Since I’ve commented on errors in Tol’s work before, I thought I might briefly comment on his Correction and Update: The Economic Effects of Climate Change. He corrects the errors in his 2009 meta-analysis (thanking Bob Ward for finding a small error) and updates the analysis using some newer studies.
The new paper corrects the original figure, which showed a net benefit (considering the mean) for warming up to just over 2 degrees. The updated figure (shown below) reduces this slightly, but doesn’t change it much. However, it does illustrate – more clearly – that there’s really only one study (Tol 2002) that shows a net positive benefit (for completeness : there is a second, but it is +0.1% for a warming of 2.5 degrees).
credit : Tol (2014)
What’s maybe more interesting, is that Tol updates the analysis by including 7 newer studies. The result is shown in the figure below, and appears about the same as Grant McDermott found when he did something similar. With the new studies included, there is essentially no positive benefit for future warming. This seems like quite a significant change to me, and I’m looking forward to Matt Ridley pointing this out in his next article 🙂
credit : Tol (2014)
This update looks both interesting and quite significant. However, what I’m unsure of is how this compares to what was presented in the latest IPCC reports. What’s presented in Tol’s analysis is the welfare-equivalent income gain or loss. In WGII, the headline numbers were
With these recognized limitations, the incomplete estimates of global annual economic losses for additional temperature increases of ~2°C are between 0.2 and 2.0% of income (±1 standard deviation around the mean) (medium evidence, medium agreement). Losses are more likely than not to be greater, rather than smaller, than this range (limited evidence, high agreement).
So, this the annual cost of climate change.
In WGIII, the headline numbers were
Under these assumptions, mitigation scenarios that reach atmospheric concentrations of about 450ppm CO2eq by 2100 entail losses in global consumption—not including benefits of reduced climate change as well as co‐ benefits and adverse side‐effects of mitigation of 1% to 4% (median: 1.7%) in 2030, 2% to 6% (median: 3.4%) in 2050, and 3% to 11% (median: 4.8%) in 2100 relative to consumption in baseline scenarios that grows anywhere from 300% to more than 900% over the century. These numbers correspond to an annualized reduction of consumption growth by 0.04 to 0.14 (median: 0.06) percentage points over the century relative to annualized consumption growth in the baseline that is between 1.6% and 3% per year.
So, this is how mitigation will influence the annual growth rate : essentially, not much – so small as to almost be in the noise.
What I’m unsure of is quite how one can compare the welfare-equivalent income gain or loss presented in Tol’s meta-analysis, with what’s been presented in the WGII and WGIII reports. If anyone knows how this is defined and how it can be compared to the WGII and WGIII numbers – if it can – maybe they could explain in the comments.
...and Then There's Physics 
A few points. First, Tol’s updated graph includes two data points not included in Grant McDermott’s figure, and therefore also not included in the IPCC data that McDermott used. Second, the derived curves are quite different in the two cases, though not at small values. McDermott had a linear parameter of 0.1, and a quadratic of -0.27; while Tol reports values of -0.28 and -0.16 respectively. That means that while Tol’s “expectation” is marginally lower up to about 2.5 C, above that it is higher showing significantly less cost at higher temperatures. On the other hand, the pessimistic error margin for Tol becomes very large at higher temperatures, so it is likely to represent a greater expected loss overall at higher temperatures. As it happens, both the original and corrected values represent greater costs with increased temperatures than the corrected and updated figures so that while the latter show greater short term costs, they would represent good news in the long term at face value.
Finally, the IPCC claim that “…global annual economic losses for additional temperature increases of ~2°C are between 0.2 and 2.0% of income (±1 standard deviation around the mean)…” are based on the distribution of data at 2.5 C alone. When all data is included, plus the updates, that becomes a 95% range of around -0.2 – -7.5% with a median(?) estimate of -1.7%. That is a substantial difference. It would be interesting to know the estimate using just IPCC available data.
Tom, thanks.
Yes, I was a bit lax when I suggested Grant’s result were about the same as Tol’s. Do I take it from your comment that the Welfare equivalent income gain or loss is essentially the same as the annual cost? I had assumed that there were some normalisations that made it a bit more than simply an annual cost.
Anders, on economics I am moderately informed, but far from expert. I believe that “welfare equivalent” gains and losses would be gains and losses normalized by a specific means. Something analogous to purchasing power parity adjustments rather than inflation adjusted. Not knowing the specific algorithm used for the measure, however, I cannot elucidate further than that. That said, I suspect that the costs specified in WG2 are in “welfare equivalent” terms, and hence they are directly comparable with Tol’s figure. It is not clear to me that the same can be said of the figures from WG3, I point I would certainly like cleared up.
Tom,
Thanks. All I know about the WGIII and WGII comparison is that I asked Brigitte Knopf (who wrote a recent RealClimate post about the WGIII report) if they could be compared and her response was that they are not directly comparable.
Pingback: The power of passive …
An welfare-equivalent income loss is a welfare loss that is equivalent to a welfare loss due to an income loss. Specifically, ~2K of warming (relative to recent times) would make the average person feel as if she had lost 1.1% of her income.
WG3 reports income losses. Income losses and welfare-equivalent income losses are directly comparable (which is why the latter are defined as they are).
The headline numbers of WG2 and WG3 are thus directly comparable (by design).
The big difference with the old and the new results is not what happens at modest warming. The change reflect the limits of parametric statistics. See http://ideas.repec.org/p/sus/susewp/6914.html
There is a big difference for large warming though.
Richard,
Thanks. As I read it, the WGII numbers appear to be income losses, but the WGIII numbers are losses in global consumption. It was my understanding that these were not directly comparable, but you seem to be suggesting otherwise.
Both WG2 and WG3 give percentages.
Consumption is a fraction of income. That fraction is by and large independent of climate change and climate policy.
Doesn’t that imply that you need to know a relationship between consumption and income to really do a comparison?
If C = (1-s)Y, then (1-d)C = (1-s)(1-d)Y.
(C=consumption, s=savings’ rate, Y=income, d=damage)
A percentage loss of income equals a percentage loss of consumption.
So – if I understand this – all else being equal a loss of x% loss of consumption is equivalent to an x% loss of income. However, the WGIII numbers do not include benefits of reduced climate change as well as co‐ benefits and adverse side, so it would seem that the two situations are still not strictly comparable. Assuming, however, that we need to make some kind of comparison, it would seem that your updated analysis suggests that the loss of income due to climate change is comparable to the loss of consumption due to a mitigation scenario that reaches CO2 levels of around 450ppm by 2100. Is that a fair conclusion?
@Wotts
No, that is the wrong conclusion.
The impact of climate policy (for 450 ppm CO2eq) in the second half of the century is 2-11%; the impact of climate change is up to 2%. The benefit of climate policy is the avoided impact of climate change, which is necessarily smaller than the impact of climate change.
In other words, 450 ppm does not pass the benefit-cost test.
Richard,
Hold on, you appear to have quoted the IPCC result which was for ~2 degrees of warming. I was referring to your updated analysis. Your updated analysis seems to show a range of 0 to about -7.5% for 2 degrees of warming. However, if we choose to do nothing and continue increasing our emissions the evidence suggests that we would reach 2 degrees – relative to today – before 2100. The RCP8.5 mean relative to today is more like 3 degrees by 2100 which your updated analysis would suggest would produce income losses of between -0.5% and -12.5% (roughly). So, I’m slightly at a loss as to how you state that they are as different as you suggest.
You also appear to have ignored this caveat
Ottmar Edenhofer has written a letter to the editor of The Economists. The whole letter is reproduced at Klimazwiebel.
The main content of the letter, as I read it, is confirming that the costs estimates of WGIII should not be taken as close to accurate knowledge:
Pekka,
I thought one of the main points in the letter was this
Which I had taken to be a response to the Economist article which had compared the reduction in the annual average consumption growth rate from WGIII with the reduction in income from WGII (which is an annual figure – I think).
Having said that, I don’t doubt that these numbers should not be taken as close to accurate knowledge, but I would suspect that that would also apply to the numbers presented in WGII.
One additional problem with the cost estimates is that they are calculated as a difference between two optimized scenarios without any emphasis on the problems of forcing the change. I have done many similar model calculations in the past and found that the cash flows between various participants of the economy change typically many times (a factor of 10 is typical) more than the calculated net cost. That means further that some benefit a lot while others loose a lot. The disruption in the operation of the economy is also likely to be much larger than one would expect from the net cost.
Actually we have again the situation that I don’t believe that any of these numbers, or any other number, can describe at all, what the change will be. The reason for that is that the way people live changes in so many different that the whole statement that consumption is x% smaller tells almost nothing on that.
People adapt to the development. They use new opportunities and they find solutions to problems. Different climate policies lead to different paths. The outcomes are different, but one is not 5% better than the other. The development influences also the measures of economy – and those of well-being. Measuring the other state by the measures of the other might give conflicting results depending on the selection of the measure.
A society that produces 80% less CO2 is likely to by quite different from another that has not reduced the releases from the present. Similarly the changes in climate may have led to quite different societies over 50-100 years.
I don’t believe that direct economic comparisons are possible at all. Arguing about the sign of the effect of first 2 degrees is totally irrelevant. Up to that point the effects are likely to be small, and that’s all that matters, but going further we enter the issues I describe above.
“No need to panic about global warming” arguments are about as sophisticated as “we haven’t driven off the cliff yet so it’s fine to carry on accelerating”.
“Arguing about the sign of the effect of first 2 degrees is totally irrelevant. Up to that point the effects are likely to be small(.)”
This optimism seems unjustified given the extent to which economic modeling seems to lag reality.
Returning to Richard Tol’s claim that WG2 costs and WG3 costs are directly comparable, I do not see how that is possible. The cost in a particular year will be, among other things, a function of reduced growth in prior years. The WG2 values, however, specify costs at a particular temperature. They are not varied based on the time to reach that temperature, or the temperature pathway over time to reach a particular temperature at a particular time. Thus, if temperatures rise to 4 C, before falling back to 2 C in 2100, the cost will be different than if they rise gradually to 2 C in 2100, a fact not reflected in the WG2 figures.
