Bryan Caplan  

The Moral Case for Fossil Fuels: We Can Live With Warming

PRINT
Cotton, slavery and capitalism... John Cochrane on modern macro...
Rolling Stone (ahem) includes Alex Epstein, author of The Moral Case for Fossil Fuels, on its list of top "global warming deniers."  Epstein:
[T]hose who dispute catastrophic global warming are accused of denying the greenhouse effect and global warming. I experienced this in 2013 when I woke up to find myself named to Rolling Stone's Top 10 list of "Global Warming's Denier Elite" --in which they cited three articles of mine, each of which explained that CO2 has a warming effect!
How can anyone believe in anthropogenic global warming yet continue to enthuse over fossil fuels?  It's a question of magnitudes, of course.  Massive warming is deadly; modest warming is fine.  Epstein thinks the magnitude of warming has been - and will remain - modest.  Which brings us to the obvious question: Why should anyone go with his judgment, rather than the scientific consensus?

There was a time, Epstein admits, that he didn't take this challenge seriously. 
But there was so much going on in discussions of global warming , I didn't know how to decide where the evidence lay. I would hear different sides say different things about sea levels, polar bears, wildfires, droughts, hurricanes, temperature increases, what was and wasn't caused by global warming, and on and on.

With such a mess to work with, I - like most, I think - tended to side with the scientists or commentators whose conclusions were more congenial to me. I will admit to reiterating the arguments of skeptics of of catastrophic global warming with the lack of rigor I think is extremely common among believers. But I didn't do this for long. I acknowledged that I didn't really know what to think, and the idea that we might be making the Earth fundamentally uninhabitable scared me.
Most us, myself included, are in the same epistemic boat.  I'm not qualified to evaluate Epstein's main claims about the magnitude of global warming.  But he expresses his main claims so clearly that experts shouldn't find it hard to confirm or deny them.  Two claims in particular stand out.

Claim #1: While hard experimental evidence confirms a greenhouse effect, this effect is logarithmic; increasing CO2 warms at a decreasing rate:
While I've met thousands of students who think the greenhouse effect of CO2 is a mortal threat, I can't think of ten who could tell me what kind of effect it is. Even "experts" often don't know, particularly those of us who focus on the human-impact side of things. One internationally renowned scholar I spoke to recently was telling me about how disastrous the greenhouse effect was , and I asked her what kind of function it was. She didn't know. What I told her didn't give her pause, but I think it should have.

As the following illustration shows, the greenhouse effect of CO2 is an extreme diminishing effect--a logarithmically decreasing effect. This is how the function looks when measured in a laboratory.
Figure 4.1: The Decelerating, Logarithmic Greenhouse Effect
epstein1.jpg

Claim #2: Complex interactions between this logarithmic greenhouse effect and other factors could generate a lot more warming, but this is not based on hard experimental evidence.  The only way to judge these more complex climate models is against observational data.  What we've learned over the past few decades is that these models systematically overstate warming:
Here's the summary of what has actually happened-- a summary that nearly every climate scientist would have to agree with. Since the industrial revolution, we've increased CO2 in the atmosphere from .03 percent to .04 percent, and temperatures have gone up less than a degree Celsius, a rate of increase that has occurred at many points in history. Few deny that during the last fifteen-plus years, the time of record and accelerating emissions, there has been little to no warming-- and the models failed to predict that. By contrast, if one assumed that CO2 in the atmosphere had no major positive feedbacks, and just warmed the atmosphere in accordance with the greenhouse effect, this mild warming is pretty much what one would get.

Thus every prediction of drastic future consequences is based on speculative models that have failed to predict the climate trend so far and that speculate a radically different trend than what has actually happened in the last thirty to eighty years of emitting substantial amounts of CO2.
Which brings us to the single most striking figure in the whole book.

Figure 4.3: Climate Prediction Models That Can't Predict Climate
epstein2.jpg
[larger version]

My question for experts: Is there anything seriously wrong with this figure?  In particular, is it really true that virtually every major climate model overpredicts global temperature?  I'm genuinely curious, but I insist on a straight answer.




