## Wednesday, February 16, 2005

### Kyoto takes effect (update)

The Kyoto protocol comes of age. Finally. How much will it - and its hypothetical future extensions - cost? Let's talk about the catastrophic scenario in which people won't abandon this weird international treaty and they will really try to stabilize the CO2 concentrations in the atmosphere.

Reuters published estimates that it would cost 18 quadrillions (18 times 10 to the 15th power):

If you don't know what a quadrillion (U.S.) means, it is roughly speaking an infinite number, and you may object to such an unrealistic result and label the news from Reuters as a kind of typo. What about the estimates from the real people behind the Kyoto protocol?

These bureaucrats and scientific bureaucrats are organized in the IPCC - the same institution that has claimed, for several years, that the temperature record in the last 1000 years looks like a hockey stick, without checking the statistics behind these claims. So what does the IPCC say about the costs of stabilizing the CO2 concentrations? Let me start with some elementary data.

The pre-industrial concentrations were about 280 ppm (parts per million, counted as volume) - the usual peak for recent interglacials. Incidentally, during the ice ages the concentration fluctuated around 180 ppm. Let's try to return further to the past.

When mammals were created 100 million years ago, the temperatures were 10 degrees C higher and the CO2 concentrations were about 3,000 ppm, eight times as high as today. The humans appeared during the "previous global warming" three million years ago in the middle Pliocene - it had to be a horrible time if monsters like humans were born; the concentrations were 380 ppm just like today and the temperatures were 2-3 degrees C higher.

Once again, the current concentration is 380 ppm and it rises by a few ppm every year. Some people want to claim that 400 ppm is already too high. We can't count the costs of stabilizing the concentrations below 400 ppm because the theory is not renormalizable. ;-)

OK, so the first realistic number that the IPCC considers is to stabilize the CO2 concentrations below 450 ppm. See the graph

which is found as 7.3 on the page

Thanks to a reader of my blog who pointed out this page. What does this particular graph say? In order to stabilize the concentration below 450 ppm, one needs to pay 400-1800 trillions of 1990 U.S. dollars by 2100, according to different models. What would such a number mean?

Note that the 1990 U.S. dollar is more than the present one. This number amounts to more than 10 trillion USD per year. That's roughly the size of the U.S. economy (GDP), 20 percent of the world's economy.

The United States are reasonable enough not to participate in the main Kyoto activities (although there are similar programs at the level of the states), so let's talk about the real signatories. The cost of "Kyoto 450" as described in the graph would be equivalent to nuking the whole territory of Japan, Great Britain, Germany, France, and several other countries every year so that it's guaranteed that life can't continue until 2100. (You could first evacuate the people from these countries.) Note that if you only nuke the European countries in this list, the costs will be lower than "Kyoto 450". I am not proposing it, just listing the fact that according to the graph, there would be alternatives to Kyoto that would have a similar effect.

However, as a leading climate expert W.S. from Harvard who read my blog has pointed out, these IPCC numbers on the graph are bogus. The IPCC later published an "important correction"

On the main page, you will find this "important correction" with a new PDF file where the expenses are lowered by two orders of magnitude: 18 trillion. Note that I am not the first victim of the blunder: Andrei Illarionov as well as all readers of this blog (except for W.S.) as well as the participants of the sci.environment newsgroup thought that the estimate 1,800 trillion was really an authentic estimate done by the IPCC. No, it was not. Nevertheless, the number is not calculable - it's a guess. 18 trillion USD is not a big deal - it's just stopping the whole U.S. for two years or so. My estimates are higher; they are based on counting the expected slowdown of the global economic growth.

Incidentally, now I guess that the report from Reuters that talks about 18 quadrillions is derived from the same graphs (note that the number starts with 18 again), and the author (Aliston Doyle) has made yet another numerical error in his counting of orders of magnitude. Note that once these discussions try to become quantitative and factual, no one has an idea what the reasonable numbers should be and two orders of magnitude don't really matter. People still call it "science"...

Figure 1: Taiwan's emperors of the Kyoto protocol have new clothes. Or maybe they're preparing for the after-Kyoto living standards. Click the picture for more information.

1. "The United States are reasonable enough not to participate, so let's talk about the real signatories."