Because of this, it appears to me that the WG2 costs are more directly comparable to the 0.06% reduction in economic growth for the 450 ppmv mitigation target than to the annual value of the reduction in any given year. Specifically, the reduction in consumption in a given year will be the compounded annual reduction in growth due to increased temperatures in preceding years up to and including that year, with the specific factor in each year determined by the temperature of that year using the formula describing the cost at a given temperature found in figure 2 of Tol’s correction (or other equivalent estimate).
The alternative view, that the losses are directly comparable to the cost in a given year as per WG3 seems equivalent to an assumption that the impacts of global warming cannot reduce economic growth, no matter how violent or comprehensive they are in any given year. To me that assumption is absurd (yet is continuously built into economic treatments of the costs of global warming).
Economic loss results from preceding lack of growth. Lesser of growth is the consequence of loss in the economic activity. One is the derivative, the other the integral. They go together and tell about the same thing.
Pekka: “I don’t believe that direct economic comparisons are possible at all. Arguing about the sign of the effect of first 2 degrees is totally irrelevant.”
I tend to agree with you about the limited value of these economic costings. Never-the-less, there are people who use such costings to argue for policies that with high probability take us above a temperature of 2 C above the pre-industrial average. That group is certainly not limited to contrarians, including as it does (for example) Nordhaus. Therefore it behoves us to understand their arguments to see if their conclusions follow even on their own terms. It also behoves us to understand their arguments so we can identify and rebut misrepresentations of the arguments such as those by various members of the GWPF.
Pekka, in 2010 Russia suffered massive economic losses. They were not the consequence of preceding lack of growth, but of a climate event in 2010. By my understanding the WG2 figures must represent estimates of equivalent losses rather than losses due to reduced growth in prior years due to equivalent events because they include no temporal temperature trajectory, If that is correct, they are not directly comparable to WG3 figures (other than the compounding annual values). Rather, we must use them, for a given temperature projection, as the compounding annual values for that projection to determine the total annual cost in any given year in that pathway.
Basically we are comparing levels of consumption and using that as a measure of well-being. Change in that level is a change in that level whether it’s due to smooth slowdown in preceding growth from diversion of resources to mitigation or to gradual deterioration of the environment.
Details of measurement of the change may vary but basically the outcome is comparable to the extent economic measures are applicable.
@ Richard,
The change reflect the limits of parametric statistics. See http://ideas.repec.org/p/sus/susewp/6914.html
I agree about the limitations of parametric stats. (Indeed, that’s partly what my blog post was about.) However, as also mentioned in my post, I don’t see how non-parametric estimation can provide meaningful results here either. The sample size is simply far too small.
Without the outlier from Tol, what is the justification for a non-linear fit in either version of the two graphs?
The fact is that we’ve moved beyond asking if Climate Change is real, to intentional malice of forethought geoengineering.
Why would anyone hasten to geoengineer planet Earth without knowing what the target is and what it will do?
There are no calculations or data on the costs of reversing damage done by climate change. It may become irreparable, and irreversible. What destruction of ecosystems is our goal?
The fact is that the costs are unknown at this time. I say this with confidence because there are no defined goals or targets. It is insensible to argue to continue geoengineering the planet without further understanding what we intend to do to it.
Grant,
I had a look at your blog post and I agree; it beggars belief that someone would look at a quadratic fit to a small set of data points and blindly draw inferences from the behaviour of the curve (in this case, “beneficial up to 2.2C”). It’s about as sophisticated as those I’ve seen fit a high-order polynomial to temperature data and use it to show an ice age is imminent and AGW is a hoax. (Yes, it’s been done.)
I think a more accurate representation of the various studies would have been:
* Tol (2002) claims there will be a 2.3% improvement to economic welfare with a small (1C) amount of warming.
* Everyone else thinks there will be anywhere from a tiny improvement to very large reductions in economic welfare with any amount of warming, with reductions in economic welfare tending to become larger as warming increases.
Rather than trying to somehow fit them all together, I think it would be worthwhile looking at Toll (2002) closely to see why it’s such an outlier.
Regarding your quote from Wikipedia: “Nonparametric regression requires larger sample sizes than regression based on parametric models because the data must supply the model structure as well as the model estimates.”
This is obviously true, but it’s easy to overlook when it has been applied. As Eli asked: Why a quadratic? Is it because we have some theoretical model that predicts a quadratic relationship between temperature and cost a-priori, or was it chosen post-hoc on the basis that it looks like it would fit the data best? In the latter case, then the data is being asked to supply the model structure, too.
I don’t dismiss the possibility that complex economic interactions could be accurately described by simple-looking formulas thanks to the Law of Large Numbers (much like the ideal gas law), but I nevertheless always find myself being somewhat suspicious of the exceedingly simple equations economists seem to proffer. As a postgrad I earned some extra cash helping an economics professor with his course, and my conclusion from the economics I learned as part of that was that all the problems economists tried to solve were either (a) trivial and unrealistic or (b) impossible. No doubt part of this was because it was only an undergraduate course, but I can’t help noticing the poorly-predicted economic turmoil of the past decade or so and wonder just how much weight we should give to statements about which course of action is most cost-effective down to a few percentage points. Perhaps we should be thinking more in terms of “taking responsibility for cleaning up our own mess” rather than “I assume that people in the future will be richer than us, let them deal with it”.
In response to Eli’s concerns, we know that some where between 5 and 15 C above preindustrial, welfare loss approaches 100%. No linear fit to the data, whether Tol 2002 is included or not, is compatable with that. Therefore the fit must at least be quadratic, and may be worse than that.
If that is not enough, we also know that welfare loss at 0.6 C above preindustrial (ie, 0 C above the 1986-2005 average) is 0% by definition; and that welfare loss at -0.6 C is likely to be greater than 0%. Further, it is likely very large at -6 C (glacial conditions). Again these facts are inconsistent with a linear fit.
The numbers in WG2 and WG3 are not comparable in any meaningful way.
WG2 simply says annual global income loss will be 0.2 to 2.0% if we reach a further 2°C warming. It’s a snapshot, unless global warming peaks at ~2.8°C. A scenario in which global warming peaks at 2.8°C requires substantial mitigation. Hence this figure only gives the cost of climate damages in a scenario which also includes substantial mitigation. The fact that it’s a smaller number than the cost of mitigation to 500 ppm CO2 is meaningless, because to limit global warming to 2.8°C requires limiting CO2 to around 500 ppm (~550 ppm CO2-equivalent). The costs have to be added – they can’t just be compared to tell us anything meaningful.
WG3 on the other hand says that achieving peak CO2 at 450 ppm will cost 0.06% of annual economic consumption (compared to a fictional perfect world baseline in which climate change has zero economic impact). Frankly I think this number is pretty useless too, because of the comparison to a fictional baseline, but at least it’s based on an actual emissions pathway rather than just being a snapshot.
Using IAMs to find the economically optimal emissions pathway is far more useful than either of those figures, and according to PAGE an DICE, the optimal pathway is around 450–500ppm CO2, which is a scenario that requires substantial mitigation.
Tom,
Yes, this is something I’ve been wondering. The WGII numbers are income losses in a particular year. However, it seems that these losses must have some influence on growth and since – as you say – the calculation doesn’t seem to take into account reduced growth in previous years, it’s hard to see how one can directly compare the WGII and WGIII numbers.
ATTP,
In a similar way income losses from diversion to mitigation do also affect further growth. There’s no difference in that.
You are on the right track in noticing that the dynamics of economic development are different in every different scenario, but you are wrong in concluding that the logic would automatically favor your conclusion over the opposite. There has been discussion in favor of each alternative. People like Azar and Sterner have discussed potential detrimental effects of warming on economic development, while many of those economists who are unhappy with the claims that low discount rates are the right conclusion argue effectively for the opposite.
Pekka,
Sure, but the WGIII numbers were presented as the annualized reduction in consumption growth. That would seem to suggest that the 3% to 11% reduction in consumption by 2100 has taken into account the reduction in growth in previous years. The WGII numbers were simply the income losses in a single year (as I understand it) and if this doesn’t take the pathway into account would then seem to not include the reduction in growth in previous years.
Pekka,
I’ve just noticed this – I really should read comments more thoroughly 🙂
I’m not sure what you think I’ve concluded. I didn’t think I’d actually concluded anything. I was simply trying to understand if – for example – the WGII numbers (and those in Tol’s analysis) take the pathway into account. If they don’t (as it seems) then it would seem that comparing these numbers (income losses for a given increase in temperature) with those from WGIII (which do seem to take the pathway into account) is not – strictly speaking – a particularly meaningful comparison.
Both tell about the income or consumption loss at specific times, the difference is only that the same loss is told in a different way. Neither approach looks at indirect effects. They are no different in that respect.
Pekka,
Sure, but doesn’t the pathway matter? The WGIII number is presented as being associated with a annualised reduction of 0.06% in consumption growth. So, for example, if in an ideal world with no climate change and no mitigation, the growth rate is 3% per year, then mitigating means the growth rate will be 2.94% per year.
The WGII number – as I understand it – is an estimate of how much climate change – in a single year – will cost (reduce income) if temperatures rise by a further 2 degrees. However, it would seem to me that climate change costs in preceding years should also influence the growth rate. So, if the WGII number doesn’t consider the pathway, then the WGII number only tells us the annual loss of income per year in a world that is 2 degrees warmer than today, but doesn’t (I think) tell us how the impact of climate change has influenced growth of the global economy up until that point in time. Maybe it does include this, but I haven’t yet heard a convincing argument that this is indeed the case.
Christ. The word “uninformative” springs to mind on both counts.
Climate change economics isn’t worth the level of serious discussion it gets.
Tom writes:
In response to Eli’s concerns, we know that some where between 5 and 15 C above preindustrial, welfare loss approaches 100%. No linear fit to the data, whether Tol 2002 is included or not, is compatable with that. Therefore the fit must at least be quadratic, and may be worse than that.
Since the most sensible way to get a handle on net-cost and benefits of climate mitigation is via IAMs (though, please note, I am certainly not claiming that they are in perfect any way), I thought that some of you may be interested in the damage functions that underpin these models.
