Yes. SRES A1F1 has projected about 600 ppm by 2050 and 1000 ppm by 2100.

1000 ppm is about 2.6x the current level, so taking the theorised logarithmic relationship into account, one would expect about 1.4 times the 2xCO2 climate sensitivity figure, whatever you think that might be.

There’s a lot of “if’s” in that, though.

Of course, if you extrapolate things a different way, you get a different answer. In 1960 the CO2 level was rising at about 1 ppm/yr, and by 2000 it had risen to 2 ppm/yr. The rise is irregular (there’s a drop in the 90s because of the collapse of Soviet Communism) but fairly linear. If you project that to the end of the century, you get about 4 ppm/yr, or if you edit out the Communist collapse, 6 ppm/yr in 2100. That gives a 21st century average of either 3 ppm/yr or 4 ppm/yr, about 300-400 ppm per century. That would put the final level around 680-780 ppm, a factor of 1.8 to 2.1 over present, about 0.85 to 1.1 times the 2xCO2 sensitivity.

You can play endless games. But of course, all of this numerology is akin to the Victorians of 1894 trying to calculate how many horses we would need to power civilisation in 50 years time, and how much dung they would generate. (London would be 9 foot deep in it, it was said.) Oh, how right they were…

http://www.nature.com/nature/journal/v458/n7242/full/nature08019.html

This is just the abstract, but the conclusions are that a total of a trillion tons of carbon emissions (half of which humanity has already emitted) will put us at the 2ºC limit — but the uncertainty band is fairly wide. The 5% to 95% confidence interval is from 1.3ºC to 3.9ºC. Note that the upper end of the confidence interval is extended.

Yes, maintaining an open market for bonds would readily address the long-term problem. I wonder if we couldn’t implement a real-world version of the various prediction markets that are currently used to predict election results and other highly uncertain events? In other words, use the market mechanism not merely to determine probabilities of various events, but to actually fund our responses to them. This is a very interesting concept.

You would put ‘bad data’ second? Interesting.

I’m not going to give a complete discussion, it would take too long. But just for the sake of experiment, try this in a spreadsheet or something.

Set the first cell to a Normally distributed random number with standard deviation 7. In Excel, this is =7*NORMSINV(RAND()).
Next set the following cells to each be 0.99 times the previous cell, plus a unit Normally distributed random number. In Excel, this would look something like =0.99*A1+NORMSINV(RAND()). You need to extend the sequence for a few thousand terms to tell what is going on, but if you start off with a few hundred, that would be fine.
Finally, plot a graph of the data.

This is called an AR(1) stochastic process, and is a member of the ARIMA processes family. There’s lots of theory on these. Box and Jenkins is probably the best textbook on the subject. The ARIMA processes model cummulative quantities. If the heat added or subtracted in a given year is random, then the temperature will be close to last year’s temperature, plus or minus a small random change. That’s not to say that this model is right, (it isn’t,) but it’s not totally without plausibility either.

If you look at the graph, you’ll see that over the short term, it appears to show significant trends. The line goes steadily up, or steadily down, for 50 steps or more. And yet it is pretty easy to calculate the actual mean of the distribution at every point, because it’s the sum of two distributions for which the mean is zero. Yes, the mean is zero at every point, and there is no trend. Extend it a few thousand terms and you can see that directly, but if you have only a couple of hundred, it doesn’t look obvious.

Now AR(1) is not the only sort of sequence with this property, I picked it only because it’s simple to implement. It’s valuable because playing around with it can give a better feel for some of the possibilities that are out there.

On bonds: yes, the idea needs a lot of work, it was thrown together quickly. My point was simply that whenever people have different expectations about the future, or assign different values to something, you can make money. Incidentally, you can solve the long-term aspect by selling the bonds on to somebody younger. As they get nearer maturity, their price converges gradually on their estimated final value. That estimate will depend on the weather, of course.

On Venus: the difference in temperature between clouds and ground *is* due to the greenhouse effect. The problem is, the greenhouse effect doesn’t work the way most people think it does. (Like the same can be said of greenhouses.)

It’s the same reason you see snow on top of mountains. If you think hot air rises, this would seem the wrong way round. Why doesn’t the air from around the base of the mountain rise up the slopes and melt the snow? Why doesn’t the IR radiation from the land around it come into equillibrium with it?

The adiabatic lapse rate, and the fact that it results in a vertical temperature gradient, is a *very* standard bit of meteorology. It explains mountains and clouds being cold.
(Please do look it up. Maybe somebody else could explain it better than I could.)

And on Venus, if you know that the cloud tops are 70 km above the surface and the atmosphere is turbulent, then the lapse rate is easily calculated to be 10 C/km and a temperature difference between the two of about 700 C is expected.

