & Tsonis are off base. In my opinion, way off base.
In fact, the findings of the authors quite possibly make one of the strongest cases against the accuracy of climate models when replicating the observed global temperature history: If the Swanson & Tsonis results are right, then climate models appear to get things right for the wrong reason—which likely means, that when using the right reasons, they would get things wrong. And by “things,” I mean the observed global temperature history of the past 100 years.
At issue is the impact of aerosol emissions on the global temperature history. Aerosols, when taken in net, are presumed to have a cooling effect on global temperatures. The problem is that, scientifically, we just don’t know how large the cooling effect is.
Aerosols were introduced into climate models when it was realized that without them, the models produced too much warming from greenhouse gas increases alone. How much cooling the aerosols provide varies from climate model to climate model—but in each case, aerosols provide just the right amount of cooling to counteract just the right amount of warming, such that each climate model closely reproduces the observed global climate history (see the results of Kiehl, 2007, for more details about this remarkable model characteristic). And, of course, while aerosol emissions were used to depress the greenhouse warming during the 20th century, 21st century emissions scenarios assume that aerosol emissions will decrease, which unleashes the full warming power of greenhouse gases and things heat up in a hurry. How convenient.
But Swanson & Tsonis find that the episodic nature of the global temperature rise during the past 100 years is an inherent part of the internal variability of the earth’s climate system—aerosols need not apply.
This is big, for it means that if internal reorganizations of the earth’s climate, not aerosols, are responsible for the interspersed multidecadal periods of warming/non-warming, then the models have the processes wrong. Take away their aerosols and the models produce too much warming—which means that their climate sensitivity is too large. This argues that climate models currently produce too much warming in the future, not too little (as Romm would have it).
So a full examination of the implications of Swanson & Tsonis reveals that Dr. Romm was far too hasty to condemn the Cato ad for this citation, and, in fact, suggests that an even broader use of its results in detailing climate model flaws could be warranted.
The bottom line is that the issue as to whether or not climate models are accurately replicating observations is one of active research, and it far from being settled. And until it is satisfactorily settled, reliance on climate-model projections of our climate future for local, state, national, and/or international policy is a risky and ill-advised endeavor.
An intent of the Cato ad was to draw attention to this fact.
References:
Kiehl, J.T., 2007. Twentieth century climate model response and climate sensitivity. Geophysical Research Letters, 34, L22710, doi:10.1029/2007GL031383.
Swanson, K.L., and A.A. Tsonis, 2009. Has the climate shifted? Geophysical Research Letters, 36, L06711, doi:10.1029/2008GL037022.
Knowing what I do about the authors of that study’s statements to the press about their recent research, I would bet that the point which Romm wishes to emphasize was put in there by Swanson. It’s an obligatory nod to Catastrophic AGW where none is really appropriate. They are saying that “Yes, natural variability is greater than we thought, but even though this calls attribution studies into question, it is possible that natural variability did not cause the warming but instead hid it from view, meaning sensitivity is greater than we thought, and things are worse than we thought”-a technically true statement, but misleading, because it is actually more likely that the ~opposite~ is true-that natural variability has exaggerated the warming, thus creating the illusion of a climate system which is highly sensitive to Anthropogenic forcing. Plus, they are assuming that the natural variability ~response~ is larger than expect from a “known” level of natural “forcing”-when it could really easily be the opposite-the natural forcings are underestimated, so the response is about right or too high, meaning sensitivity is actually low. These to qoutes are quite illuminating:
““When the climate kicks back out of this state, we’ll have explosive warming,” Swanson said. “Thirty years of greenhouse gas radiative forcing will still be there and then bang, the warming will return and be very aggressive.”” From the article in Discovery
““We realized a lot of changes in the past century from warmer to cooler and then back to warmer were all natural,” Tsonis said. “I don’t think we can say much about what the humans are doing,” he added.” From a local paper in Wisconsin.
Great points! When you put it that way, it definitely is clear that these shifts probably mean less warming, not more.