COMMENTS (32 to date)
Hana writes:

It is true. I regularly visit wattsupwiththat.com blog led by Anthony Watts. A hotbed of deniers, the writers and vast majority of commenters critically review the data. The failure of the climate models has been a noted feature of the blog for years. I defer to information there.

On a secondary note, the positive externalities of higher CO2 are regularly glossed over. Higher yields, lower water usage and more rigorous growth are all by-products of elevated CO2. (There is a reason those CO2 units have been featured in High Times for years). An interesting source for this information is CO2science.org. Under the education heading are the sources for large numbers of studies showing the higher growth and yield rates resulted from elevated CO2. There are, of course, some plants that are negatively impacted by higher CO2, those studies are listed as well.

mucgoo writes:

I thought the plateau for the past 15 years was relatively well known.
Here's an economist piece one one of many explanations
link

[Broken url and html code fixed. --Econlib Ed.]

jb writes:

Ideological Turing Test time:


1) More recent models show less warming as our understanding of the climate grows. By choosing older, less mature models, he is making the discrepancy larger than it would be if he chose the more recent models.

2) By choosing which temperature dataset to use, you can make the models look worse - for example, if the models use GISTemp or GHCN-M/ERSST surface measurements as the baseline, and the "real world" line on that graph uses, for example, the RSS satellite measurements, you'll see a big discrepancy, because the RSS satellite measurements show a lot less warming than others.

If you were to use GISTemp in your comparison to the models, the "real world" rise of 0.5 C since 1979 would be much, much closer to the models.

3) Volcanic activity has a dampening effect on warming that is impossible to predict with these models. Assuming the volcanic activity reduces back to historical frequency, the sunlight-reflecting particles diminish, the warming will increase rapidly

4) There are a number of legitimate ways that the temperatures associated with increases in CO2 can be compounded by other effects - melting ice exposes dark water and/or dark land, which absorbs more sunlight, instead of reflecting it back into space. Higher temperatures mean more water vapor, which means more clouds, and we believe that more clouds contribute to more warming. There's also evidence that large amounts of methane are trapped in tundra, and held in check by very cold water, and if that tundra melts and that water warms, we'll see a lot more more methane in the atmosphere, which has a very potent warming effect.

In summary, while the models have over-predicted the temperature rise, it isn't as much of an over-prediction as shown on this graph. While the rate of warming may not be quite as high as we expected, the change in albedo of melting ice and the potential for trapped methane to be released are not (or should not be) controversial. There are good reasons to believe that CO2's logarithmic effect is not the end of the story, and we shouldn't assume the warming will continue to be modest.


Ok. My personal lukewarmist response to my Ideological Turing Test persona:

First, the most fundamental flaw with the "runaway warming" theory is that it was warmer than it is now a few thousand years ago. If the runaway warming was going to happen because of melting ice and tundra, it would have happened then.

Second, the land temperature records are much more sparse, and far more subject to local "interference" than the satellite record. (and the ocean temperatures are even more sparse). Satellite measurements are much more complete. If the satellite record showed more warming than the land record, there's no doubt in my mind that the two sides would reverse their opinion on which was more accurate, and come up with semi-hilarious rationalizations of the switch. Personally, I think the satellite measurements seem more legitimate, and I'm prepared to accept that if they show significant warmth in the near future, I'll need to accept that global warming is more severe than I think.

Third, there is as of yet no solid understanding of the overall impact of increased cloud cover: more warming due to insulation, less warming due to increased reflectivity, no change). IMO, this makes the whole debate somewhat ludicrous - if one of the most central aspects of the climate can't be understood, yet alone modeled, it is perfectly legitimate to be skeptical of the overall efficacy of the models. I suspect (and this is just my ignorant opinion) that clouds are a negative feedback, because that's the only non-miraculous way the warming 4000 years ago would have stopped.

[comment edited per email with commenter--Econlib Ed.]

Matt B writes:

Mucgoo:

I think the real issue isn't where the warming went, but why none of the models predicted that the warming would go into the ocean. The value of the models is their supposed ability to predict the future, but they all failed to do so. It was only after the fact that a huge investigation went on to discover where the predicted warming for the last 15 years went.