The most famous IAM, William Nordhaus’s DICE model, using a quadratic damage function that is calibrated on Tol (2009)… But offset by an (arbitrary) 25% adjustment factor to account for the many accounted factors such as biodiversity loss, extreme events, etc (see page 10, 11 here). The function in question is: D(T) = 0.00267*T^2, where T is change in global surface temperature relative to the pre-industrial period.
[Mod : As per Grant’s request, I’ve corrected the equation. There is a typo in Nordhaus’s notes. Eq. 5 on page 11 makes it seem as though there is both a linear and quadratic term when, in this particular implementation, the coefficient in front of the linear term is actually zero.]
However, as everyone realises — not least of Nordhaus who has spend a great deal of time discussing this very issue — the above damage function is only “suitable” for moderate temperature increases. I like the polynomial damage function that Weitzman (2012) has proposed to mimic the DICE damage function at lower temperature increases, but then diverge sharply.
To get a better sense of what I’m talking about, here I’ve made a graph of the two damage functions. You can clearly see how this change in functional form matters.
Grant,
Thanks, I’ll have a look through that.
Grant,
If you have a moment, maybe you can explain the significance of Figures 3 and 4 in the DICE model PDF. To me it looks as though whatever we do, the model produces the same global output and per capita consumption. I’m presumably missing something obvious, but my first take from that is that we can do whatever we like and the global economy will grow in a similar way regardless.
Jason B – Frank Ackerman found that Tol is very optimistic mostly because of his assumptions about agriculture. (the divide by zero thing is real, but handleable and in reality a distraction from the main point)
Grant, considering that temperature increases of 10.6 C relative to the preindustrial will make the tropics seasonally uninhabitable for any non-aquatic mammal larger than a medium size dog, a welfare loss of circa 20% at that temperature is so implausible as to be a reductio ad absurdum to the damage function that predicts it. Therefore Weitzmann’s function as you graph it is of very limited interest.
Tom, Grant,
From my reading of the DICE PDF and even the Weitzmann paper, the damage function is simply an input to a typical IAM. If so, where does it come from in the first place?
@Tom, ATTP
Woops! Sorry, in my haste to graph the figure I attached the wrong names to the red and blue lines in the legend. See the correct version here. (It is therefore the DICE damage function — Nordhaus (2013) — that is off very limited interest. Weitzman’s function more accurately shows that welfare loss will be around 100% of GDP at 10 C increase, which is why I like it.)
ATTP, can you change the link above as well? Thanks
Grant,
Done.
@ATTP
Grant,
If you have a moment, maybe you can explain the significance of Figures 3 and 4 in the DICE model PDF. To me it looks as though whatever we do, the model produces the same global output and per capita consumption. I’m presumably missing something obvious, but my first take from that is that we can do whatever we like and the global economy will grow in a similar way regardless.
No, you aren’t missing something obvious 🙂 That is what IAMs produce and it shows how small the calibrated effects are in these.models. Though it must be said that, with specific regard to FIg.’s 3 and 4, we should also bear in mind that the really bad effects will only start to take hold at temperature increases that occur past the year 2100. (In determining the “optimal” abatement path, you are thus effectively playing off the twin forces of discounting and really long-lived — but far into the future — negative effects.)
On a somewhat related note, see this paper by Azar and Schneider (2002). If you want to argue that the differences in outcomes are small no matter what we do, then you could similarly argue that the costs of stabilisation aren’t very high either.
The above is from the introductory chapter 2 of the thesis of Tommy Ekholm: Risks, costs and equity. Modelling efficient strategies for climate and energy policy. While I have met Tommy several times and know well most of the people he thanks in the preface, I haven’t contributed in any way to the thesis. I think Tommy describes well many of the problems in the above quote.
The rest of the rather short (15 pp. of text) summary part of the thesis is also worth reading.
Yes, I thought that this was the take-home message from the WGIII report. A reduction in global consumption in 2100 of 4.8%, essentially means the difference between the global economy being 10 times greater in 2100 than today and being only 9.5 times greater than today. So, even if income losses due to climate change in 2100 are still less than this, the argument that we should risk ecosystems over a few percent of GDP in 2100 doesn’t seem particularly strong.
From my reading of the DICE PDF and even the Weitzmann paper, the damage function is simply an input to a typical IAM. If so, where does it come from in the first place?
Partly theory, partly estimates.
Theory: We would expect climate damages to become increasingly severe in changing temperature — hence the quadratic damage function (or higher order polynomial).
Estimates: As mentioned above, the estimates are largely based on the studies described in Tol (2009)… The very topic of this post! The quadratic damage function is calibrated to match the best regression fit, often with arbitrary offsets as described above to account for things like biodiversity loss, etc.
PS – I believe that earlier IAMS like Nordhaus’s first version of the DICE model were calibrated very roughly to correspond to what a group of experts thought the damages at various temperature increases would be. Now, I obviously agree that it is better to provide quantitative estimates of the damages and then further try to find the average among all the studies that do so… But we are still dealing with very imperfect and limited data! IAMs are certainly better than guessing, but are no more than a very rough guideline to what we should be doing.
Grant,
That’s very interesting. So, if I understand this, IAMs came about because people realised that we needed some way to model how to proceed given the risks associated with climate change. An important part of such a model would be an estimate of future damages due to climate change. This is something that is quite difficult to determine, but you need some estimate in order to use it in an IAM, so you do the best you can using various theories and estimates. This allows one then to use an IAM to inform climate policies. However, because the damage function suggests that damages will be low for small increases in temperature, some are essentially arguing that we shouldn’t bother doing anything because the damages will be lower than the cost of mitigation. However, it seems as though the damage function was never intended (not initially at least) by itself to suggest a possible policy option. It was really just intended as an input to IAMs. So, we make a best guess at a damage function that can be used in IAMs (and that many agree is possibly one of the most uncertain aspects of these models) and this damage function then gets used – by itself – to argue against any kind of aggressive policy action. Is that a reasonable assessment of the position we’re in?
Sorry for linking to a bunch of papers, but since biodiversity loss and ecosystem risk keep coming up…
One of my all-time favourite articles on the economics of climate change is “An Even Sterner Report” by Sterner and Persson (2008). Not only do they give a great summary of the whole discount debate, but they show how important it is to account for the imperfect substitution between man-made and natural goods in determining optimal climate policy. (Conventional IAMs just abstract from these issues and just assume that — to take a deliberately absurd example — we’d just able to replace forests with iPads if we cut down all the trees.) Of course, what would actually happen is that the relative price of natural goods will go through the roof as they become scarcer, making the BAU emissions scenario increasingly untenable… even if we assume a high discount rate.
I have given a brief overview of the paper and its implications on my blog here and here if anyone is interested.
Grant,
Linking to papers is good 🙂
ATTP,
Yes, I think that’s a very respectable summary.
Both Grant and myself have referred above to work of Thomas Sterner. He has written also a nice book on Policy Instruments for Environmental and Natural Resource Management. I haven’t seen the update edition with Jessica Coria as coauthor but I liked very much the first edition.
He visited my university soon after writing the article “An Even Sterner Report” and we had a really nice small seminar where that work was discussed. I think that the ideas of that paper are very important, but there’s still much open in the understanding of those issues.
ATTP
And a fine summary of the reasons why we ignore the majority of climate change economics in favour of listening to ecologists and marine biologists.
To be fair, Grant’s links to his blog and the Sterner & Persson paper are extremely interesting, to the extent that I have saved them.
ATTP: “to argue against any kind of aggressive policy action”. Except that Tol does, in fact, argue for strong policy action (not sure it’s aggressive!) in his 2009 paper (PDF) saying, “there is a strong case for near-term action on climate change, although prudence may dictate phasing in a higher cost of carbon over time, both to ease the transition and to give analysts the ongoing ability to evaluate costs, benefits, and policy mechanisms”.
That might not be as strong as some would like, but it’s a long way from saying “do nothing until 2050”. What I don’t know: how much Tol has taken people to task if they have used his work to argue there’s “no need to panic” and action should be delayed, as Ridley has.
Anders (May 5, 7:44 PM) writes, “However, if we choose to do nothing and continue increasing our emissions the evidence suggests that we would reach 2 degrees – relative to today – before 2100. The RCP8.5 mean relative to today is more like 3 degrees by 2100…”
The RCP 8.5 scenario is almost certainly not what we will reach “if we choose to do nothing.” As I’ve tried to point out several times previously, the RCP 8.5 scenario involves consuming approximately 40 billion metric tons of coal in the year 2100. No organization or individual with knowledge of the history and likely future of coal production thinks that is a credible scenario, absent developments that certainly weren’t described in the paper that introduced the RCP 8.5 scenario.
The RCP 8.5 scenario…not “impossible,” but extremely unlikely
Grant McDermott; I’m ever off topic, but can you recommend any papers on ecological collapse? I’m curious about prediction, and the cost of restoration.
I probably should have linked to my comments, rather than David Friedman’s blog post (which was based on a misunderstanding of my previous comments on his blog):
RCP 8.5, extremely unlikely based on past history
China’s coal use peaking prior to 2030 will virtually guarantee that RCP 8.5 will not occur
Mark,
I know that you’ve pointed out that the RCP8.5 scenario is unlikely to be the scenario that we will actually follow and I don’t disagree, I just don’t quite share your optimism that it is unlikely as you claim. Also, by “do nothing” I actually “do nothing to reduce our emissions”.
Dan,
Certainly, it seems that if you read the literature, what that says is not always what one would guess if based on what people say publicly.
Grant, is D(T) = 0.00267*T + 0.00267*T^2, for the DICE damage function a miscopying or is that really it?
Dan: Tol is pulling the full Lomborg there. RP play the same game.
Eli,
Certainly seems to be the same form as the function on page 10 of the DICE model User’s manual.
Figure 2 of the users manual shows a damage function that has no linear term as it has zero derivative at origin.
Pekka,
Isn’t that because given the function Grant presented, the gradient at T=0 would be 0.00267, which seems pretty close to zero.
No, looking at the curve shows clearly that the coefficient is much smaller than 0.00267 = 0.267%
The value is consistent with zero with a maximum perhaps one tenth of the value given.