The speculations on climate sensitivity in Swanson et al paper’s conclusions may logically go both ways: if natural variability has cooled the Earth over the past decades, then yes, it has masked the models warming and sensitivity would be much “higher than predicted”. But if natural factors have warmed the Earth, then the GHG warming effect and hence sensitivity is much lower than predicted.
Over the past 3 decades, there have been many observations that suggest natural variability have increased temperatures rather than decreaesed : el Nino dominant phase of the PDO, el Chichon+Pinatubo in the 80 & 90 and the exceptional 98 el Nino which conjure to skew the temperature trend up, the high solar activity of the past 3 solar cycles (they are among the 6 cycles with most sunspots for more than 3 centuries).
The fact that Swanson et al -and Dr Romm- chose only the high sensitivity hypothesis in their conclusion is evidence that their speculation is biased by at best PC conformity pressure, at worse unscientific and preconceived views.
I keep seeing statements like this one from the Swanson & Tsonis conclusion section cited above:
Mojib Latif has made similar arguments when discussing current GWMs inaccuracy at the decadal time scale. I wonder why no one has stepped forward to point out that a superpositional decomposition of this sort is completely invalid for the analysis of coupled non-linear, time-variant systems such as global climate dynamics. One simply can not separate “natural variations” from anthropogenic forcing in a system whose feedback sensitivities are functions of the output state, a fact easily provable and taught in every first year physics curriculum. Superposition only applies to linear, time-invariant (LTI) systems, which our climate is most assured not.
For the same reason, the only scientifically valid conclusion one can draw when observations exceed the confidence intervals of the model, as recent MGT observations have, is that the model is invalid. This is true whether exiting on the low side or high side. It is inaccurate to say AGW is better (or worse) than we thought and therefore or century scale projections must be adjusted downward (or upward). The only thing one can say is that at this point in GCMs development, they have nothing predictive to offer to the policy debate nor can they be used to identify the source of observed warming (or cooling) trends.
Check out Dr. Roy Spencer’s latest book, THE GREAT GLOBAL WARMING BLUNDER, in which he shows with global satellite data that the climate exhibits strong NEGATIVE FEEDBACK (limits temperature changes due to forcings) in contrast to the IPCC models which all have varying degrees of POSITIVE FEEDBACK
(amplifies temperature changes due to CO2 forcings).
Andrew,
It is definitely complicated. I agree with Dr. Swanson that when the climate shifts, then the temperature will increase at a faster rate (then present), but, I don’t necessarily agree that it will be at a rate much greater than expected. For instance, the warming during the 1980s and 1990s (presumably part of which was being stored up by the cooling state of the 1940-1970s) wasn’t greater than expected (from climate models). Observations seem to indicate that during Swanon&Tsonis’s “cool” periods, global temperature rise doesn’t occur very fast (despite increases in radiative forcing from growing greenhouse gas concentrations), and that during Swanson&Tsonis “warming” periods, the warming occurs along the lines of model expectations. Take the two periods together, and you get less total warming than projected.
-Chip
In case anyone needs further proof of the implications of Swanson&Tsonis concerning sulfate aerosols, the same researchers, along with Geli Wang, just published an article that appeared today in Geophysical Research Letters that examines some potential reasons that the climate goes through a state-change every couple of decades or so. One thing that is conspicuously absent in their latest work is any mention of sulfate aerosols in playing a role. Perhaps one could argue that the role of sulfates is tied together with greenhouse gases to influence the interplay of the major climate modes that the authors have identified as being behind the state-changes. But that argument becomes largely untenable as the authors find the same mechanisms at play in climate model control runs—i.e., in runs not forced by changes in atmospheric greenhouse gases or aerosols. Thus, these state-changes seem to be part of the background climate system, rather than something unique to the era of human climate influence.
Again, I’ll make the point that if sulfates are not the primary reason for the episodic nature of the global temperature history of the past 100 years or so, then the climate models have some serious issues to deal with.
-Chip