Well, you are misinformed.

http://www.sfgate.com/cgi-bin/article.cgi?file=/chronicle/archive/2005/02/16/EDG49BAVBP1.DTL

...Nowhere is this truer than in the automobile industry. The countries that have signed Kyoto bought more than 60 percent of new cars in 2002. Although the U.S. government has not signed onto the agreement, many individual states are aggressively pursuing reductions in greenhouse-gas emissions. California has taken a lead role, adopting an emission standard for vehicles that will take effect in 2009. The seven Northeastern states that already follow California's current emissions standards, including New York and Massachusetts, have indicated they intend to adopt the new rules. Together, these states represent more than 25 percent of all new car sales in the United States....

....Regardless of your position on global warming, there is no doubt that the Kyoto Protocol and other efforts to curb greenhouse gas emissions will affect the way companies do business around the world. The companies that invest in new technologies will be best equipped to compete in the global markets of the future....

2. Lubos,

could you give me a hint, how these numbers are calculated? It seems hard to believe that you sum up costs in that order of magnitude if you motivate people that the next car they buy has a lower consumption (e.g. by increasing the tax on gas, god forbid) and to spend a bit more on insulation (it might already pay if the AC has less to do in summer...) when you build a house. As far as I know that would already amount for a substantial decrease in CO2 emission.

My other question concerns all your posts on climate issues: I agree with you that doubts about interpreting recent weather events as indicative for an observation of climate change are in order.

However, I would be interested if you could agree with the following statements: Climate is described by a set of coupled non-linear partial differential equations that describe many different observables (wind, rain, temperature of air and sea, currents etc) of Navier-Stokes type. Equations of that type are known to have many meta-stable solutions. It is unknown which concrete pertubations could trigger the transition to another metastable state (in popular speak: stop the gulf stream). However, if such a transition happens on a short timescale (compared to evolutionary timescales, say order of 100 years) it is not unlikely that results for vegetation are (locally) desastrous (again on that timescale). If you like you can also assign costs to these results.

Don't you agree that given that, we should make some effort not to trigger such a transition. And not changing CO2 concentration on short timescales might be a likely measure to that effect (even if we even don't know if an increase of CO2 results in a fall or raise in temperature)?

I would say, we cannot prove that at concentration x the world climate will change, it's a save bet that we should better not try to map out the complete domain of stability of our current metastable state.

3. Yes, I am aware of the activities in the U.S. that try to reduce the CO2 emissions at the state level. My guess is that they will be much less harmful for the economy than the "federal" activity would be.

Robert: I don't know exactly how the IPCC calculates the numbers. But you know, you must simply reduce the total CO2 emissions by some amount that is known from the simulations, in order to keep the concentration below some limit.

Once again, this means that you know how much CO2 you need to save. The last thing you need is to know how much money you pay for reducing the emissions by 1 ton. You know that this number, according to the Kyoto mechanism, is a subject of market economy, and the estimated price is tens of dollars per ton of CO2. So my guess is that the costs may be just calculated as the product of the price and the number of tons.

If you calculated how much more the companies and individuals pay for the same products whose only difference is their treatment of CO2, you would probably get a comparable number. Sorry, I am not sure about the details. Even if I were sure, the total estimate is just a guess anyway.

It's a good shift in the discussion to talk about the factors that could stop the Gulf stream. Yes, the differential equations are the best tool we have to analyze these possibilities. I don't think that there exists a controllable calculation that correlates the CO2 emissions with the probability of stopping the Gulf stream.

But I agree with your general philosophy that the local events should be those that we're really thinking about and planning. If we think about the sea levels, it affects the Netherlands as the #1 country, and one can see that the NL opinion about the sea levels is very different.

I think that the history of Earth has already mapped the region we're mapping again. It's tried 3,000 ppm of CO2, 10 C higher temperatures, and mammals were happy.

Frankly speaking, let me say that stopped Gulf stream would not have to be a universal - or Euro-scale - disaster either.