Picking two points from the curve in addition of the origin and solving for the coefficients, I got a linear coefficient of about 0.0002, but consistent with zero. Coefficient of the quadratic term is about 0.00233 according to this solution. The curve is not consistent with the value 0.00267.
Pekka,
I don’t know the answer. Certainly equation 5 seems to have a linear and a quadratic term, but I can’t find what value is actually used for Psi_1.
Anders, based on the same reasoning the IAMs themselves can’t be trusted as a basis for policy. Yet you seem to be discussing them as if they can be.
From one of Grant’s blog posts:
“So, the choice of discount rate has a potentially massive impact on the perceived efficiency and timing of climate policy. Again, it is worth emphasising the intergenerational nature of climate change, where present (and future) sacrifices will have to be made in order to protect future incomes and utilities. The corollary of this is that acting too slowly may harm future generations in the form of rising climate costs, whilst acting too quickly may be equally as harmful by unnecessarily limiting economic growth.”
This reasoning doesn’t allow for discontinuities, of which there are all too many lurking in the physical/biological science weeds. The idea of allowing for them isn’t even especially controversial, underlying as it does the adoption of the 2C “dangerous” limit, so IMO it’s a fair statement that discounting (whether formally via the IAMs or informally via the aggregate of our actions) is the tool we use to ignore them.
Steve,
At this stage, I think I’m just discussing them. The more I’m learning, though, the more I’m realising that there may well be quite serious issues with using them as a basis for policy.
Steve,
Actually, to clarify what I said above, there does seem to be some value in the IAMs. From what I can tell, they provide no real evidence that mitigating against the risks of climate change presents any significant economic risks.
This recent paper on declining discount rates is interesting and in principle would allow IAMs to begin to bear some resemblance to reality, although as the conclusion notes it’s not possible to avoid a subjective judgement about the timing of climate discontinuities.
I looked at Nordhaus book “The Climate Casino” The figure 22 of the book is very similar to the Figure 2 of the User’s manual, but the damages shown are somewhat higher. The value at 4.5C is about 5.2% in the book while it’s about 4.3% in the manual. I picked this time five points from the curve and searched for the best fit with linear + quadratic. The answer gave a small negative coefficient for the linear term (-0.000071) and a somewhat larger coefficient 0.00263 for the quadratic term (0.00261, when linear is forced to 0). Thus the book is consistent with the 0.00267 quadratic term with no linear term.
Pekka,
Also, on page 11 it says
So, maybe there is some issue with Equation 5. It certainly doesn’t look like there is a linear term in the curve in Figure 2.
Mark Bahner: Available coal is higher than 860 billion tons.
Your articles are based on Proven Reserves. That’s what David Rutledge is using;
https://gsa.confex.com/gsa/2013AM/webprogram/Paper225602.html
Here’s his data source.
Click to access statistical_review_of_world_energy_2013.pdf
“Proved reserves of coal – Generally taken to be those quantities that geological and engineering information indicates with reasonable certainty can be recovered in the future from known deposits under existing economic and operating conditions.”
This means that current proven coal reserves represent the amount of coal that is currently economically viable to extract. Amount of Proven Reserves increases with increases in coal prices. Given how cheap coal is, the ability to extract more at a higher price is easily possible.
To avoid burning more coal;
All nations on earth must cease searching for coal. (they haven’t)
Coal prices must stop going up. (higher prices spur exploration, and increase reserves)
Chances of hitting RCP 8.5 are low… the chances of burning more than our current Proven Reserves of coal is very very high.
When the damages grow strongly it’s extremely difficult to give quantitative estimates for them, perhaps some lower limits, but not best estimates. Similarly it’s not possible to give justifiable quantitative cost estimates for damages of any size far in the future. Neither is it possible to estimate mitigation costs far to the future. These facts limit the range of applicability of the IAM type cost estimates. The estimates might be considered accurate enough to be useful as long as:
– the damages are not very large
– short term effects (like first 50 years or less) dominate the estimated value either because of discounting or because it’s concluded that the difference between the alternatives being considered diminishes in time through adaptation. (This is a more realistic possibility for specific immediate decisions than for full development paths, which include effects of anticipated future decisions.)
If it’s not accepted that both of the above requirements are satisfied, I would not consider any estimates quantitatively fully meaningful.
ATTP,
I don’t see any problem in the situation that the model allows for a linear term, while the present applications set that to zero. There’s nothing wrong with that.
Pekka,
I didn’t mean that it was a problem. I was just meaning that in Eq. 5 the coefficient in front of the linear term seems to be the same as the coefficient in front of the quadratic term (i.e., they’re both Psi_1). It’s just poor practice to write an equation in that form and then later suggest that the linear term is zero. Normally you would make sure that the two coefficients were explicitly different so that you could then specify that one was zero. As it stands, from Eq. 5 alone one would typically assume that the linear and quadratic terms had the same coefficient.
I didn’t look at the equation closely enough to notice that. It’s very likely a printing error. The coefficient of the quadratic term should probably have the subscript 2.
Pekka,
I agree, that’s all that I was really getting at.
Climate scientists can be part of the problem too (from here):
What a truly stupid and dangerous thing to say.
I looked at the manual a bit more. On page 67 of User’s manual we can read from the Excel spreadsheet that the present version uses quadratic only, while lines do exist for both coefficients separately (lines 7, 8, and 9).
The page 97 tells that the quadratic term coefficient is 0.00267. Linear term is not mentioned, but the same value is given also to Initial damage quadratic term, whatever that means.
AOM: “Chances of hitting RCP 8.5 are low”
That’s only true *if* the models correctly incorporate future feedbacks. Sadly, they are known not to. Fast feedbacks from soil carbon losses, permafrost degradation, ocean biological carbon sink losses (and/or simply a saturation limit being reached), and tropical deforestation aren’t even the complete list. So the anthropogenic emissions path can be too high even while the resulting concentrations are about right or, more likely, too low.
Pekka, Eli, ATTP.
Sorry folks, but there is a slight mistake in the damage function given for DICE above. (I’d like to blame Nordhaus since that is actually a typo in *his* notes, but I was rushing and should have remembered instead of just copying blindly.) Anyway, Pekka is correct: The linear term in the quadratic function is indeed zero. The correct function is therefore:
D(T) = 0.00267*T^2
Although, Eli, I’m not sure whether that will put you more or less at ease!
ATTP, please feel free to edit my earlier comment to avoid further confusion..
PS – You can be sure that it is zero by looking at the actual GAMS (programming) code given in the appendix of the book. Specifically, look at the damage constants — “a1”, “a2”, “a3” — given on page 97, and then the damage equation (“damfraceq(t)”) that takes these numbers as arguments on page 100:
a1 Damage intercept /0 /
a2 Damage quadratic term /0.00267 /
a3 Damage exponent /2.00 /
…
damfraceq(t) .. DAMFRAC(t) =E= (a1*TATM(t))+(a2*TATM(t)**a3)
Note: ** indicates exponent in GAMS
Grant,
Thanks. Have corrected the equation in your first comment and added a moderation comment pointing out the typo in Nordhaus’s notes.
Not a problem Grant, the version in your comment just looked odd. More concerning is the fact that Nordhaus calibrated his damage function to Tol 2009, but looking at the D(T) = 0.00267*T^2 without any linear term, Nordhaus may have spotted the errors in Tol (2009) and modified it, of course, if he did so then the question is when.
Eli,
It doesn’t look like Nordhaus noticed the error in Tol (2009) as Fig 2 in Nordhaus’s notes still seems to have Hope (2006) at +0.9, rather than -0.9 (well, in the case of damage rather than benefit, as -0.9 rather than +0.9).
Pingback: Singolare ingannevole – Ocasapiens - Blog - Repubblica.it
Steve Bloom: (Weren’t you supposed to email me?)
I think the point that is being Cherry Picked here is that we aren’t likely to hit RCP8.5 as defined. I’m of the opinion that we will find more fossil fuels that are currently unaffordable, or that we will find different fossil fuels.
Fracking Natural Gas and Shale Oil was virtually unheard of 20 years ago. They are dirty (high carbon) technologies.
FYI… with the EPA’s new Green Completions regulations, fracked natural gas only has half the emissions of coal. Coal plants that run out will likely be converted to natural gas.
However, the rest of oil and gas should get a lot dirtier. The big push in the industry is to go deeper for oil. Higher oil prices mean that oil found at higher temperatures and higher pressures are relatively affordable.
And as an engineer I don’t like rolling dice. That’s why I commented that we are geo-engineering the planet with intent at this point, and yet we have no clue as to what our goals are or what the final state will be. In industry, badly defined projects with no clear goals almost always end in lawsuits and still born products.
Very soon now, AOM.
I think this very recent and very strong critique of the IAMs by Sterner and others got mentioned in a prior related post, but it’s worth noting again here. Note the timing relative to publication of the WG2 report.
AOM: “I think the point that is being Cherry Picked here is that we aren’t likely to hit RCP8.5 as defined.”
Exactly, and maybe even more to the point no one ever thought or claimed otherwise. Except Bahner, apparently.
Oh Dear!
Arguing about possible economic impacts… I don’t want to argue about any of the economic models being right or wrong. I just want to point out that they are only about the effects on the HUMAN economy. These models have a severe blind-spot. Perfectly possible WE are better off in 2100 than we are now. Would there be anything left in that world except for US though? What is the price we have to pay in their models aside from the pure economic effects? I’d like to know. I would happily pay some tax to avoid the economist’s valuation of EVERYTHING in human terms only.
Reich,
Absolutely, I agree.
Over at Rabett Run, Taylor B. points to a recent paper by Robert Pindyck that asks the question:
And then answers it (abstract):
As Jon Stewart is wont to say at times like this: BOOM!
Many people have pointed to various limitations and problems if IAMs and other similar economic comparisons. The dilemma is that decisions must be made, there are always many objectives that should be taken into account, and the many objectives would lead to very different conclusions, when each of them is considered separately. Thus making decisions means unavoidably that weights are chosen for the conflicting objectives. The weights are chosen either explicitly or implicitly. The explicit choice allows for analytic discussion, the implicit not.
If we wish to discuss the the choice using arguments that can be compared, and in a way through which we may wish to convince others that our weights are better supported that some others, we must have some way of making the criteria comparable. We must set a scale for the importance of each difference. When we have a scale, we may present that in monetary units.
