Error Theory

For the first decade of global warming alarmism (dating from James Hansen dire warnings to Congress in 1987) the skeptics were in a tough position. They could only point to the illogic of thinking that the greenhouse warming effects of carbon-dioxide could be anywhere near as large as the alarmists were claiming. But how to account for the fact that some warming seemed to be occurring, if not by the greenhouse gas effects?

The logical arguments against greenhouse gas effects are strong. Carbon-dioxide traps the same wavelengths of infrared as water vapor does, but on average there is about sixty times as much water vapor as carbon-dioxide in the atmosphere. The only time an increase in carbon-dioxide can have a significant marginal heat-trapping effect is in the absence of water vapor, which only occurs in the coldest air masses, in the middle of the polar night, when a tiny bit of warming effect is going to hurt what, exactly? (See Satanic Gases by Patrick Michaels and Michael Balling, p.89.)

The tiny amount of extra heat trapped by carbon dioxide is imagined by the alarmists to start a viscious cycle, where warming causes increased evaporation of water vapor from the oceans, which traps more heat, etcetera. Some such feedback must occur, but if the cycle is really vicious, why would carbon dioxide be needed to start it? Water vapor is doing all the work. Why put the blame on carbon-dioxide? Talk about the tail wagging the dog.

Not suprisingly (given that water vapor feedback effects do not spin out of control) there are some compelling explanations for why water vapor feedback effects might be self-limiting. For one, water vapor creates clouds, which reflect away sunlight. If they reflect away more energy than they trap--as cumulus clouds do--then cloud effects would place a natural limit on water vapor’s “vicious cycle.” (MIT climatologist Richard Lindzen has studied this process.) Also, the more water vapor in the atmosphere, the more efficient the rain cycle, leaving drier air behind, opening up a hole in the sky for heat to escape, and rain (condensation) puts the heat right there (the opposite of the cooling generated by evaporation)in a position to rise thorough the window.

Still, without an alternative explanation for observed warming, the greenhouse gas skeptics got steamrolled. One candidate for an alternative explanation is the solar weather: the storms of solar-magnetic flux that erupt from sunspot activity. Correlations between sunspot activity and climate have been observed for over a century, but no one understood the mechanism. Solar storms cause a slight increase in solar luminescence, but not enough to create a significant warming effect. Where could the warming effect of sunspots be coming from?

Starting in the late 1990’s, astrophysicists started figuring out the answer to this riddle. The storms of solar-magnetic flux that erupt from sunspot activity shield the Earth from cosmic radiation. Cosmic radiation ionizes the atmosphere, which seems to be crucial for cloud formation, so sunspots have the effect of blowing away the cloud cover, giving the Earth a sunburn. Add that sunspot activity has been very high since the 1940's, and the slight global warming observed since the mid 70's could easily be due to this effect.

Now the skeptics are armed, not just with a debunking of the alarmists’ greenhouse gas explanation for observed warming, but also with a compelling alternative explanation of their own. How are the alarmists responding? They are fighting like mad not to acknowledge the competing reason and evidence. Listen to Gavin Schmidt, a global warming alarmist working under the original global warming alarmist, James Hansen, at NASA/Goddard Institute for Space Studies. Schmidt is explaining to a non-scientist (science fiction writer Jerry Pournelle) why it is appropriate to ignore the new sunspot science:

“[T]here is no obvious need for ‘new’ or unknown physics to explain what was going on,” Schmidt says That's it! He doesn't "need" it. He and his colleagues have been able to tweak their models so that, when a list of historical variables not including sun-spot behavior is entered, the output of the model matches historical temperature measures reasonably well. So what if they are leaving gigantic natural temperature forcings out of their model? They have simply decided that they can do without this new-fangled information.

Earth to NASA. Fitting the temperature data isn't enough. You have to comprehend ALL the data: all the physical effects that science can discern. Any model can be tweaked to fit historical temperature data, but if it leaves out physical mechanisms that we know are at work, then it is WRONG. When the alarmists leave the warming effects of solar weather out of their models, the warming that ought to be attributed to sunspot activity ends up being wrongly attributed to other potential warming effects, primarily greenhouse gas warming.