James Broughel writes:

[Comment removed pending confirmation of email address. Email the webmaster@econlib.org to request restoring this comment. A valid email address is required to post comments on EconLog and EconTalk.--Econlib Ed.]

Tom West writes:

The risk of catastrophic (i.e. Venus redux) global warming is a tail risk. Probably very small, but definitely not zero.

However, ignoring that, the mild effects of global warming are only mild when you have at least a small amount of robustness in the system to adapt.

For desperately poor countries small changes in climate often produce catastrophic results because there is no margin for error. As well, these tend to put stressors on the system that also result in disease and war, just adding to the toll.

Luckily for those of us who want all the benefits of fossil fuels, but aren't in the "sucks to be them" category, the effects of global warming are likely well mixed with the natural variations that already produce catastrophe, so we can pretend fossil fuel usage isn't costing huge loss of life elsewhere.

And, to be honest, if we were able to prove exactly how much AGW costs the desperately poor countries, I suspect it would just move a lot more people into the "sucks to be them" category. At least this way we get some ameliorative efforts by Western populaces.

Brian writes:

"My question for experts: Is there anything seriously wrong with this figure? In particular, is it really true that virtually every major climate model overpredicts global temperature?"

Bryan,

The graph is correct--all models run too hot--but note that the temperatures and models are describing not the surface but the mid-troposphere, where the discrepancies are larger. Of course, humans should be interested mainly in the surface temps, so one can debate the relevance of the graph. The two graphs can be seen side-by-side herehttp://wattsupwiththat.com/2014/08/28/quote-of-the-week-reality-is-in-the-eye-of-the-beholder.

The reason why the mid-troposphere models are worse is interesting in its own right. Climate models predict a tropical hot spot in the troposphere due to water feedback. Not only should more water evaporate in a warming world and dump its latent heat in the atmosphere, but it should also trap more long-wave (IR) radiation and contribute to overall warming. Climate models predict that the mid-troposphere in the tropics should warm even faster than the surface because of these effects. Yet, THE DATA SHOW NO EVIDENCE OF THIS HOTSPOT. The discrepancy suggests that the models fail to simulate the water feedback correctly, but it is precisely this feedback that is responsible for claims of CAGW for surface temperatures. If the model feedbacks are wrong, as they appear to be, surface warming is likely to remain modest (less than 2 C for a doubling of CO2). Even the IPCC admits that this warming range provides a net positive economic gain. In other words, we should welcome warming of this magnitude.

It should also be noted, however, that Epstein appear to misunderstand the significance of claim #1. The warming response to increased CO2 is indeed logarithmic, but this means that for exponentially increasing CO2 levels, the temperature goes up linearly. The effect, so far as temperature is concerned, is not "decreasing" or in any way limited, so it can't be used to argue that warming effects are small or unconcerning. The shape of the function itself has no bearing on whether the problem is serious or not.

All in all, though, comparisons of models with data as well as temperature data analysis to determine climate sensitivity without models increasingly suggest that climate impacts of CO2 will be much less than previously thought (by about a factor of 2--1.6 C instead of 3 C for doubling) and that there is really no problem to worry about at all, since the warming is likely to remain beneficial.

Note: My attempt at the link above is not working, so here is the url:

http://wattsupwiththat.com/2014/08/28/quote-of-the-week-reality-is-in-the-eye-of-the-beholder/

[html for link fixed--Econlib Ed.]

Derek Lowe writes:

mucgoo, the plateau may now be "relatively well known", but only recently does it seem to have moved into "everybody knows that" territory. Just a year or two ago, talking about it was taken as a sign that you were a climate-change-denying yahoo. For examples, see Phil Plait's blog at Slate and this post at Climatewire, at Scientific American.