4. OK, let's get some grip on the numbers: According to Wikipedia the us emits 5x10^9 tons CO2 per year. The protocol asks for a order of 5% reduction. So assuming the 10$/ton cost does not lead to your number (by several orders of magnitude). Of course, assuming linear cost is quite wrong: From TV this morning I learned that Germany achieved a substantial part of their reduction goal virtually for free by phasing out the brown coal based powerplants in east Germany and replacing them by state of the art power plants but that the remaining reduction would be more expensive. Regarding the question of mapping the domain of stability: I think the timescale is quite important here: If things happen over a period of some centuries than migration would be much less of a problem than the need of say evacuating Bangladesh in a couple of years. Another example of what has been mapped out is the ice age: True, mankind has survived that but freezing up northern and central Europe even over a period of 100 years would probably change a number of things. So, I do not quite agree that all parameter space probed so far is safe. Plus the non-linearity of the equations that are possibly not far from a chaotic regime means that we could just be unlucky and hit a bad combination of parameters with which similar parameters would have been save. But this just shows how difficult predictions in this area are no matter whether they predict we don't have to worry or that doomsday is close... 5. OK, let's get some grip on the numbers: According to Wikipedia the us emits 5x10^9 tons CO2 per year. The protocol asks for a order of 5% reduction. So assuming the 10$/ton cost does not lead to your number (by several orders of magnitude). Of course, assuming linear cost is quite wrong: From TV this morning I learned that Germany achieved a substantial part of their reduction goal virtually for free by phasing out the brown coal based powerplants in east Germany and replacing them by state of the art power plants but that the remaining reduction would be more expensive.

Regarding the question of mapping the domain of stability: I think the timescale is quite important here: If things happen over a period of some centuries than migration would be much less of a problem than the need of say evacuating Bangladesh in a couple of years.

Another example of what has been mapped out is the ice age: True, mankind has survived that but freezing up northern and central Europe even over a period of 100 years would probably change a number of things. So, I do not quite agree that all parameter space probed so far is safe. Plus the non-linearity of the equations that are possibly not far from a chaotic regime means that we could just be unlucky and hit a bad combination of parameters with which similar parameters would have been save. But this just shows how difficult predictions in this area are no matter whether they predict we don't have to worry or that doomsday is close...

6. Robert said: "OK, let's get some grip on the numbers: According to Wikipedia the us emits 5x10^9 tons CO2 per year. The protocol asks for a order of 5% reduction. So assuming the 10\$/ton cost does not lead to your number (by several orders of magnitude)."

Even the 5x10^9 tons CO2 per year figure is doubuious. That would be about 50 tons per household, or 4 tons per household per month. We know the biggest energy consumption is usage of gasoline to drive cars. There is no way an average family consumes a few tons of gasoline per month. Plus I do not think the US production or importation rate of coal and petroleum reach any where near 5x10^9 tons per year. So the 5x10^9 tons figure was probably exagerated by 2 or three orders of magnitude.

The earth had experienced far higher concentration of CO2 than 4000 ppm. Pre-life earth was 20%+ Co2 and virtually no oxygen. Only because of the occurence of plantation it turned the 20% CO2 into 20% of O2.

Quantoken

7. I have always thought "climate science" is really an illfounded field full of alarmists and scaremongers. A guy I graduated with got into this stuff and for his phd started using supercomputers to model climate related things, when they were meant for more sensible stuff like lattice qcd. There is no real evidence that humans are affecting the climate or whether they even can. CO2 concentrations reached a maximum during the Cretaceous Thermal Maximum, some 65 million years ago, a time of massive volcanic activity arising from very active plate
tectonics/continental movement. In particular, what we know call India was really just an ocean of basaltic lava, poisoning the atmosphere. What is left of it is called the Deccan Traps, a basaltic lava slab about 2000m thick and covering about 500,000 km squared in India. However, the lava sea would have covered about 1.5 million km squared during the Cretacious pouring colossal amounts of CO2 into the atmosphere.(Incidently, basaltic lava floods are believed to occur on Venus too). Life on Earth, dominated by the dinosaurs, had a particularly hard time but the planet survived.
(It certainly helped push the dinos over the edge). I don't think our CO2 emissions amount to very much in comparison. The earth can cope (provided we don't cut down all the forests and jungles, which is the more important issue)

Another thing climate scientists and alarmists seem ignorant of is helium-3 mixing in the Sun's core and its effect on solar luminosity. This is the part of the proton-proton cycle that requires the highest temperature. Helium-3 can be produced in "cooler" regions outside the core but the fusion of He-3 to make He4 only takes place at the highest temperatures in the core. This means it might accumulate for a while around the core until enough forms then it sinks into the core and gets burned. This shifts the luminosity output of the sun. Intervals between theoretical solar core mixing and intervals between Ice Ages on Earth seem to match. Certainly even small changes in solar luminosity could affect global climate and will have a bigger impact than any CO2 we can produce.
Computer climate models simply can't take this into account.