We have the dilemma that we cannot make good enough economic calculations, but at the same time we cannot argue rationally with those whose judgments differ without something that’s essentially equivalent with an economic analysis.
People who have built IAMs have realized that, and they have made their best efforts to produce numbers for comparison. I agree that those numbers are often of questionable value, but agreeing on that is of little help until some alternative approaches can be proposed to solve the unavoidable questions of decision-making.
The main idea that I have in mind that the economic approach should be simplified. Reliance on specific models should be reduced. In particular the most difficult part of weighing the distant future and extreme risks should be approached in a different way. In that the importance of those factors should be retained, but the futile attempts of full quantitativeness should be replace by some semiquantitative approach that saves comparability to the extent it can be done, but avoids the arbitrariness of IAMs in handling those issues.
I don’t really know, how far that’s possible, but that seems to me the only possible way forward.
Pekka I would suggest that with the sort of problem that climate change presents and the limitations of economic analysis its a severe mistake to give CBA a privileged position at the table. it’s wasn’t done for other global problems that pose existential threats (e.g WW2). CBA implicitly frames climate change as an optimization problem when it’s better addressed as a risk management problem.
Marlowe –
Some questions.
–> “its a severe mistake to give CBA a privileged position at the table.”
How is that to be avoided? With an accompanying agenda, partisans will twist CBA (deliberately or otherwise) to use as part of their arsenal. For example, “skeptics” will claim that near term costs are high and unavoidable, and the long-term benefits are either minimal or non-existent or at least unverified so as to be inadequate for basing policy (and therefore it is “alarmist” to consider them in trade-off for what they see to be certain near term costs).
We see this happening constantly. Even those who claim to stake out some kind of nebulous middle ground in the battlefield, such as Judith or RPJr or putative “luke-warmers,” claim that near-term costs are certain. A CBA (and a resulting determination of certain near-term costs) in inextricably hard-wired into their arguments.
But this is understandable, as construing the issue as one of trading costs for benefits is an intuitive component of decision-making. Risk assessment the in the face of uncertainty is far more complicated, so people naturally gravitate towards creating a certainty in CBA outcomes even when it can’t be supported.
–> “it’s better addressed as a risk management problem.”
So you say that is how the existential threat of Nazism was dealt with – but can there really be a parallel given that the risks associated were much more apparent on a short-term time frame? Is there any way that climate change will be approached as anything other than a CBA problem unless people see the risks as being immediate?
And are you suggesting that CBA and risk assessment are somehow mutually exclusive? Isn’t there, necessarily, overlap?
Something that I’m getting from this (which may not be quite right) is that it’s almost as if people have forgotten the initial assumptions. It seems as though much of this work (IAMs, damage functions,…..) was developed with the assumption that climate change presented a risk and we needed to use modelling (however imprecise) to guide policy makers.
So, the initial assumption is : we must do something, what can we do that both minimises the problem without impacting our economies too adversely.
Now we seem to have people who are using these imprecise models/calculations to argue that actually the best thing to do is to do nothing. But that’s not really consistent with the reasoning behind developing these models in the first place. They were developed to try and guide us in what we should do, not to determine if we should do something or not.
My assessment of this may be wrong, but that’s just a sense I’m getting from what I’ve been reading and from what others are saying.
Joshua,
I agree that CBA will inevitably be misused by some to further political objectives. the subject of this post being the most obvious example. that doesn’t make it appropriate. i don’t agree that people think of long-term problems in a CBA framework though. quite the opposite. posterity is usually considered in normative terms. the question of ‘what kind of world do I want to leave behind for my children?’ is usually framed in terms of equity, quality of life etc, not how many flat screen TVs will they be able to buy.
Finally, I don’t mean to suggest that CBA and risk assessment are mutually exclusive. it really boils down to what your objective is and how you treat uncertainty.
This is the basic idea that my favorite development economist Partha Dasgupta has taken as starting point in his book Human Well-Being and the natural environment.
In that approach an attempt is made to value the state of the world at a point of time not too far in the future. That valuation depends on what we expect to happen later. Thus it’s not possible to get rid of long-term considerations, but even so this may be a more promising approach than standard IAM or CBA. The result of that valuation tells what we do leave behind for our children.
See also Richard Gardiner’s Climate Change: The Perfect Moral Storm”
That hits the nail squarely on the head, I suspect, Eli.
We love out gesture politics, we do.
Please God, make me pure, not just yet. Was Augustine of Hippo a climate economist? http://en.wikipedia.org/wiki/Augustine_of_Hippo
Public copy of the Gardiner paper, well worth a read.
Also mentioned at Eli’s, it turns out Stern very recently went after the IAMs (title/abstract):
All of this stuff has me wondering how the related WG2 text ended up the way it did. Was there a widespread assumption that the WG2 authors would take these concerns into account, and public pushback began only when it became clear that wasn’t going to happen?
Steve: Maybe the WG2 guys were playing around on the internet when they should have been keeping current? I can think of one who has.
Anders –
–>”So, the initial assumption is : we must do something, what can we do that both minimises the problem without impacting our economies too adversely.”
I think that you are zeroing in on the crux of the biscuit (Zappa reference, for the youngsters). IMO, people start with an assumption as to whether we should do something or not, and then try to use these assessments to try to justify that assumption. But in fact, if someone assumes that we should not do something (which applies to the “skeptics” that I have run into), then there’s no point in using these tools. “Skeptics” should simply stay consistent with their starting assumption, in which case these tools, in the sense that they are intended, would be useless.
Hi Anders,
You write, “I know that you’ve pointed out that the RCP8.5 scenario is unlikely to be the scenario that we will actually follow and I don’t disagree, I just don’t quite share your optimism that it is unlikely as you claim.”
I think you would find it as unlikely as I claim if you took the time to do the arithmetic in terms of 40 billion metric tons (44 billion “short” tons) of coal usage in 2100. Seriously…if you think it’s possible for the world to be producing 44 billion short tons (40 billion metric tons) of coal in the year 2100, why not sketch out where that coal is being produced?
Here was the production in 2000 (in billions of short tons):
China 1.51
United States 1.07
India 0.370
Australia 0.338
Russia 0.265
South Africa 0.249
Germany 0.226
Poland 0.179
Kazakhstan 0.085
Indonesia 0.084
Rest of world 0.742
Total world 5.13
According to the RCP8.5 scenario, the production will increase steadily to the year 2100, to reach 44 billion short tons. Come up with the numbers for the various countries that sum to 44 billion. For example, China’s government expects its consumption to peak at 4.7 billion tonnes (5.2 billion short tons), and decline by 0.43 percent per year thereafter:
http://www.chinadaily.com.cn/business/2014-03/04/content_17321848.htm
So unless you expect China to be exporting a whole lot of coal in 2100, their production should be under 4 billion short tons in 2100. And since a lot of Australian coal goes to China, you shouldn’t have their production in 2100 much over 1 billion tons. And since U.S. production has actually declined since 2000, U.S. production shouldn’t be massively larger in 2100 than it was in 2000.
Mark
Hi,
I forgot to include: “For example, China’s government expects its consumption to peak at 4.7 billion tonnes (5.2 billion short tons), ***in 2020*** and decline by 0.43 percent per year thereafter:”
In other words, China’s consumption is expected by it’s government to peak in 2020, versus the IPCC RCP 8.5 scenario have world conumption continuously increasing all the way to 2100 (and beyond!).
Pingback: Ukraine: cartoon [Stoat] | Gaia Gazette
Mark Bahner: I’d need to know what the industry defined as Proven Reserves, as that is exceedingly central to your point.
By definition: Higher Coal Price = Higher Proven Reserves
A fact overlooked in any analysis you’ve pointed to.
Mark Bahner points to the intention of the Chinese government for its coal consumption to peak in 2020. That intention, however, is based on an intention to reduce carbon intensity as a response to the threat of climate change. He therefore arguing that no mitigation strategy is needed because the Chinese government has a mitigation strategy.
Well, its a nice rhetorical strategy if you can get away with it…
Why is Mark banging on about RCP8.5 as if it is the only thing that matters?
Perhaps looking at CO2 ppm, CH4 ppm and carbon cycle feedbacks (as Steve mentions above) would be more illuminating.
Did somebody say “rhetoric”?
One essential starting point for the discussion of RCP8.5 is given by authors of that scenario
Click to access Riahi_RCP8.5_2011.pdf
IIASA has worked continuously with energy system models and energy scenarios since 1970s. Nebojsa Nakicenovic and Arnold Grübler are the senior scientists behind RCP8.5, and they where among the main scientists of that early work of IIASA already. One of the participants in the early part of that was also Hans-Holger Rogner, whose 1997 estimates of coal resources have been used in many later papers and are one reference also in the RPC8.5 paper.
Thus lets see, how they characterize the scenario (emphasis mine):
and
We see that the scenario has been built taking into account estimates of coal reserves, and building in assumptions of rising fuel prices. The scenario has the nature of upper limit at least among existing scenarios, but in many ways also more generally as a combination of factors that weigh heavily towards high consumption levels. It’s not simply business-as-usual but rather business-as-usual under circumstances that favor strongly high consumption of fossil fuels.
The Fig. 5 of the paper shows clearly how extreme the growth in coal supply is in the scenario. My own feeling is that the estimate is, indeed, unrealistic. The marginal cost of coal required for that consumption would probably be far too high. Many of the scenarios assume that the fuel consumption of billions of people could keep on growing at a fast space over the rest of the century. That’s not a likely scenario even in absence of any climate policies or climate related damages. China, India, and other large populous countries are affected also by other restrictions that make that type of growth highly unlikely. Hopefully they can find alternative positive paths of development. Copying, what the present rich countries did many decades earlier is not possible – or even desirable. Limited energy supply is just one of many reasons for that.
Pekka,
I agree with your last paragraph in general – it may well be unrealistic. However, what you say would also seem to imply that continued growth at 3% per year (leading to a world economy 10 times richer in 2100 than today) is possibly also unrealistic unless we can find ways to continue increasing the energy supply.