The range of those responses is well captured by those two posts. Plait's is exasperated by the idea of there still being Neanderthal shills who could even think such a thing, while the Climatewire post makes a good point about surface temperatures versus ocean temperatures. But it does seem to be true that the surface temperature predictions seem to be off.

mico writes:

That particular graph looks cherry-picked; notice for instance that there is essentially no warming at all in the entire time series when there has been surface warming throughout much of that period. I'm not expert in this field but I guess it's because he chose a graph depicting temperatures in the mid-troposphere.

The trend, however, is correct. In the decade or two for which we have forward predictions by computer models, rather than just back testing, the models have overpredicted warming. Whether that is because the trend predicted by the models is wrong or because the variance is understated is not clear, but at the very least there are not, currently, strong grounds for confidence in the model predictions.

As jb mentions, the physics estimate of how much warming should result from a given percentage increase in CO2 has been decreasing as new information becomes available. Meanwhile, the sceptic community has bifurcated into "deniers", who dispute the basic physics, and "soft sceptics", who accept the basic mechanism but dispute the predicted magnitude of warming. I suspect at some point the official number and that advanced by soft sceptics will converge. Whether that occurs at a level still high enough to be worth worrying about is less clear.

On a fundamental level I would dispute that all this is really science at all. If you are asking a policy maker to make a decision based on a quantitative estimate of an effect that, on account of too short time series of data with which to compare, can't be empirically tested until after the policy will have taken effect, you are really advancing a position on faith.

What is still not clear to me is how the magnitude of economic damage varies with the magnitude of warming. At what point does it become worth doing something about?

Brian writes:

"The risk of catastrophic (i.e. Venus redux) global warming is a tail risk. Probably very small, but definitely not zero."

Tom West,

There may be some risk of CAGW, but no risk at all of anything resembling Venus. A runaway greenhouse effect is impossible on Earth.

I would add that it's not clear that poorer countries are at greater risk for "catastrophe." The only really potentially "catastrophic" effect of global warming is in the cost of displacing current industries and infrastructure, such as in a shift of agriculture. Poorer countries, which generally have less to displace, may be less affected than wealthier ones. Indeed, when we consider what poor countries really need--better access to food, more infrastructure, more access to energy--these things are likely to get better in a warming world. There's no reason not to expect the eventual eradication of poverty and the attainment of prosperity for all under warming conditions.

Brian writes:

"What is still not clear to me is how the magnitude of economic damage varies with the magnitude of warming. At what point does it become worth doing something about? "

Mico,

This question has been answered repeatedly by the IPCC and the consensus is not controversial. Warming of less than 2 C above the pre-industrial level is expected to provide a net positive economic impact. It follows that if the equilibrium climate sensitivity (ECS) for a doubling of CO2 is less than 2 C, AGW cannot be viewed as a problem at all, unless we have reason to expect that Earth will exceed the doubling level (560 ppm).

If we continue to have carbon-based fuels as our primary energy source indefinitely, we will of course exceed that level. But given that we have at least 50 years before getting there, it seems likely that technology will advance enough to create a market-based shift to less carbon-intensive approaches. In the absence of such a shift, we may have to go with some kind of CO2 capture or recovery technology, but we probably have at least 30 years before such decisions need to be made.

Tom West writes:

Okay, Venus redux was there for the rhetoric, but I don't think that our civilization would well handle every condition that the planet has already been in (including snowball). We just don't know what the tail risks are. (including doing nothing, but most accept there's more risk when mucking about with the system).

The only really potentially "catastrophic" effect of global warming is in the cost of displacing current industries and infrastructure, such as in a shift of agriculture.

Um, agriculture is kinda important. A few hundred thousand people starving to death is catastrophic - no quotes. If a foreign reactor blew up, and killed 100,000 Americans, we'd probably not consider that simply "that's just the cost for them getting cheap, convenient energy".

There's no reason not to expect the eventual eradication of poverty and the attainment of prosperity for all under warming conditions.

Nice for future generations. However, its not usually morally acceptable to sacrifice millions now for a "future prosperity" that may (or may not) arrive.

Dan Carroll writes:

Tom West

Um, agriculture is kinda important. A few hundred thousand people starving to death is catastrophic - no quotes.

