Guest commentary from Steve Ghan
A good writer knows their audience, and Roy Spencer knows his. There are plenty of people who would love to hear a compelling argument for why no action is needed to mitigate global warming, and Spencer’s book “The Great Global Warming Blunder: How Mother Nature Fooled the World’s Top Climate Scientists” will give uncritical readers the argument they’ve been looking for. As Sarah Palin said, “while we recognize the occurrence of these natural, cyclical environmental trends, we can’t say with assurance that man’s activities cause weather change”. That is really the essence of Roy’s argument.
What is the Great Blunder? According to his book, “a fundamental mistake has been made in previous interpretations of satellite data”…”a mix-up between cause and effect when analyzing cloud and temperature variations”.
Who made this mistake? Invariably, it is “the IPCC researchers”. He cites a couple of specific papers by Piers Forster, but finds no fault with them. So he casts aspersions into the wind.
Spencer’s assertion in his book of that there has been a “mix-up between cause and effect” is quite a different conclusion from his recent article published in the Journal of Geophysical Research – Atmospheres in 2010, which concluded innocuously that “since the climate system is never in equilibrium, feedbacks in the climate system cannot be diagnosed from differences between equilibrium climate states” … despite the fact that this is the exact diagnosis supporting his conclusion in the book.
In his book Spencer contends that short-term fluctuations in the energy balance and surface temperature are consistent with a low climate sensitivity: “A careful examination of the satellite data suggests that manmade warming due to a doubling of atmospheric carbon dioxide could be less than 1ºC – possibly much less.”
Why does Spencer consider his “discovery” of a mix-up between cause and effect to be so important? Because “natural cloud fluctuations in the climate system will cause a bias in the diagnosed feedback in the direction of positive feedback”, which means those careless IPCC researchers have vastly overestimated the climate sensitivity. He then asserts that “if the real climate system looks sensitive to climate modelers, they will build their models to be sensitive also.” But he never mentions the fact that climate models have produced climate sensitivities of 2-5ºC per doubling of CO2 concentration for decades, well before the Forster papers and before satellite measurements were available for those careless anonymous IPCC researchers to produce biased estimates of climate sensitivity.
Moreover, how could models explain the observed warming during the 20th century if the climate sensitivity is as low as it is, unless aerosols don’t cool and there is some other warming mechanism? Spencer addresses this question with a hypothesis that natural cloud variability is the cause of longer term trends. He proposes a relationship between the Pacific Decadal Oscillation (PDO) and clouds by considering a variety of combinations of initial ocean temperature, ocean thickness, cloud feedback, and forcing by clouds (neglecting forcing by CO2 and the water vapor feedback entirely) in a simple energy balance model, and finds a relationship between PDO and clouds using 9 years of satellite data. By exploring parameter space randomly he found the agreement with the observed 20th century warming was best for an initial ocean temperature 0.6ºC below normal, which means almost all of the warming that his model explains is simply the ocean returning to normal, not the response to decadal variability in clouds. (Ed: note that the details of this calculation are heavily criticised by Barry Bickmore in a series of posts). Of course, decadal variability in clouds can only be a response to decadal variability in the surface conditions or atmospheric circulation that drive cloud formation, because the lifetime of cloud systems is days rather than decades.
How, then, does Spencer explain the ice ages? He essentially punts, saying he believes the ice core record is “irrelevant”, that “we don’t have a clue” what was causing those climate variations. But if climate really is as insensitive as he claims it to be, the climate forcing producing the ice ages must have been huge, much larger than the radiative forcing from orbital changes, surface albedo, and greenhouse gases. He claims to prefer empiricism to theoretical models, yet when the paleo data supports the higher climate sensitivity simulated by climate models, he blames some unknown mechanism. It reminds one of the old cartoon of the physicist drawing mathematical theories madly on the blackboard, completing his theory with the statement “and then a miracle occurred”.
Spencer does make a valid point about the potential for bias toward exaggerating problems because it can bring in more funding. We all must be wary of this. On the other hand, it’s worth noting that the book market tends to financially reward a bias toward contrarianism.
But for me his credibility as a climate scientist was most compromised with his assertion that “it would take only one research study to cause the global warming house of cards to collapse.” So much for weighing the evidence. As Arnold Schwarzenegger said about the diversity of views of climate scientists, if your child is ill and 98 out of 100 doctors call for life-saving surgery and 2 say it is not necessary, your decision is obvious.
Roy Spencer is respected for his remote sensing expertise, but the conclusions of his book are nothing like those in his JGR article. What a difference an audience can make.
Suspension of comments: Due to Roy Spencer being caught up in a loss of power related to the tornado outbreak in Alabama, we are suspending comments on this post until he is in a position to respond (should he choose to).
Update (05/05/11): Comments have been re-opened.