ATTP
What I say is that continued growth at that rate cannot be identical to the past growth, or the present growth, but it’s no news that there’s more change than simple linear growth. How the growth is measured, when it’s combined with change is a complex question. It’s usually done looking at differential changes based on current prices of each moment and compounding annual changes. That’s technically possible, but how meaningful that is?
Another logical choice would be to pick a staring point (say 1914) and this year and ask, how much richer we are now or how much more we earn in real terms compared to 1914. If the measurements of growth would be fully well defined we would get the same result in both ways, but even assuming that we had perfect statistics from all countries in 1914, it would be impossible to compare that year directly with 2014. The impossibility applies both to inflation and to GDP and other similar measures. Compounding annual changes gives an answer, but that answer gets less and less meaningful when the time span gets longer.
Andrew Revkin has written recently on the Joint Workshop of the Pontifical Academy of Sciences and the Pontifical Academy of Social Sciences, 2-6 May 2014. Through this site I found two interesting papers of Partha Dasgupta from 2004 and 2006, both on intergenerational justice and paradoxical questions raised by attempts to find right solutions for that.
The real issue with coal is the quality of what is mined. If prices go up, there is a huge reserve of low quality coal (lignite and even peat). Burning dirt is possible.
Tom Curtis: Analysis of Chinese CO2 Mitigation Strategy
Click to access p263.pdf
BBD: Its standard [redacted] tactics to try and get either first comment, or last. That way if someone wandered into the conversation randomly they’d fast forward to the end and somehow grab that sound bite. (I monitor for that behavior, and bomb it where ever possible. I don’t actually want the last post privilege.)
xkcd covered this tactic thusly;
https://xkcd.com/1019/
Pekka: I find talks about wealth amusing. Every year my bank account says the same thing. “Please insert more cash.”
In North America we are clearly wealthier… but if you look closely, people living in the burbs are slaves to multiple car payments, and rather high energy expenditures.
AnOilMan writes, “Mark Bahner: I’d need to know what the industry defined as Proven Reserves, as that is exceedingly central to your point.”
It’s not really “exceedingly central” to my point. It’s not even exceedingly central to Dave Rutledge’s point. Dave Rutledge’s point is that–historically–coal regions do not come even close to consuming all their proven reserves. Here’s an excerpt from his many writings:
Coal and the IPCC
“On the other hand, for coal the pattern has been that countries produce only a small fraction of their early reserves, and then late in the production cycle the reserves drop to match the coal at the last working mines. This pattern is seen in the UK (cumulative production of 19% of early reserves), Pennsylvania anthracite (42%), the Ruhr Valley (14%), France and Belgium (23%), and Japan and South Korea (21%). This means that the reserves criteria have been too optimistic, but it also means that world coal reserves are a good upper bound on future production. An IPCC scenario that burns two times or seven times the reserves is utterly at odds with the historical experience.” (Emphasis added.)
To his excellent analysis, I could add many things, such as that photovoltaics, wind, nuclear, and natural gas are all much preferable to coal…*totally disregarding CO2 emissions.* Therefore, *totally disregarding CO2 emissions* those technologies are likely to replace coal. For example, China’s horrendous particulate, sulfur dioxide, and tropospheric ozone air pollution problems are significantly related to coal. Therefore, there is a tremendous incentive in China to switch to natural gas and nuclear…which is exactly what is happening. According to the latest issue of MIT’s Technology Review, China has 95,000 MW of nuclear capacity either under construction or planned. That’s *95* 1000-MW plants. Regarding photovoltaics, in the U.S. Southwest (not to mention much of Australia, the Middle East, Southern Europe, etc.) photovoltaics will likely be the *least expensive* electricity source in the next 1-3 decades. China could supply all its electricity by covering relatively small amounts of the Gobi desert.
But if you think you need to know something about global coal reserves, I’m certainly not trying to stop you from obtaining it. In fact, in case you haven’t heard of Wikipedia, here’s a Wikipedia article that discussed proven reserves of coal:
Wikipedia on world coal reserves
AnOilMan continues: “By definition: Higher Coal Price = Higher Proven Reserves”
OK, whatever.
He concludes with, “A fact overlooked in any analysis you’ve pointed to.”
Whatever! If you don’t like any analysis I’ve pointed to, point to some analysis that you think refutes the analyses I’ve provided. Please provide some analysis that you think supports the RCP 8.5 scenario of coal usage increasing steadily throughout the 21st century, reaching 44 billion short tons (40 billion metric tons) in 2100.
My request to Anders (and to you, and every other person who has objected to my argument that RCP 8.5 is definitely *not* “business as usual”) was that he/you/they provide some “best guess” of how the world production of coal will look in 2100. Just to get y’all started, here’s my best guess for world coal production in 2100 (in billions of metric tons):
China <0.01
United States <0.01
India <0.01
Australia <0.01
Russia <0.01
South Africa <0.01
Germany <0.01
Poland <0.01
Kazakhstan <0.01
Indonesia <0.01
Rest of world <0.05
Total world <0.2
Mark Mahner: I’m using Dave Rutledge’s words and yours, so I don’t need another source. This is Dave Rutledge’s data source;
Click to access statistical_review_of_world_energy_2013.pdf
Page 30 of Dave Rutledge’s source states;
“Proved reserves of coal – … with reasonable certainty can be recovered in the future from known deposits under existing economic and operating conditions.”
You will also see on that page, the historical price of coal. If you read up on China they are closing coal mines that produce undesirable coal. A higher price will likely shift their position. They have been buying environmental sensors hand over fist for the last while.
In any case, RCP 8.5 is what happens if all energy favors fossil fuels to the exclusion of other sources. This is Pekka’s quote from the IPCC;
If you swap coal for natural gas, you’re going to hit 75% of RCP8.5, and that assumes we don’t find anything dirtier or cheaper. (LNG adds another 8-10% to your Natural Gas emissions. So, I guess 80% of RCP 8.5?) Who knows, maybe the tar sands will finally flow and grow, that can’t possibly affect emissions could it?
http://www.desmogblog.com/2014/05/05/first-time-transcanada-says-tar-sands-flowing-gulf-keystone-xl-south
RCP8.5 is what it is.
Point is that there is no bottom to the quality of coal that can be extracted.
Mark,
Firstly, AOM and Eli have said. Secondly, I don’t have a view as to what the world coal production will be like in 2100. My only point – which you seem completely unwilling to even consider – is that you may not be right.
“Saved by the invisible hand”? From RCP 8.5? But we don’t need RCP 8.5 to cause very serious problems. Mark is persistently ignoring comments that point to his construction of a strawman using RCP 8.5.
Mark – and others – please note the use of bold face type above.
Ah, but this only happened because of access to cheaper coal imports from elsewhere.
One might wonder how this could affect Mark’s hypothesis in an energy-hungry future with a developing India, Indonesia, China, Brazil etc. In the past and at present, the vast majority of new demand for electricity is met by increased demand for coal. Historical fact.
So on what paper does IPCC:s “With
these recognized limitations, the incomplete estimates of global annual economic losses for
additional temperature increases of ~2°C are between 0.2 and 2.0% of income (±1 standard
deviation around the mean) (medium evidence, medium agreement).” rest?
Cant pin it down… links to the reports:
Click to access SREX-Chap4_FINAL.pdf
Click to access IPCC_WG2AR5_SPM_Approved.pdf
Click to access SREX-All_FINAL.pdf
From the Executive Summary of WGII Chapter 10 we can read:
The SPM has replaced 2.5°C by ~2°C. In the paragraph of the Executive Summary we see a reference to 10.9.2. Aggregate Impacts, but no similar numbers can be found in the text of that section. Figure 10.1 forms, however, part of the section, and Figure 10.1. does indeed present values in the range +0.1 .. -2.5% (positive is benefits, negative damage) for the loss due to warming of 2.5°C. References for all these estumates can be found in Table 10.B.1.
–> “My only point – which you seem completely unwilling to even consider – is that you may not be right.”
Uncertainty is such a convenient tool. It materializes whenever you need it, and disappears obediently when you don’t want it around.
“The real issue with coal is the quality of what is mined. If prices go up, there is a huge reserve of low quality coal (lignite and even peat).”
First, we should probably make it clear that peat isn’t coal. The lowest quality coal is lignite. The IPCC RCP 8.5 scenario has ~40 billion metric tons of coal (including lignite) being burned in 2100. I’m still waiting for anyone to offer an opinion on what the production breakdown would be in various countries to get to that amount:
China ?
United States ?
India ?
Australia ?
Russia ?
South Africa ?
Germany ?
Poland ?
Kazakhstan ?
Indonesia ?
Rest of world ?
Total world ?
“Burning dirt is possible.”
Oy vey. Presently, the humanity produces over 500 exajoules of primary energy. What percentage of that is obtained from burning dirt? And in 2100, according to RCP 8.5, humanity’s primary energy production will be 1730 exajoules. How much of that do you figure will come from burning dirt?
Mark Bahner, are you aware of the technology for Underground Gasification of Coal (UGC). This technology has the potential to add huge quantities of recoverable coal reserves. Of course, the coal will not be physically removed from its source but will still produce the same amount of CO2 has if it had bee mined and burnt on the surface.
An example is the potential reserves of Norway, which does not even appear on your list of coal producing countries. It is estimated that Norway has 3000 billion tons of off shore coal which could be gasified.
This discussion is getting paradoxical. The less we can extract fossil fuels the more certain is the urgency of switching to alternative solutions. What would allow for a different policy is trust in the possibility of getting all non-conventional resources into use, and taking advantage of that possibility in disregard of the environment.
Ah, thank you Pekka. Is it justified though? First of what is the rule for what can be taken in to the SPM? Should it now have been supported in the text in the full text?
Second limiting it to about 2 degrees sure limits the amount of studies available (which would be really stupid and then perhaps not justified to make a statement about it at all). Or is the about supposed to take in to account all the 2.5 studies? In that case using the wording they have should they not incorporate the uncertainty rage? Which at least for one study (does not the other use uncertainty?!? seams rather pointless just giving a number (and wrong)) is far outside that range?
(and perhaps know with new studies we could throw it out of the window any how?)