The agricultural shifts that would likely occur are shifts in the types of crops that grow better in the new climates, not a shift to less food. There is actually no empirical or theoretical reason to conclude that the rate of starvation will increase under AGW; it may in fact go down. Indeed, the local impacts of AGW are extremely difficult to predict - some areas will get less rainfall, some more rainfall, some hotter temps, some cooler temps, some more severe storms, some less, etc. The countries with more rigid economic systems will be less able to adapt, while the more flexible ones will adapt better. Poor countries would likely get hit in two ways: poor countries tend to have inflexible top-down economic management, and if ocean levels rise, flooding of low-lying areas will pose a problem.

Brian writes:

"Um, agriculture is kinda important. A few hundred thousand people starving to death is catastrophic - no quotes."

Tom West,

Yes, that would be catastrophic, but note that I said a "shift" in agriculture, not a reduction. In a warming world, Earth is likely to produce MORE food, not less. We already produce more food per capita than the world population needs to be well fed. Hunger and starvation is almost always a politically caused problem, not an agriculturally caused one.


"Nice for future generations."

Yes, and nice for current ones too. The point is that the current level of warming is in the range where the net economic benefits are positive, so warming is not inhibiting us from making progress with eradicating poverty. If anything, it's an advantage. What we do with that advantage is up to us.

Brian writes:

I had previously submitted a comment that appears to have been eaten by the system. Since I don't have time to type it all in again, I am including the link that shows the comparison between the mid-troposphere data above and the model-to-measured surface data. The graph above, while not cherry picked, has a larger discrepancy than the more relevant surface data.

Dave Tufte writes:

I made a similar point to your Claim 2 a few years back.

I proposed back-of-the-napkin estimates of elasticities for climate change: first in "Elasticity of Carbon Dioxide and Economic Growth", and then in "Elasticity of Temperature and Carbon Dioxide".

What did I find? That both relationships are ridiculously inelastic: far more inelastic than, say, smoking. This makes the end goal that some have — of reducing atmospheric carbon dioxide by limiting economic growth — pretty close to impossible.

The analogy I gave is that what if a smoker's cigarettes in turn were hooked on smoking little tiny cigarettes. And the proposed method of dealing with the health risk to the human smoker was to go after the little tiny cigarettes smoked by something like Doonesbury's Mr. Butts.

But, of course, when you put it that way, the whole thing sounds silly, rather than **important** policy work.

FYI: Interested readers might also consider global warming policy in the context of the following post by Scott Sumner riffing Cochrane's view of monetary policy. The policymakers are overly concerned with "what to do now" rather than on what their goal might be, or whether they even have one.

Jeff writes:
As the following illustration shows, the greenhouse effect of CO2 is an extreme diminishing effect--a logarithmically decreasing effect. This is how the function looks when measured in a laboratory.

That seems like a really basic and fundamental question. Is it true? If so, how is it that it never comes up in discussions of climate change?

MikeP writes:

A few hundred thousand people starving to death is catastrophic - no quotes.

A few hundred thousand people die every year from causes that are primarily due to poverty.

The sooner these people can lift themselves from poverty the better. And cheap energy is one of the most important paths to that better future for them.

Cutting off the bottom rungs of the energy ladder for developing societies in order to assuage the sensibilities of developed societies is monstrous. The world's truly poor simply do not have the luxury of worrying about global warming's effects on the much richer planet a century from now.

Josiah writes:

One big problem with determining the likelihood of catastrophic global warming is that what counts as "catastrophic" is vague and undefined. In this thread, for example, you have people disagreeing about whether hundreds of thousands of people dying from global warming would be "catastrophic" or not.

Brian writes:

"That seems like a really basic and fundamental question. Is it true? If so, how is it that it never comes up in discussions of climate change? "

Jeff,

The logarithmic character of the response is well known and has been forever. See my comment at 11:14 AM for relevant commentary.