Also there have been many climate fluctuations during human history. The Medieval Warm Period from about AD 1000 to AD 1400 is well documented. Wine grapes were grown 300 miles north of where they currently are, and Vikings made it to Greenland since a lot of ice had melted. This was obviously a pre-industrial climate change and we may be just about to go through another one.

Steve M

8. I have always thought "climate science" is really an illfounded field full of alarmists and scaremongers. A guy I graduated with got into this stuff and for his phd started using supercomputers to model climate related things, when they were meant to be there for more sensible stuff like lattice qcd. There is no real evidence that humans are affecting the climate or whether they even can. CO2 concentrations have been much higher in the pre-industrial past and reached a maximum during the Cretaceous Thermal Maximum, some 65 million years ago, a time of massive volcanic activity arising from very active plate
tectonics/continental movement. In particular, what we know call India was really just an ocean of basaltic lava, poisoning the atmosphere. What is left of it is called the Deccan Traps, a basaltic lava slab about 2000m thick and covering about 500,000 km squared in India. However, the lava sea would have covered about 1.5 million km squared during the Cretacious pouring colossal amounts of CO2 into the atmosphere.(Incidently, basaltic lava floods are believed to occur on Venus too). Life on Earth, dominated by the dinosaurs, had a particularly hard time but the planet survived.
(It certainly helped push the dinos over the edge). I don't think our feeble CO2 emissions amount to very much in comparison. The earth can cope (provided we don't cut down all the forests and jungles, which is the more important issue)

Another thing climate scientists and alarmists seem ignorant of is helium-3 mixing in the Sun's core and its effect on solar luminosity. This is the part of the proton-proton cycle that requires the highest temperature. Helium-3 can be produced in "cooler" regions outside the core but the fusion of He-3 to make He4 only takes place at the highest temperatures in the core. This means it might accumulate for a while around the core until enough forms then it sinks into the core and gets burned. This shifts the luminosity output of the sun. Intervals between theoretical solar core mixing and intervals between Ice Ages on Earth seem to match. Certainly even small changes in solar luminosity could affect global climate and will have a bigger impact than any CO2 we can produce.
Computer climate models simply can't take this into account.

Also there have been many climate fluctuations during human history. The Medieval Warm Period from about AD 1000 to AD 1400 is well documented. Wine grapes were grown 300 miles north of where they currently are, and Vikings made it to Greenland since a lot of ice had melted. This was obviously a pre-industrial climate change and we may be just about to go through another one. We will have plenty of time to adapt to it and I dont think Kyoto will have much (if any) of an overall effect anyway.

Steve M

9. Sorry, I accidently posted this twice. The second one is the complete one.
Steve

10. In my opinion, there is a chicken-and-egg problem. Solar cell power at its cheapest costs about 25 cents per kilowatt hour, while conventional sources cost around 10 cents per kilowatt hour. But presumably, if solar cells were a mass market product, the cost would come down significantly.

A sun-drenched country like India, despite all its economic growth, simply does not have the buying power to create the mass-market for solar cells; but if the mass-market started off in the First World, then India could ride the bandwagon. That is, I think there is an enormous market that would open up for solar power, if the market could be kickstarted to achieve a scale where costs were reduced.

The open question is how to do this.

----
I see claims like, for the United Kingdom, "Between 1990 and 2002 the output of the economy increased by 31 per cent whilst energy consumption only increased by about 6 per cent."

Here are GDP(PPP) per unit energy consumption:
http://hdr.undp.org/statistics/data/indic/indic_200_1_1.html

So, IMO, it is not at all obvious that one has to sacrifice economic growth to achieve Kyoto goals.

11. In my opinion, there is a chicken-and-egg problem. Solar cell power at its cheapest costs about 25 cents per kilowatt hour, while conventional sources cost around 10 cents per kilowatt hour. But presumably, if solar cells were a mass market product, the cost would come down significantly.

