The Thought Experiment That First Made Me A Climate Skeptic

http://www.forbes.com/sites/warrenmeyer/2014/02/26/the-thought-experiment-that-first-made-me-a-climate-skeptic/

 The Thought Experiment That First Made Me A Climate Skeptic

Preface
Last night, the accumulated years of being called an evil-Koch-funded-anti-science-tobacco-lawyer-Holocaust-Denier finally caught up with me.  I wrote something like 3000 words of indignation about climate alarmists corrupting the very definition of science by declaring their work “settled”, answering difficult scientific questions with the equivalent of voting, and telling everyone the way to be pro-science is to listen to self-designated authorities and shut up.  I looked at the draft this morning and while I agreedwith everything written, I decided not to publish a whiny ode of victimization.  There are plenty of those floating around already.

And then, out of the blue, I received an email from a stranger.  Last year I had helped to sponsor a proposal to legalize gay marriage in Arizona.  I was doing some outreach to folks in the libertarian community who had no problem with gay marriage (after all, they are libertarians) but were concerned that marriage licensing should not be a government activity at all and were therefore lukewarm about our proposition.  I suppose I could have called them bigots, or homophobic, or in the pay of Big Hetero — but instead I gathered and presented data on the numberof different laws, such as inheritance, where rights and privileges were tied to marriage.  I argued that the government was already deeply involved with marriage, and fairness therefore demanded that more people have access to these rights and privileges.  Just yesterday I had a reader send me an email that said, simply, “you changed my mind on gay marriage.”  It made my day.  If only climate discussion could work this way.
So I decided the right way to drive change in the climate debate is not to rant about it but instead to continue to model what I consider good behavior — fact-based discussion and a recognition that reasonable people can disagree without that disagreement implying one or the other has evil intentions or is mean-spirited.
A Thought Experiment on Climate Models
This analysis was originally published about 8 years ago, and there is no longer an online version.  So for fun, I thought I would reproduce my original thought experiment on climate models that led me to the climate dark side.  The charts are from a number of years ago, and I still stand by them, but two things have changed since then:  1)  The forecasts are from the IPCC 4th edition — since then new 5th edition forecasts have been released; 2)  The current atmospheric CO2 ppm has gone up 10-15 points, sitting near an annual mean of 397 or so.  These do not affect the analysis in the least so I have left the charts as I originally produced them.
We need to begin with a definition.  You will often see a term called “climate sensitivity”**.  Typically this is the number of degrees Celsius Earth’s average near-surface atmospheric temperature would rise given a doubling of  atmospheric CO2 concentrations.   So, for example, if CO2 is at 400 ppm and then rises to 800 ppm, and the climate sensitivity were 3, we might expect average temperatures to rise by 3 degrees C.   Greenhouse warming is thought to be a logarithmic or diminishing return effect, so that the 400 point rise from 400 to 800 would cause the same warming as the 800 point rise from 800 to 1600.  Another way to put this is that each ppm of CO2 added has incrementally less effect than the last.
One of my first frustrations with the IPCC reports was that they have thousands of graphs of temperature over time, based on embedded CO2 production forecasts.  These are fine as far as they go, but they complicate analysis of a forecast because they include many embedded assumptions about economic activity and thus CO2 levels.  I understand the need to consider time-delayed responses, but I have always thought it would be far more useful to leave out the complication of dueling economic and CO2 forecasts and just look at temperature vs. CO2 level.  Unfortunately, the IPCC never produces such a chart, at least that I can find.
But it did provide a formula, from Michael Mann in 1998, giving a hypothesized relationship between CO2 and temperature leaving aside all system feedback effects.  The formula and its graph are below (the selection of the zero temperature point is arbitrary so I set it to zero at the then-current CO2 level).