Ben Kalafut writes:

"By contrast, if one assumed that CO2 in the atmosphere had no major positive feedbacks, and just warmed the atmosphere in accordance with the greenhouse effect, this mild warming is pretty much what one would get. "

Did anybody else spot the mistake here? Epstein is trying to pass off something qualitatively plausible (and then only if we don't look too hard at what the ocean is doing, what solar forcings have looked like, etc.) as the explanation or the most likely explanation for a measured effect. He makes no case for the no forcings position among the sea of alternatives--certainly doesn't compare it here to what working scientists have actually offered--but he does this so casually he pulled the wool over Prof. Caplan's eyes and I suspect many others without them even realizing it just got a whole lot darker.

Let's set aside the problems with setting up "no warming, just random variability" as a privileged null hypothesis. (The correct approach is to look for a statistically significant change in the trend.) Let's just look at the physics.

(1) The warming trend since the late '90s was underestimated. See http://onlinelibrary.wiley.com/doi/10.1002/qj.2297/abstract

(2) More heat than the older models expected has been going into the oceans. See http://onlinelibrary.wiley.com/doi/10.1002/2014GL061456/abstract
or for a popular introduction to two different studies,
http://www.nature.com/news/atlantic-ocean-key-to-global-warming-pause-1.15755
and links therein.

It looks like Epstein, yet again, is pulling a fast one, and that he fooled Bryan Caplan. However unreliable Rolling Stone is these days, Epstein's place in the top-10 list of liars is justified and secure.

Chad writes:

Brian wrote:

The warming response to increased CO2 is indeed logarithmic, but this means that for exponentially increasing CO2 levels, the temperature goes up linearly.

Brian, over what time frame are CO2 levels are rising exponentially? How often are CO2 levels going to double?

My understanding is that CO2 emissions are rising, but while they may be rising faster than linear, they're not anywhere near fast enough, nor accelerating enough, to be doubling CO2 concentrations at a constant pace and thereby produce a constant temperature rise. Do you have data that shows otherwise?

Ben writes:

The logarithmic character of the curve seems rather irrelevant, since the sharply curved portion of the curve is below preindustrial levels. A straight line from preindustrial levels through to 2x preindustrial levels is a very close approximation to the shape of the curve in that range. Yes, if we get up to 3x or 4x preindustrial levels the shape of the curve will start to matter – although still not very much, in the big picture. The fact that Epstein makes such a big deal about it just goes to show that he's spouting denialist propaganda rather than making solid points.

The "single most striking figure in the whole book" seems adequately debunked in what has been written above; it isn't even showing surface temperatures! Cherry-picked and misleading. That ought to give you an idea of the company you're keeping, and the misinformation you're promoting, Bryan.

Yes, the present "pause" in warming is an interesting phenomenon that is not well understood and was not well predicted by models. The article linked to before (http://www.scientificamerican.com/article/has-global-warming-paused/) seems like a good summary of the state of affairs. As the article says, researchers "link this change to shifts in winds, especially in the Pacific Ocean, related to decadal weather patterns in the Pacific". That means it is probably a temporary state of affairs; when those decadal weather patterns oscillate back again, the "pause" will likely end. In any case, as the article makes clear, global warming continues unabated; it's just that less of the heat is going into the atmosphere than expected. That is not necessarily a good thing; it means that sea level rise will actually be faster than predicted, and coral reef ecosystems might die even sooner than expected.

Notice the difference in reactions between the scientists and the denialists. The scientists say "that's interesting, the models aren't performing well, I wonder why?" And then they come up with hypotheses, and start trying to improve their models. The denialists, on the other hand, just crow about how the models are worthless and therefore we don't have anything to worry about – seeming not to realize that the fact that we know even less about the dynamics of the climate system than we thought we did is hardly a reason to feel comfortable with the fact that we're perturbing it.

As I've emphasized before, Bryan: if you really want to get your "epistemic boat" to dry land, why don't you actually start talking to climate scientists? I imagine you'd have no trouble finding several who would be happy to talk with a prominent economist about their research. But instead you ask your questions here on your blog, which is pretty much a denialist echo chamber.

Sam writes:

I'd like to add my voice to Chad's in asking Brian for more information about what's known regarding the likely path of future CO2 levels.