Projecting these exaggerated warming effects forward, when we know they are exaggerated, is grotesquely irresponsible. Not only is it a gross violation of scientific integrity, it is a highly damaging violation, aiming to shut down economic activity on the basis of what is essentially a lie. The responsible and scientifically sound thing to do is to take our best estimate of the effects of sunspot activity on solar forcing and incorporate it into our computer models (the GCM's, or General Circulation Models).

This would be easy enough to do. Using satellites, we can directly measure the effect of sunspots on the average reflectivity of the Earth. We have an extensive historical record of sunspot activity, both visual (ever since Galilleo invented the telescope) and in the geologic record (from measuring the isotope residue of cosmic radiation). Accounting the implied variation in solar forcing would improve the calibration of the model when it is fit to historical data. No longer would the warming effects of higher solar forcing be wrongly attributed to greenhouse gas effects.

The problem is that the computer modelers don’t WANT to incorporate the new findings. They apparently like their models, and the known-to-be-untenable predictions that they yield, just the way they are. Listen to some more of Gavin Schmidt’s rationalizations:

“We have proxy measurements (sunspot counts, cosmogenic isotopes 14C and 10Be) that go back that far, but they can't be indepentedly [sic] calibrated to solar output. They are correlated to the irradiance over a sunspot cycle, but they are really magnetic-related phenomena, not irradiance-related. So, absent a good physical model for the sun that links these things, people make estimates based on their best guesses. Thus there is a huge variation in estimates of this long-term component.”

Schmidt is breathtakingly misleading here. First, he is pretending that the important thing about sunspots is the variation in solar output that goes with them. The cosmic radiation residue doesn’t measure changes in luminescence, only changes in how much cosmic radiation reaches Earth, so it supposedly is not the data we really need. But surely Schmidt knows that the important thing about sunspots under the new theory is not the slight increase in solar output but the change in the percentage of solar output that gets absorbed by the Earth when there are fewer clouds to block it. For this, the critical data is exactly what we have a geological record of: the effect of sunspots on cosmic radiation.

After minimizing “magnetic-related phenomenon” and pretending that it is luminescence that matters, notice what Schmidt does next. He goes on to suggest that, since estimates of the effects of sunspot activity are subject to widely differing estimates, it is best to leave them out altogether. Really? It isn’t better to go with an expected value, or a maximum likelihood estimate? The best way to deal with stochastic variables is to leave them out? This guy's whole job is to deal in widely varying estimates! Now he is going to pretend that such estimates should be eschewed?

We certainly do need to figure out how much credence to give to estimates based on immature understanding, which is the state of the entire climate science field. But the alarmists are not being cautious in this respect. They are predicting catastrophe and calling for drastic action. How about a little moral sanity? ALL best estimates need to be incorporated, and the result needs to be taken with the appropriate many grains of salt. Instead, the alarmists dismiss what militates against their alarmist predictions while paradoxically proclaiming that the cautious thing to do is embrace the most extreme predictions they can come up with, and engage in drastic action to avoid those predictions.

In short, the alarmists--the dominant force in climate science today--are not scientists. They are propagandists, doing what propagandists always do: picking and choosing what reason and evidence to account or dismiss in order to fashion the best case for their preferred conclusions. This can be done as easily with hard science as with any other sphere of reason and evidence. Of course it requires a willingness to embrace illogic. Schmidt’s rationalizations for dismissing contrary reason and evidence are, upon inspection, patently untenable. But he can still, in this mode, compile data sets and solve equations and build computer models. He can still be a "scientist," just an exceedingly bad and dangerous one.

There has been a recent spate of global warming alarmism, with two major alarmist studies hitting the newspapers in the last week. Both of these studies do exactly what Gavin Schmidt does. They set aside the new understanding of solar weather in order to maintain exaggerated claims about the warming effects of greenhouse gases.

