More On The "Science" Behind The Global Warming Scare

If you are still reading or watching legacy "news" sources like the New York Times, Washington Post, or one of the major TV networks, you may have the impression that all is going swimmingly in the consensus (groupthink) "science" supposedly proving that human CO2 emissions are destroying the climate.  Certainly, there continues to be unanimity among Democrat politicians that human use of energy from fossil fuels is a drastic problem that can somehow be cured by building thousands of windmills and solar panels at great cost.  See, e.g., Andrew Cuomo, Gerry Brown, and the 49 Democrat Senators who voted unanimously against the confirmation of suspected climate "skeptic" James Bridenstine to head NASA.

Lately, most of my posts on the subject of energy and climate have been devoted not so much to the science in question, but rather to the ridiculous costs and complete futility of trying to reduce CO2 emissions by means of the preferred solutions of windmills and solar panels.  See, for example, "Some Perspective On Efforts To Reduce Carbon Dioxide Emissions" on May 13, "Everyone Knows That Trying To Control The Climate By Reducing CO2 Emissions Is A Joke" on May 3, and "What's Really Happening In The World Of CO2 Emissions" on April 6. 

But still, it's fair to ask, is there any strong basis to believe the human CO2 emissions operate as some kind of direct control knob on atmospheric temperatures?  So today, let's return briefly to that subject.  

If you are an aficionado of climate science, you have undoubtedly many times seen two scientifically-related things cited in the effort to sell you on the proposition that CO2 emissions are a big problem:  (1) the fact that CO2 has been demonstrated in laboratory experiments to be a "greenhouse gas," and (2) the correlation between rising atmospheric temperatures over the past 40 years or so with rising concentration of CO2 in the air.  QED!  On the other side, the pesky skeptics point out that, despite CO2 being a known greenhouse gas, the proposition that it operates in a way to have direct statistically significant effect on world temperatures -- not in a laboratory environment, but in an enormously complex climate system filled with confounding factors like a hydrological cycle, clouds, oceans, volcanoes, and who knows what else -- is just a hypothesis in need of proof.  The minimum beginning of such proof would be a valid mathematical demonstration of a statistically significant relationship between atmospheric CO2 levels and temperatures.  Does that exist?  I've never seen it.  Have you?

Into this mix there has dropped today a new paper from James Wallace, Joseph D'Aleo and Craig Idso, titled "Comment on “Examination of space-based bulk atmospheric temperatures used in climate research” by John R. Christy et al."  The title doesn't give you much clue to the contents; but the main subject of the piece is testing the hypothesis that a statistically-significant relationship can be shown between CO2 in the atmosphere and measured global temperatures.

Before getting to the paper, just a little background.  Over the past 40 years or so, if you calculate a linear trend line through the best data on world temperatures, you get an upward-sloping line.  Similarly, if you do the same with the best data for atmospheric CO2 concentrations, you also get an upward-sloping line.  And CO2 is a greenhouse gas!  Isn't this all the proof you need?

Actually, no.  Those with any math background beyond high school (I know that this is not a lot of people) will recognize that trying to establish a statistically significant relationship between two time series by the mere fact that the two have upward-sloping linear trend lines is a mathematical fallacy.  By this logic, we could prove that climate is controlled by anything that has increased over the relevant period, from the number of snowmobiles to the level of immigration into the United States to my age.  Any serious effort to show a statistically-significant relationship between two fluctuating times series like atmospheric CO2 concentration and temperature must deal with actual temperature data; and those data show temperature levels that bounce around in a way not closely matching the CO2 data (which have basically been on a steady increase, with a smaller embedded annual cycle).  Here is the published graph of the UAH satellite temperature record from inception (1979) through the most recent month (April 2018) (Note that Dr. Christy is one of the people responsible for generating the UAH satellite temperature record):


What's immediately obvious is that it's not a smooth upward trend, but rather has periodic dramatic drops, for example in 1991-93, 1998-2000, and most recently 2016-18.  No such drops were recorded in the CO2 record.  The temperature record also shows a "pause," where temperatures did not increase, for approximately 20 years, from 1995 to 2016.  Do these periods without a relationship of CO2 concentration to temperature destroy the ability to demonstrate statistically-significant relationship between the two?  That is the question addressed by the new paper from Wallace, et al.

And the answer is, a statistically-significant relationship cannot be demonstrated.  From the abstract of the Wallace, et al., paper:

[T]he Christy et al (2018) work does not claim that this lower trend slope finding implies anything whatsoever about whether CO2 has had a statistically significant impact on the Earth’s temperature over the last 50years or so.  . . .  This Comment argues that such a claim must be addressed using appropriate mathematical methods. Such methods are used in this Comment to prove that increasing atmospheric CO2 concentrations did not have a statistically significant impact on the UAH TLT 6.0 temperature data set over the period 1979 to 2016. In fact, this structural analysis demonstrates that there was a “Pause” in temperature trend increases over the 1995 to 2016 period. This is a time period over which atmospheric CO2 concentrations increased by over 12.0%.

Go ahead and read the whole paper at the link above and see if you agree or disagree with its conclusions.  The underlying data are all available, and all methods are fully accessible for replication or refutation.  Have at it!

I would comment that a statistically-significant relationship between CO2 concentration and temperature would be a necessary but not sufficient proof of causation.  If the relationship could be shown, then the hypothesis of CO2 as the control-knob of world temperature would remain viable.  However, lacking a demonstration of a relationship, the hypothesis is refuted.  It's as simple as that.

Meanwhile, over at No Tricks Zone, they have produced a new compilation of some 81 graphs from 62 scientific papers in 2018 "undermining claims that modern era warming is climatically unusual."  Each graph shows, to some degree or other, that recent temperatures "are not unprecedented, unusual, or hockey-stick-shaped — nor do they fall outside the range of natural variability."  Go to the link and start going through the material.  There's enough there to keep you busy for days.