Pekka… Mark is trolling. IMO He doesn’t want to recognize what you said (IPCC specified RCP 8.5 as an outlier that is excessively focused on fossil fuels), and he wants to cherry pick, and hyper focus on RCP 8.5.
Heck… even if Mark is completely right, we’re going to hit at least 75% of RCP 8.5, because fracked Natural gas emissions are so high. (Coal plants will convert to Natural Gas easily enough.)
Here’s Ingraffea’s paper natural gas emissions;
http://link.springer.com/article/10.1007%2Fs10584-011-0061-5
Fracked natural gas emissions are currently higher than coal, this is because we vent so much methane on completion. New EPA regulations call for flaring of completions. I’m being incredibly generous when I say 75% of RCP 8.5… I’m assuming the world will adopt EPA regulations. As of now, if nothing changes, we’ll hit full RCP 8.5 with natural gas.
Probably the only real salient point he’s brought up is China’s climate change mitigation plan, namely to stop building new coal power plants.
On the other hand we are finding dirtier and dirtier (read: cheap) fossil fuels, as Dr. Forrester Eli, and I pointed out.
Kjel Aleklett a professor in Sweden that for some years know have been accusing IPCC for similar things that is beeing discussed here… did two blogs on the debate in Sweden… Google translated in the links. Bottom line is it is extremely difficult to predict the price for different sources and to try to figure out political strategies (What will russia do know e.g.?) Will infrastructure be built in time and so on…. any way you can also reed comments from his research team in the comments to the posts.
http://www.google.com/translate?hl=en&ie=UTF8&sl=sv&tl=en&u=http%3A%2F%2Fuppsalainitiativet.blogspot.se%2F2009%2F11%2Ffinns-det-nog-med-fossila-branslen-for.html,
http://www.google.com/translate?hl=en&sl=sv&tl=en&u=http%3A%2F%2Fuppsalainitiativet.blogspot.se%2F2010%2F03%2Fbrist-pa-olja-och-kol-raddar-inte.html&sandbox=1
Magnus,
I haven’t quite worked out what he’s suggesting, but it seems that he’s partly suggesting that there isn’t enough fossil fuels (coal mainly) to ever reach the 2 degree target. That would seem to not be correct. He also seems to be suggesting that we will soon have trouble providing enough energy for continued economic growth. The latter point does seem interesting as I it does seem that fossil fuels will become more expensive as they become harder to extract, and so I do find it odd that there isn’t more agreement that whether or not we should worry about climate change, we should still be working hard on alternative energy sources as we will probably want to find alternatives to fossil fuels anyway.
The discussion is paradoxical independently of Mark’s motivation.
I’m sure, we have many who argue against the severity of climate issue based on the limitations on availability of fossil fuels, but who also support investments in development of technologies that would make presently unavailable carbon resources available. That’s one example of paradoxical thinking.
It’s, however, also true that many people do genuinely consider it almost certain that the limits of recoverable fossil fuel resources are well below the amount needed for RCP8.5. They might agree that the ultimately recoverable resources will be higher in very long run of centuries, but it’s clear that decisions on, how much fossil fuels will be burned in far future are not done now.
It’s sobering to remember that the most immediate decisions are not final. Whichever way we decide now, new decisions will be made at all future times. Some consequences are irreversible, but few are that to a large extent. Some decisions lead also to path dependence, they either open new paths or close others, but those effects are temporary and take often only a few years to reverse.
Postponing action by one or two years means little, but repeating that every year for long may have a large effect. The well known problem of climate policies is that both the warming and the effects of the policies develop slowly and are masked by variability for years, for long enough to build up development that may be irreversible.
Pekka,
Yes, I think your last paragraph sums the situation up really well.
How natural gas compares with coal in CO2 releases depends on many things. The amount of methane leaks from fracking is a controversial question. I don’t believe that claims that they are so large that they would make the total CO2 equivalent GHG emissions as large as for coal is widely accepted. Such claims are made, but also contested.
Another advantage of gas concerns power plants. The efficiency of combined cycle natural gas plants is so much higher than that of coal plants that the CO2 releases are only about 50% of those from coal. Local conditions affect that ratio, but it’s around that value. Natural gas is used also in gas turbines without the steam cycle, but such plants are not directly comparable with coal fired plants, but used for peaking power and other uses where coal is not practical anyway.
Presently 50% may be a more representative value for CO2 releases of natural gas use vs. coal use than 70%, because that’s the value in situations where coal is replaced by natural gas on large scale. There may be somewhat more potential in improving the efficiency of coal plants. Thus the ratio will change in comparisons of best present technology of future times, but in replacement of old coal plants by new gas plants the value of 50% will remain good for long, the ratio may well be even lower than 50%.
Mark still waving his invisible hand, I see.
Pekka still trying to focus on what is relevant.
Thanks, Pekka.
Pekka… I hadn’t considered efficiency. But I do know that we burn 8-10% of natural gas to compress it for LNG ports. Global market = More emissions.
First the important one… it really bugs me the way WG2 and WG3 reports are being used and the writing in them as I mentioned above… no one to bite on it? Is this worth sending to the IPCC for clarification?
About fossil fuels… It is not clear exactly where Aleklett stands just that he has been wrong on several occasions in the press.
The thing that could get wrong here is when the “scarsity” of fossil fuels is used as a reason not to put a tax on it because the problem with climate change will solve it self when the fossil flues get more expensive… which at the moment seams far from certain.
Just to help put the various RCPs into some kind of perspective, here’s a useful visualisation tool that lets you compare each one along a timeline to 2200.
Bearing in mind that the total GHG emission is *not* simply a function of coal burned, we can look at RCP 4.5 vs RCP 6.0 and decide if the invisible hand is waving, or drowning.
Magnus,
Yes, I find that argument a little frustrating. Why would the scarcity of fossil fuels just fortuitously happen to lead to the development of alternatives that just happened to occur at about the right time to prevent too much damage from climate change?
I have just been looking at one of the new studies added by Tol in his correction, specifically Roson and van der Mennsbrughe (2012). Tol records it as showing damages at -4.6% of GDP for plus 5.4 C above the 1986-2005 average.
As it happens, the warmest temperature for which the paper reports damages is +4.87 C relative to 2000 in 2100. Using HadCRUT4, I added the difference between the 1991-2010 and 1986-2005 to align it to the figures used in the IPCC, with the resulting value being +4.95 C. It appears, therefore that Tol has inflated the temperature by 9%.
Further, in the concluding section we are informed that lost labour productivity represents 76% of lost GDP, or -4.6% of global GDP in 2100. That means the net loss from all factors is 6.05% of GDP. Tol reports it as -4.6%, deflating the value by 24%.
I am not certain that Tol has made any error here. He may have used a different temperature series to normalize the temperatures, and I may have misread the comment on labour productivity. It looks to me, however, that he has made two mistakes, both reducing the estimated costs of climate change. Given the overwhelming influence of that one study in reducing estimates of the economic impact of global warming by Tol, these mistakes, if real, are very significant.
“Mark Bahner, are you aware of the technology for Underground Gasification of Coal (UGC). This technology has the potential to add huge quantities of recoverable coal reserves. Of course, the coal will not be physically removed from its source but will still produce the same amount of CO2 has if it had been mined and burnt on the surface.”
I’m aware of the technology. I’ve thought for many years (probably a decade or more) that it was potentially very attractive. It avoids many of the environmental problems of extraction (e.g. removing overburden) and avoids the above-ground gasification equipment. I was not aware until very recently of the potential for UCG exploitation of coal under the North Sea.
But coal is currently being produced worldwide at a rate of more than 7.5 billion metric tons per year. How much of that comes from UCG? When do you expect UCG production to hit, say 1 billion metric tons per year? 5 billion?
And RCP 8.5 projects that coal production will continue to increase throughout the 21st century, reaching more than 40 billion metric tons in 2100 (that’s my figure, based on subbituminous coal…if a substantial portaion of coal production was lignite, the value would be much larger than 40 billion metric tons per year).
Do you think North Sea UCG is going to be producing multiple billions of metric tons of coal in 2100? OK, add those potential countries to the list, and come up with an estimate:
China ?
United States ?
India ?
Australia ?
Russia ?
South Africa ?
Germany ?
Poland ?
Kazakhstan ?
Indonesia ?
Norway?
Britain?
Rest of world ?
Total world ?
What games do you play Mark? I just watched some interesting football matches today and in that game the goal posts are fixed. You seem to like playing a game with moving goal posts, does it have a name?
You started your posts in this thread stating that there was not enough coal to get to RCP8.5 projections for CO2 concentration in 2100. When I showed that there was both sufficient coal reserves and potential technology to satisfy these projections you replied by asking the irrelevant question of how much of the coal was being produced (gasified) by that technology today. We are talking about the future, not today, please stop moving the goal posts backwards.
Ian Forrester writes, “What games do you play Mark? I just watched some interesting football matches today and in that game the goal posts are fixed. You seem to like playing a game with moving goal posts, does it have a name?”
I’m not playing any games.
“You started your posts in this thread stating that there was not enough coal to get to RCP8.5 projections for CO2 concentration in 2100.”
I wrote no such thing. In fact, I made it clear that the issue was *not* whether there is “enough coal to get to RCP8.5 projections for CO2 in 2100.”
I wrote:
“…Dave Rutledge’s point is that–historically–coal regions do not come even close to consuming all their proven reserves. Here’s an excerpt from his many writings:
‘On the other hand, for coal the pattern has been that countries produce only a small fraction of their early reserves, and then late in the production cycle the reserves drop to match the coal at the last working mines. This pattern is seen in the UK (cumulative production of 19% of early reserves), Pennsylvania anthracite (42%), the Ruhr Valley (14%), France and Belgium (23%), and Japan and South Korea (21%). This means that the reserves criteria have been too optimistic, but it also means that world coal reserves are a good upper bound on future production. An IPCC scenario that burns two times or seven times the reserves is utterly at odds with the historical experience.‘ (Emphasis added.)”
Ian Forrester continues, “When I showed that there was both sufficient coal reserves and potential technology to satisfy these projections you replied by asking the irrelevant question of how much of the coal was being produced (gasified) by that technology today.”