To my eye the data clearly show that the period from 1750-2000 included some amount of exponential growth. It's also safe to bet that world GDP will continue to grow exponentially. And CO2 emissions are clearly correlated with GDP on some timescales. At the same time, So the question is what's known or expected to be true about CO2 emissions as a function of GDP (or GDP per capita if you prefer -- either way it's the dominant long-term exponential component of GDP growth).

A quick look on the internet indicates to me that E(CO2 | GDP) across countries is almost certainly sublinear in GDP (although the initial growth rate is quite large). This is quite commonsensical. There's not really enough data to distinguish sublinear-but-polynomial (e.g. square root) from logarithmic. But if we assume it's logarithmic, then the long-run rate of change of CO2 emissions can be assumed at worst polynomial. We care about the integral of this polynomial over time (atmospheric concentration), but that's still polynomial.

Hence, in principle, on very long time scales the logarithmic nature of the greenhouse effect (of which like many non-climate-scientists I was unaware) might in fact mean that the temperature will increase sublinearly in the long run.

Of course this depends crucially on E(CO2 | GDP), which I think is known as "carbon intensity" and which seems to be basically completely not understood. I cannot rule out convincingly that it is a polynomial function of GDP, but then again, nor can I prove it is a monotonically increasing function of GDP.

Finally, there's the question of whether of this matters at all, which is probably what mainstream climate scientists would deny. Who cares about the long term temperature trend if catastrophic warming sets in well before the effects of exponentially increasing GDP are relevant? That is a strong argument, but again it is quite difficult to bring a quantitative approach to bear on this question. The IPCC tried to do so using "Scenario analysis". This is an extremely complicated business, and thus amounts to another layer of modelling in the CAGW debate.

Mark Bahner writes:
Brian, over what time frame are CO2 levels are rising exponentially?

In the time period from 1959 to 2014, CO2 levels rose super-exponentially:

Global CO2 atmospheric concentration changes

An exponential increase means increase at a constant percentage rate (e.g., 0.2 percent per year, or 0.5 percent per year, or 1 percent per year). If one puts all the data in the table on that webpage into a spreadsheet and takes the ten-year rolling average CO2, one will get about 0.25 percent per year in the 1960s, up to about 0.53 percent per year.

So we'll all die, right? Well, even at 1 percent per year, it takes roughly 70 years to double. Even if that increased to 0.7% per year, it would take roughly 100 years to double.

How often are CO2 levels going to double?

Opinions differ. But *good* opinions ;-) say doubling from the pre-industrial concentration of 280 ppm to 560 ppm has a 40-80% chance of happening. And doubling again (to 1120 ppm) is virtually impossible.

My understanding is that CO2 emissions are rising, but while they may be rising faster than linear, they're not anywhere near fast enough, nor accelerating enough, to be doubling CO2 concentrations at a constant pace and thereby produce a constant temperature rise. Do you have data that shows otherwise?

No, you're misunderstanding. The rate of annual percentage increase is actually increasing. But it's presently at about 0.5 percent per year, which means a doubling roughly every 140 years. And as I noted, even if it increased to 0.7% per year, that would be a doubling every 100 years.

Back in 2005, I predicted a 50% likelihood of a concentration of 558 ppm (almost exactly double the pre-industrial concentration) in the year 2100. I stand by that prediction:

Projections of global warming

P.S. The prediction of 1.2 degrees Celsius warming from 1990 to 2100 still looks pretty good, too. ;-)

mico writes:

Brian,

I was interested at what point the net damage due to CO2 emission outweighs the benefits of CO2 emission, not at what point it begins to cause just any net damage.

The reason I am so sceptical on this point is that it occurs at a weak link between the scientific and economic fields. Climate scientists have no expertise to assess economic damage and I doubt (though may be wrong) that economists' models take into account the different gradations of physical effects that may occur at different levels of warming.

And this impression is to some extent confirmed by listening to the media. A lot of emphasis for instance is placed on the likelihood that global warming would reduce land values in the third world. This is presumably intended to be viewed as a worst case: people who are already poor are made even poorer! But land value in the third world is a negligible fraction of global wealth; even from a charitable point of view the third world would be better off if we just kept out industry and gave them some money. That's before we get to the truly vague interdisciplinary effects like apparent increase in frequency of wars.