A sun-drenched country like India, despite all its economic growth, simply does not have the buying power to create the mass-market for solar cells; but if the mass-market started off in the First World, then India could ride the bandwagon. That is, I think there is an enormous market that would open up for solar power, if the market could be kickstarted to achieve a scale where costs were reduced.

The open question is how to do this.

----
I see claims like, for the United Kingdom, "Between 1990 and 2002 the output of the economy increased by 31 per cent whilst energy consumption only increased by about 6 per cent."

Here are GDP(PPP) per unit energy consumption:
http://hdr.undp.org/statistics/data/indic/indic_200_1_1.html

So, IMO, it is not at all obvious that one has to sacrifice economic growth to achieve Kyoto goals.

12. lumo: "they will really try to stabilize the CO2 concentrations in the atmosphere." CO2 will eventually stabilize, the question is at what level.

The question of Kyoto cost strikes me as silly since we are discussing different positive powers of exponential growth of economy - whatever turnover we want to achieve we will eventually get there.

It isn't a matter of _what_ we want to to do, only _when_ we want to do it.

Another stupid argument is questioning if Kyoto will achieve stopping (the heretofore non-measurable) GW effect. Of course not, it was intended as a starting diplomatical, economic and technical effort and the main blockage, US, has since left the process.

But is still in, despite arguing the opposite, since the 6th economic power, California, is within the Kyoto protocol demands...

_If_ GW will start, the current _verified_ trend of killing of 25 % of the species within the next two generations will apparently double (BBC News 2004).

Not only will we have to tell our grand children what humanity is one of the major life disasters on Earth.

We will also unnecessarily exponentially increase our own risk of extinction, assuming ecological diversity is essential for its survival, and some other life is essential for _our_ survival. It was, last time I checked...

Contrary to capital, this resource is non-reversibly wasted. Maybe the question is, should the price be infinite on such resources?

if so, even an infinitesimal risk of increasing the loss would be _really_ stupid.

Torbjorn

BTW, quantoken, I have posted a belated reply to your post at 'Peace in Iran, LHC, and other news'.

13. Regarding changes in solar activity: When the solar neutrino deficit was not as established as it is now, I once learned that the photons from the solar core take about one million years to diffuse to the sun's surface.

Thus any change in the inner workings of the sun is smoothed out over that timescale. So, this is no need to worry for me right now.

14. torbjorn:
I looked at the link you gave regarding "secondary atmosphere". It does not contradict what I said. Actually it supports what I said. Note what it said: "Oxygen came into the atmosphere as a waste product from the photosynthetic activity of early life forms" So it, too, agrees that today's O2 came from photosynthetic activity, which means it came from CO2. Please note I was talking about "pre-life" atmosphere and they were talking about primodal "first atmosphere". The two concepts are different.

Now, back to GW. I do not think GW is a viable theory to worry about. But we do have far more imminent and serious problems to worry about. Problems like depletion of useful natural resources and man made species (including human!!!) disstinction scale catastropies. The nuclear weapons along could wipe out human race many times over.

Martin Rees concluded the human has only 50% chance to survive the next one hundred year. I am less persimistic. But I want to point out that even a survival rate of 99% over the next 100 years, still does not look very good.

Let's do some calculation. We human today are evolved from pre-human apes a few millions of years ago. Let's count human history from the point when human first learned to use fire. That's the beginning of human capabile of making man made disasters. That history is approx. 1 million years.

Countless natural and man made disasters occured in early history of mankind, many threatened the very survival of human, like ice age, volcano, tunami, large meteror hitting earth. Highly contageous and deadly diceases like anthrox wiping out whole nations and there was no medication available, etc. etc.

Early humen were very fragile, don't you agree? Now let's say once in a hundred year, there were a huge disaster that could potentially wipe out the whole human race. (think about the black death that wiped out 2/3 of Europe for example). Let's be a little bit optimistic, say humen would have a 99% chance of survival, and only 1% chance of not make it.

A one million year history is many hundred years, 10000s of one hundred years to be exact. Now, multiply 99% repeatedly 10000 times, see how good a chance human would have survived 10^6 years?