Meeting the Climate Challenge, the report of the International Climate Change Taskforce, suggests that if carbon-dioxide reaches 400ppm, temperatures will rise to more than 2 degrees centigrade above pre-industrial temperatures (we are at 380ppm now). “Beyond the 2C level,” the report intones ominously, “the risks to human societies and ecosystems grow significantly,” and “the risks of abrupt, accelerated or runaway climate change also increase.” With carbon-dioxide increasing at a rate of 2ppm/year, well, you get the picture! Within ten years, we may have already reached the tipping point of global climate catastrophe! STOP BURNING FOSSIL FUELS!!! IT IS A MATTER OF IMMEDIATE LIFE AND DEATH FOR THE PLANET!!!!!!

Bullshit. This is pure propaganda, derived by discounting the known warming effects of solar-magnetic flux, errantly attributing these warming effects to greenhouse gases, then projecting the misattributed warming effects forward, in combination with every other damaging assumption that can be mustered. True, the study has the sanction of the head of the IPCC, the U. N. umbrella organization for climate research. Unfortunately, that is not a recommendation. Thanks to the political success of the alarmists over the last fifteen years, they have managed to gain control of most of the hiring and the purse strings for almost the entire discipline. A climate scientist coming up today had better be a Gavin Schmidt, ready to dissemble for the cause.

The second alarmist study, reporting the results of climateprediction.net’s distributed processing project, does exactly the same thing. It uses the same GCM’s that the IPCC does, but uses the power of distributed processing to test more parameter variations. First the different specifications are calibrated using historical temperature data, then they are projected forward. With no solar-weather effects in the model, the warming effects of the last fifty years of severe solar weather get attributed to greenhouse gases. Not suprisingly, the most extreme of the resulting predictions are alarming.

The distributed processing approach could actually be very useful for integrating solar weather into the GCMs. Start with what seems to be the logical way to specify solar-weather effects, then toy around with the specification, looking for improvements in the fit to historical data. How can a group that is explicitly taking an “experimentalist” approach to identifying the best model justify leaving a major theoretical effect entirely out? Maybe Schmidt can pretend that effects we aren't quite sure how to specify are better left out, but experimentalists? Come on people. Start doing your jobs. Are you scientists, or are you propagandists? So far you are acting like propagandists.

It isn't just the strength of the sunspot theory that is determinative here. Careful scientists knew all along that greenhouse effects were being exaggerated. Also crucial is the infancy of climate science. We don't yet have the science to predict how warm a jacket of greenhouse gases we will want the Earth to be wearing fifty or a hundred or two hundred years from now. Alarmist prescriptions--that people should stop impacting the environment--are predicated on the assumption that human impacts on climate are large compared to natural impacts, implying that if we like our current temperature (we are adapted to it, and we know it is not unstable) then we should avoid human impacts.

Denial of natural temperature variation was a keystone of the IPCC's Third Assessment Report (2001), appearing as the notorious "hockey stick" diagram of figure 1. This picture of flat natural temperature, until human's entered the picture and caused temperature to turn sharply upward, now lies in shreds. It turned out to be a pure artifact of data handling errors. At the same time, the new sunspot theory adds a powerful natural driver of temperature change to the mix (to go along with volcanoes, orbital fluctuations, and who knows what we might yet discover).

As soon as the assumption of natural temperature stability is set aside, the prescription that human impacts should be avoided disappears with it. We cannot say whether a particular human impact is good or bad until we know whether it will offset or exacerbate natural temperature change. Research by William Ruddiman at the University of Virginia suggests that human activity may have already staved off the next ice age. Others, studying long term sunspot cycles, predict that solar activity is headed for a downswing, in which case we would want still more anthropogenic warming.

More generally, in the absence of any clear idea where natural temperature change is headed, it makes no sense to pay a high price in order to change the level of human impact one way or another. Would you pay a high price to gamble on how warmly you will want to be dressed next Thanksgiving? The fact is, we don't know how warm a jacket of greenhouse gases we will want the Earth to be wearing ten or twenty generations from now. The best thing mankind can do is advance scientifically, economically technologically as fast as possible in order best to be able to deal with whatever theats await.

Want more? See my chapter in World Ahead Publishing's recent book: Thank You President Bush. (I thank him for rejecting Kyoto. That man saved our future, when Gore, a truly radical global warming alarmist, would have chopped our legs off.)