I asked how much coal was being produced by UCG today to (gently) point out that the simple existence of UCG–mostly in the research/pilot-scale demonstrations–does not magically convert the coal under the North Sea to “proven reserves.” I could have asked you what the proven coal reserves are for Norway, Britain, and other countries near the North Sea. Both questions would yield the same answer: the idea of UCG, promising as it may seem, has *not* converted all the coal under the North Sea to “proven reserves.” And “proven reserves” are what I’ve been discussing.
Ian Forrester concludes with, “We are talking about the future,…”
Yes, I asked you three questions about the future: 1) when do you think production of coal by underground gasification will reach a billion metric tons per year, 2) when do you think it will reach 5 billion metric tons per year, and 3) what do you expect the production of coal will be in the year 2100 for:
China ?
United States ?
India ?
Australia ?
Russia ?
South Africa ?
Germany ?
Poland ?
Kazakhstan ?
Indonesia ?
Norway?
Britain?
Rest of world ?
Total world ?
What are your answers?
“Pekka… Mark is trolling. IMO He doesn’t want to recognize what you said (IPCC specified RCP 8.5 as an outlier that is excessively focused on fossil fuels),…”
No, that’s clearly wrong. If the “IPCC specified RCP 8.5 as an outlier that is excesssively focused on fossil fuels” then ATTP would have written:
“However, if we choose to do nothing and continue increasing our emissions the evidence suggests that we would reach 2 degrees – relative to today – before 2100. The RCP8.5 mean relative to today is more like 3 degrees by 2100…”
And he is certainly not alone in being fooled by the IPCC. Here are other excerpts that characterize the RCP 8.5 scenario as “business as usual”:
UK Met Office says RCP 8.5 is “business as usual”
“A new type of pathway is being used for the next Assessment Report from the Intergovernmental Panel on Climate Change in 2013 — the Representative Concentration Pathway or RCP. These represent very different views of how the world may look in 2100, with RCP 2.6 showing the effects of strong mitigation and RCP 8.5 the impacts of ‘business as usual’ in which we continue to use fossil fuels with no mitigation.”
USA National Oceanic and Atmospheriic Administration says RCP 8.5 is “business as usual”
A comparison of temperature anomalies from the CCSM4 climate model using RCP 3 (mitigation) and RCP 8.5 (business as usual).
James Cook University says RCP 8.5 is “business as usual”
Select future greenhouse gas emissions
O substantial reduction (RCP 2.6)
O stabilising before 2100 (RCP 4.5)
O stabilising after 2100 (RCP 6)
O high – “business as usual” (RCP 8.5)
University of Washington says RCP 8.5 is “business as usual”
New greenhouse gas scenarios used in IPCC 2013[1][3] range from an extremely low
emissions scenario involving aggressive emissions reductions to a high “business as
usual” scenario with substantial continued growth in greenhouse gases. Although these
scenarios were created in a different way and span a wider range of possible 21st century
emissions, many of them are similar to scenarios used in previous assessments…
> You seem to like playing a game with moving goal posts, does it have a name?
ClimateBall ™.
And, Mark, you’re not scoring so well…
To be clear we’re well on our way to RCP 8.5.
There’s more coal than you claim, and your data states this.
IPCC states that RCP 8.5 is an outsider.
Did you know we burn coal to manufacture petecoke in Alberta? Sound low carbon to you?
AnOilMan writes, “To be clear we’re well on our way to RCP 8.5.”
If that were true, you and everyone else here would be able to easily come up with production in various countries that added up to over 40 billion metric tons in the year 2100. But neither you nor anyone else has.
And what happened to what you just wrote“Pekka… Mark is trolling. IMO He doesn’t want to recognize what you said (IPCC specified RCP 8.5 as an outlier that is excessively focused on fossil fuels),…”
Is RCP 8.5 an outlier, or is it business as usual?
Oops. I forgot the bold closing:
And what happened to what you just wrote “Pekka… Mark is trolling. IMO He doesn’t want to recognize what you said (IPCC specified RCP 8.5 as an outlier that is excessively focused on fossil fuels),…”
Is RCP 8.5 an outlier, or is it business as usual?
Still strawmanning frantically on RCP 8.5. Still ignoring the fact that we don’t need emissions that high for huge problems to arise. Still trolling.
Read the thread, Mark.
Bahner, stop being so disingenuous. There are lots of reserves which could be utilized with exiting or newly developed technology which would allow production of 40 billion tons per year of coal. Just because the production by UCG is close to zero today says nothing about how much could be produced by 2100.
For a simple example of this, answer the following three questions.
1. How much tar sands oil was produced in Alberta in 1965?
2. How much is being produced today?
3. How much is expected to be produced in the future?
Note that the technology for producing the tar sands oil was invented long before it was actually used, 30 to 40 years for the Clark hot water process.
I’ll answer the first question for you – ZERO. I’ll let you do some research so you can answer 2 and 3.
So stop moving the goal posts and stop introducing straw-men arguments. That is typical [redacted] tactics.
Oops, exiting should be existing.
Mark, stop moving the goalposts. You know full well what is meant with “Business as usual”: energy needs primarily satisfied by fossil fuels. It’s the current situation where a very small percentage of energy, worldwide, is generated by non-fossil fuel sources.
Mark Bahner: Your last post is IMO the only real reasonable statement so far. “Is RCP 8.5 an outlier, or is it business as usual?”
That is a fair question.
The answer is that RCP 8.5 is “business as usual” unless we do something about it. And thank you for pointing to China’s solid and concrete efforts to deal with Climate Change. I sure hope we don’t undo that with in situ burning or some such as Dr. Forrester pointed out. Perhaps burning coal to extract petecoke isn’t wise either.
All the data suggests that we will hit RCP8.5 if we continue to develop and utilize low cost fossil fuels. Even if its by other means, ie other dirty technologies.
Many technologies in use today were not possible or in use when the IPCC first started. Fortunately they get to adjust this every 5 years or so to take into account current trend, like fracking. Fracking as is currently being done was invented in 2003 in Calgary Alberta Canada. Its dirty dirty dirty, but it reduced the cost of extraction by a factor of 10, so we use it.
http://www.cbc.ca/news/canada/calgary/multi-stage-fracking-sparked-energy-revolution-1.1338662
Mark Bahner, I suggest you read Pekka’s post about the IPCC again;
“Secondly, availability of large amounts of unconventional fossil resources extends the use of fossil fuels beyond presently extractable reserves (BP 2010).”
Ian Forrester asks:
“1. How much tar sands oil was produced in Alberta in 1965?”–>His answer is zero.
“2. How much is being produced today?”–>1.9 million barrels per day in 2012
3. How much is expected to be produced in the future? –>3.8 million barrels per day in 2022
Answers 2 and 3 per this website: Alberta oil sands
So how about you give your answers/estimates to my questions:
1) When will worldwide UCG production exceed 1 billion metric tonnes per year?
2) When will worldwide UCG production exceed 5 billion metric tonnes per year?
3) What will be the per-country coal production breakdown in 2100 that adds up to over 40 billion metric tons per year?
More moving goal posts. Your questions just show how [Mod: small snip. Ian wants to say incorrect ;-)] your arguments are. My example for oil sands in Alberta in 1965 is where we are now in regards to UCG. No one in 1965 would have been so bold as to predict 5 million barrels per day of oil from the oil sands. Thus no one can say how much energy (coal) will be produced by UCG by 2100.
The whole point of my comments is to show that you are completely wrong when you try and insist that we cannot reach RCP8.5 CO2 levels. The ultimate level for CO2 concentration will only be decided by economics, availability of newer non CO2 technology or politics when the predicted problems become obvious to even diehard [redacted] like you. So stop spreading your FUD on a science blog.
You are just as wrong in your comments on the other Tol thread.
Mark
Still with the straw:
Why *just* coal? The better-posed strawman question would be “all fossil fuels”.
I’m watching you blank me when I point out that your whole RCP 8.5 focus is a crude strawman. I’m watching you ignore every comment I make to this effect. I’m noting your lack of interest in RCPs 4.5 and 6.0 and the centennial consequences of same.
Your behaviour here is poor, Mark.
Pingback: Sense and climate sensitivity, more evidence we are in for hot future | marketspace
Pingback: How Sensitive is Climate? New Paper by Andrew Dessler | Climate Denial Crock of the Week
Twittering on about corrections, http://andrewgelman.com/2014/05/27/whole-fleet-gremlins-looking-carefully-richard-tols-twice-corrected-paper-economic-effects-climate-change/
And another critique. A good one. http://www.synapse-energy.com/Downloads/SynapseReport.2014-07.0.Tol-Climate-Policy.14-075.pdf
jsam,
Yes, very interesting. Quite compelling. Seems broadly consistent with Gelman’s criticisms too.
Pingback: Going for the record | …and Then There's Physics
Y’all will be shocked–shocked, I tell you!–to learn that Richard Tol continues to refuse to acknowledge that his own graph shows effectively no benefit of warming. He is pushing replacement of the quadratic fit with a piecewise linear fit so that the only study showing benefit (his study) dominates. He wrote that the editors of his 2014 correction refused to allow him to replace the quadratic with the linear. I imagine that’s because Tol was perfectly happy with the quadratic when it showed the result he liked, but shopped for any other fit after the result was not to his liking. Tol’s comment is a few down from my comment here: https://wattsupwiththat.com/2016/08/07/politifact-or-polititi-fiction/comment-page-1/#comment-2273324
I think that one was already discussed here some time ago, but I can’t find the thread. Major problem is that the model is predicated on a probably outlier (Tol, 2002), and Richard was unwilling to state exactly what sort of penalty term was used in fitting the model, so the work is not reproducible.
I think it was this thread.
ATTP indeed, your memory is obviously not a short as mine ;o)
No, I can search the comments 🙂
I did try that looking for Richard’s “suck eggs” hubris, which was the only bit I could remember (and that he flatly refused to answer direct questions about the technical methodology and ignored the point about Tol 2002 being a potential outlier).
http://lmgtfy.com/?q=site%3Ahttp%3A%2F%2Fandthentheresphysics.wordpress.com+%22suck+eggs%22
WP search box sucks eggs.
Pingback: Singolare ingannevole – L'archivio di Oca Sapiens