The chance is astronomically small. After you multiply 99% for 10000 times, the odd of human survival is 2x10^-44. That's 44 decimal places AFTER the decimal point. It's practically zero for all purposes. There are NOT even 2x0^44 many solar systems in the whole universe!!!!

But we do beat the odds and survived a million years! What does it tell us? What it means is the early human have FAR MORE chance of survival than current generation, because they did not have nuclear weapons and other terrible man made catastrophies to worry about.

It's a shame that we now need to calculate the chance how we are going to survive the next 100 years, not the next one million years!!! As I said, even if we have a 99% survival chance for the next one hundred years, which is very unlikely and much more optimistic than Martin Rees's 50% chance. It still does not look very good if we look down 1000 or 1000000 years down the road ahead.

Quantoken

15. Rather than focus on the trillions, or the opposite extreme (do nothing), I think we should take a shot at the easy ways to reduce CO2.

Let me add one thing I can do without, one thing that won't really deprive me:

In one of the more upscale supermarkets locally, we have half dozens of free range eggs. But these aren't just any eggs. They are brown, but that isn't the kicker ... no, they are flown from New Zealand to California each week!.

See, that stuff is ridiculous. I think it is fine to have eggs, free range eggs, brown eggs, and even New Zealand eggs ... but flying them to me half-way around the world is just crazed.

There is so much crap like that that would just fall out of the equation if people just started thinking about global warming and fuel.

But for that to happen, I think the "do nothings" are going to have to shift to "do the easy things."

(john in california)

16. quantoken: "... today's O2 came from photosynthetic activity, which means it came from CO2."

I don't think that follows since photosyntesis/metabolism uses/produces water from O2 and hydrocarbons. And carbon is added from other sources, see the other posts. It isn't a tightly closed system.

Your description of the three generations of Earth atmosphere seems confused; see Wikipedia for an explanation of this.

Returning to your risk description, it is interesting but you have forgotten that we also improve our ability to understand and handle risks. As the current discussion shows.

I agree with the conclusion though; if there is a miniscule risk for a threatening situation we must nevertheless see to it that we minimise it.

If GW exists it will apparently increase the extinction rate considerably.

The current species extinction will both be imprinted in the fossil record to our eternal shame and increase the risk for biotope and sooner or later human extinction.

17. The moderate posistion that global warming is happening, and that it has a human-caused component, got some good press today.

Two summaries of recent events are available at Wired News and Worldchanging.org.

I think hardliners should move from the weak posistion that all this is false (some kind of global conspiracy theory), and to what kinds of mitigating actions make sense to them.

For me, improved efficiency is the most logical first step. We have new technologies that continue to give us better quality of life at lower energy costs. Lets, go ...

18. Some people could accuse me that I've written the previous anonymous comment (whose most frequent word is "postition") in order to show how difficult it is for the alarmists to use their brains.

No, this comment is completely authentic, believe me.

I don't accept the weird rules of the game outlined in the anonymous comment, and I disagree with the assumption that we should be paying for "mitigating actions".

Also, I added the Kyoto counter to the right sidebar of this blog, to show how breathtakingly crazy the Kyoto-like arrangements are.

19. You know what's really funny? In that anonymous post above I didn't actually mention paying for anything.

Improved efficience saves us money.

The fact that political zealots still think this a cost, proves my point. Talk about closed minds!

20. Dear Anonymous,

my understanding is that in your latest work, you have exposed your belief that the "mitigating actions" don't cost anything.

Look at the counter on the right if you want to see whether it's true.

Your opinions about economics are not a too promising starting point for a rational debate - but it may be a good enough point to call medical assistance for you.

Besides "position", you should also try to learn how to spell "efficiency".

Best
L.M.

21. I'm sure many mitigations do cost money. I am even confident that there are people out there with ideas about how to spend trillions.

The thing is, that fact that some of them cost money does not imply that all of them cost money.

I'm sure, if you unlimbered your mind, you might think of some that don't cost money as well.

BTW, I hope you didn't try to put a block on my comments. That would make you look just too weak.

22. oops, I see the comment problem was a glitch on my side. sorry.

23. Lubos, do you not want to subscribe to the State of Fear that Michael Crichton postulates, is running the American resistance to the Kyoto Accord?