I find it much harder to understand what major economic damage could be inflicted on the developed world, where agriculture is a tiny fraction of economic output. Rising sea levels may be one, but I imagine it is much cheaper to build defenses than to end CO2 emissions.

Josiah writes:

mico,

Richard Tol has a survey of studies that have looked at the probable economic impacts of different amounts of warming, which you might find interesting.

Josiah writes:

Bryan begins the post by noting a seemingly embarrassing oversight by Rolling Stone. According to Epstein, he was "named to Rolling Stone's Top 10 list of "Global Warming's Denier Elite" --in which they cited three articles of mine, each of which explained that CO2 has a warming effect!"

If you look at the Rolling Stone piece, they actually cite two lectures and one article, not three articles. But leave that aside. I don't know exactly what Epstein said in this lectures, but I did track down the one article RS cited. Here is the closest thing I was able to find in the article to a statement that CO2 has a warming effect:

Our cultural discussion on “climate change” fixates on whether or not fossil fuels impact the climate. Of course they do—everything does

Is that an explanation that CO2 has a warming effect? No. Someone who denied any warming effect from CO2 could equally well agree to the statement.

If this level of slipperiness is typical of the book as a whole, then Bryan had better be wary.

Wojtek Grabski writes:

There is a much more fundamental physical problem with the argument for CAGW. It is:

Proponents of the theory basically claim that the atmosphere is a system at an unstable energy minimum -- maybe even an inflection point. According to their claims, it takes but a tiny change in the composition of the atmosphere ( on the order of 500 ppm) to leave this minimum and trigger a state change, i.e., enter a region of exponentially increasing sensitivity to changes, or cause an outright runaway effect. This seems ludicrous on its face. To wit, if the system were really that unstable, then it would have already left this state a long time ago. Especially considering it's already been in states neighboring this one, and no sudden changes occurred.

An alternative explanation; one might call it the "weak CAGW theorem", claims basically that the atmosphere is at a local minimum, i.e., the system is locally stable, but a large enough change in CO2 can knock it out into a change of state. Same argument applies. If changes on the order of 500/1,000,000ths in atmospheric composition can knock the system out of this state; then the Earth would already resemble venus, or whatever other ridiculous final steady state is being posited.

It seems obvious to me that the atmosphere, due to its long term stability, is almost without question in a very deep energy minimum. That is, changes in composition need to be significant to cause changes of state -- if those are even possible at all. Note that even cataclysmic events have not managed to permanently prevent it from returning to a state that is able to support tons of life.

Tom West writes:

MikeP - A few hundred thousand people die every year from causes that are primarily due to poverty.

Agreed, but there's some difference between whether our actions are directly contributing to those deaths and deaths in which we have no hand in. Death is death, but as reactions to the trolley problem clearly indicate, the how and why clearly matter.

Cutting off the bottom rungs of the energy ladder for developing societies in order to assuage the sensibilities of developed societies is monstrous.

I suspect that most who feel "fossil fuels are evil" would feel we're morally on the side of the angels if we restricted our fossil fuel usage, even if others did not.

Daublin writes:

Bryan, if you are up for some homework, you might want to dive into the latest IPCC report.

My understanding is that it agrees with you, but the blogosphere hasn't realized it yet. The latest IPCC has backed away from the models it used the push, and has conceded that their predictions were way high.

FWIW I think the other arguments about CO2 controls are better, though.
- CO2 controls are definitely very harmful to humans if enacted to a degree that it will actually lower atmospheric CO2.
- A hundred years is a long time to prepare, especially during a time of technology growth.
- We don't know if greater CO2 will be a harm at all. Much evidence suggests that it would be an improvement.

stargirl writes:

David Friedman posts very convincing evidence that essentially all the major climate models endorsed by the IPCC have over predicted warming. I don't about the record of organizations other than the IPCC.

Comments for this entry have been closed
Return to top