I was thus able to create my own version of a temperature vs. CO2 concentration graph I felt the IPCC lacked.  So far so good, but anyone familiar with climate forecasts will note an immediate problem.  The implied sensitivity to CO2 in this chart of 1.3C is way below any recent IPCC forecast.  But Dr. Mann, whose formula we are using,  is a leading advocate of high sensitivity estimates.  What gives?
This is one of the most important points of the climate debate that is almost never explained in the mainstream media:  that catastrophic warming forecasts are actually a chain of two theories.  In the first, a doubling of CO2 raises temperatures directly via the greenhouse effect by 1.3C.  Then, positive feedbacks in the climate system multiply this warming 3-8 times.  As a result, most of the warming in catastrophic forecasts is not from greenhouse gas theory, but from a second independent theory that the Earth’s climate system is dominated by very strong net positive feedbacks.
So, based on various estimates for these positive feedbacks, we can now add lines to our chart that correspond to the 4th IPPC’s mean sensitivity estimate of 3.4C, high range sensitivity of 5.4C, and top alarmist estimates of 10C.

So here is the thought experiment I did years ago.  We did not just start producing CO2 in the last few years –we have done so for over a hundred years.  In fact, we have an estimate for pre-industrial CO2 levels of around 270ppm, meaning that at close to 400ppm today we have already increased CO2 by about 130 ppm, or nearly half a doubling.  If sensitivity forecasts work going forward, they should work just as well going backwards.  So I re-scaled my chart so I could look backwards:

Then it was simple math to project the forecast lines backwards.  Since we scaled current temperatures to zero, then the values where the forecast lines hit the 270 (pre-industrial) ppm mark represent how much warming we should have seen over the last 100-150 years given each sensitivity level.  In other words, if the sensitivity is really 10C (the red line), we should have seen over 4C of manmade warming over the last 150 years.
We know that there has been about 0.7C of warming since the middle of the 19th century.   Looking at our chart, this is hard to square with the high-sensitivity IPCC forecasts.   Even at the IPCC median sensitivity number, we should have seen more than twice the warming we have had historically.  Higher sensitivities, including those of the variety used by folks like Al Gore and Bill McKibben, would require that we have seen 2-4C of historic warming.
I think any physical scientist should be extremely skeptical that a long-term stable system is dominated by positive feedback.   Systems dominated by positive feedback — and we are talking about incredibly high implied feedback percentages to get to these catastrophic forecasts — don’t tend to be very stable, but it is Michael Mann himself who has argued over and over with his hockey stick chart that past temperatures have only varied in very narrow ranges for thousands of years.  Not the behavior one would expect of a system dominated by strong positive feedbacks.
To me, this thought experiment demonstrated that it was more likely that net climate feedbacks were zero or even negative (if only half of past warming was due to man, and half due to nature, it would imply a sensitivity around 0.7C).   In either case, the resultant warming would be far from catastrophic.  To believe the IPCC forecasts, one would have to believe there were either really long time delays, or natural and manmade cooling factors off-setting the warming.  These have all been debated and I won’t go into them today, but I didn’t find the higher forecasts of 5-10C to be at all credible.
Of course, none of this constitutes proof one way or the other.   This is a sanity check, the type of thing that good modelers perform all the time.  But if catastrophic warming were really so settled that it constitutes a form of secular heresy to question it, then analyses like this should be far less equivocal.  While the proposition that CO2 acts as a greenhouse gas does border on settled science, there is by no means any consensus on the magnitude, or even the sign, of the feedbacks that account for the majority of the warming in catastrophic forecasts.  This is the heart of the disconnect in the climate debate and is discussed at greater length here (100,000 views and counting).
Postscript:  Based on this analysis done about 8 years ago, we would expect these climate models to over-predict warming.  And so far, they have.

One of the ironies of these forecast misses is the arguments that catastrophic warming supporters are trotting out to explain the divergence.  Modelers are arguing that cool phases of ocean cycles and a decreased solar cycle are offsetting some of the warming.  Which may be, but for years climate skeptics argued that in prior years warm phases in ocean cycles and increased solar cycles explained some of past warming, and we were called heretics as the “consensus” was declared that these natural effects were too small to affect global temperatures substantially.

The other argument is that the extra heat is skipping past the atmosphere and upper ocean and is hiding in the deep ocean.  The mechanics of this are odd but not impossible, but again an irony as skeptics have argued for years that ocean heat content is a better measure of CO2-driven warming than atmospheric temperature.  Heat may be hiding in the deep oceans, but they have not warmed nearly enough to be consistent with high-sensitivity climate models.

** footnote:   There are also different “flavors” of climate sensitivity depending on whether we are talking about quick transient response or longer term 10-15 year effects.  We will leave that aside for this analysis, since it is meant as a simple thought experiment.