Global Warming: You can't verify climate models with more suppositions

Over on the newsgroups, folk are discussing a rather sad article from ScienceDaily, called "Apparent Problem With Global Warming Climate Models Resolved" [2].

Apparently the folk at ScienceDaily, as well as a large fraction of the general public, need some basic lessons in how real science, the science that increases our understanding and helps us make sense of the world, works. Here's a textbook example of how not to do it:

Yale University scientists reported that they may have resolved a controversial glitch in models of global warming: A key part of the atmosphere didn't seem to be warming as expected.

Computer models and basic principles predict atmospheric temperatures should rise slightly faster than, not lag, increases in surface temperatures. Also, the models predict the fastest warming should occur at the Tropics at an altitude between eight and 12 kilometers. However, temperature readings taken from weather balloons and satellites have, according to most analysts, shown little if any warming there compared to the surface.

By measuring changes in winds, rather than relying upon problematic temperature measurements, Robert J. Allen and Steven C. Sherwood of the Department of Geology and Geophysics at Yale estimated the atmospheric temperatures near 10 km in the Tropics rose about 0.65 degrees Celsius per decade since 1970—probably the fastest warming rate anywhere in Earth's atmosphere. The temperature increase is in line with predictions of global warming models.

“I think this puts to rest any lingering doubts that the atmosphere really has been warming up more or less as we expect, due mainly to the greenhouse effect of increasing gases like carbon dioxide,” Sherwood said.

Note that the thermometers (i.e. instruments designed to measure temperature) are said to be "problematic"—a euphemism for flat-out disagreeing with the AGW theory. So instead, they come up with yet another unverifiable model, this time claiming that wind changes are a better measure of temperature than a thermometer. Gimme a break! If you believe that, there's a nice harbour bridge in Sydney I wanna sell ya. So we have:

• Thermometers disprove AGW, so:
• hypothesise a connection between wind and temperature that shows different temperatures
• (Note that this hypothesis cannot agree with experiment because the theory is designed to differ from the raw temperature readings—the thermometers)
• on the basis on this, by its deliberate construction, false theory, claim you have corroboration of your AGW theory.

The problem, of course, is that this supposed "verification" is circular. One theory (wind tells you temperature), completely unverified, is somehow supposed to provide the evidence that removes "any lingering doubts" about another, also unverified, theory. Truly sad!

Real science works by connecting theory to observation. Two theories cannot corroborate each other in an evidence-free vacuum. Unfortunately this is a classic ploy of the AGW shysters. From now on, any time someone points out the demonstrable fact that the temperature measurements flatly contradict the AGW theory, they will be told "Ah, but that claim has been discredited." "Discredited" is a powerful word; most unwitting onlookers will be taken in by it without further ado. Bogus pseudoscience works hand-in-hand with dishonest psychological manipulation. That is what the great scientific edifice, won with the blood and courage of heroes of the enlightenment like Giordano Bruno, has degenerated into. Have we arrived at the threshold of the next dark age?

[Update] Here's another story from ScienceDaily with all the same mistakes: Changes In The Sun Are Not Causing Global Warming, New Study Shows [3]. This time, it is the Sun's influence on cosmic rays [4] that is "laid to rest" by a mere computer program. Here's what I know, as someone who has taught computer science for near thirty years: computer programs embody whatever knowledge, prejudices, and outright mistakes that their programmers put into them. Only if the assumptions are correct will the output be.