That still leaves open the question of whether the surface will be cold and the clouds even colder, or the surface hot and the clouds cool, but the temperature *difference* is just a matter of basic gas thermodynamics.

I shall look forward to your discussion of feedbacks.

Yes, I would. I don’t think that reasoning applies to the AGW problem, though, because we don’t have multiple and similarly likely hypotheses competing. I see AGW as easily our best explanation, bad data as a distant second, solar variability as an even more distant third, and so on. I perceive the likelihood of all the other hypotheses added up as STILL being less than the likelihood of AGW. That’s the position of the NAS, too. But again, I recognize this as a matter of judgement.

I think that your first funding idea, the after-the-fact tax, seems like a generational heads-I-win-tales-you-lose deal. It says, “Hey, grandkids, have we got a deal for you! We’re going to release all sorts of carbon, and IF it turns out that it causes damages, YOU’LL have to pay a tax for it!” That doesn’t seem very fair to me. My guess is, if we pursue that line of thinking, the kids will just throw out the idea when it comes due.

The second idea (bonds) is much more intriguing. I like the put-your-money-where-your-mouth-is aspect of it. But there are a couple of nasty gotchas. First, there won’t be enough skeptics to buy all the bonds we need to sell. Second, bonds with lifetimes that long won’t appeal to many people — most people with money to invest will be dead before the bonds mature. (This, BTW, is a serious problem with all long-term capitalization issues; it puts a real kink into present-value calculations.)

Let’s apply that idea, for example, to the decision to invade Iraq in 2003. Suppose that Mr. Bush had said, “OK, all you nitwits who think that Mr. Hussein doesn’t pose a threat to us, put your money where your mouth is. We’ll forego invading Iraq if you guys post a bond of $1 trillion to cover the losses incurred by an Iraqi attack. If there’s no Iraqi attack in 20 years, we’ll pay back your bond at 10% interest rate. But if there is and Iraqi attack, all your money goes to cover the costs incurred by it.” Boy, I would have signed up for that one. But there’s no way we could have found enough underwriters to pay for that bond. And yet the bond you propose would last longer and require lots more money. Although in principle it’s a splendid idea (I *REALLY* like the put-your-money-where-your-mouth-is aspect), it just wouldn’t work in practice. But I’m going to mull that one over. It might be possible to come up with some variation that might work.

On the Venus problem, I don’t think we’re making any progress on explaining the difference between the surface temperature and the cloudtop temperature. I still think it’s due to greenhouse effect, but I’m not going to press the issue.

I was going to start discussing that matter of climate feedback effects, but while checking some facts I ran into a surprise that I need to sort out before making my claims. I’ll have to defer that until tomorrow.

Too bad, punk.

Better luck next time.

brian.valentine@hq.doe.gov at work

if anyone cares to write to me

And saying Good-Bye! for now on Chris’s web log

(hey – want to have some fun? Read Chris’s Storm World. It’s a trip and a half.)

Conceding *any* uncertainty in the science behind AGW is more than a lot of its supporters would do, and I respect that. We can agree to differ on the *degree* of certainty, if necessary, but it’s a healthier framework for the debate in my opinion.

I think that with your black-and-white choice, you might be thinking in Pascal’s wager terms – of thinking that with two alternatives and no information, that the default should take them to be of roughly equal probability.

Let’s take a trivial example of throwing 2 normal six-sided dice, and guessing the total. This is subject to uncertainty, but the hypothesis that the total will be 7 has the greatest support for it. If we are asked to choose between the hypotheses 2, 3, 4, …, 11, 12, then 7 is the best bet. However, if we are asked to choose between the hypotheses ’7′ or ‘not 7′, then the first hypothesis has probability 1/6, and the second 5/6. It is far more likely to be something other than 7. Picking 7 when we have the opportunity to pick “We don’t know” is the wrong choice.

Whether or not you agree that scientific confidence in AGW might be so low, would you agree that simply being the best of the hypotheses available isn’t *necessarily* sufficient to justify believing it?

On your tax proposal, how would you feel about a carbon tax linked to sea level or something like that? We pay 1% for every centimetre the sea rises past 50 centimetres. Then anyone who doesn’t believe a word of it can happily sign up, and if you’re right, it ought to kick in with a vengeance within a few decades. Not as good as you wanted, perhaps, but better than nothing, yes?

Or if you want to raise money for your measures faster, then how about climate bonds, that mature with a good interest rate in 40 years only if there is no sign of catastrophe? Sceptics can buy them from you, you can build windmills with the proceeds, and if we’re wrong, we end up paying for them. Good deal, right?

Sorry, slightly confusing thermodynamics usage: in thermodynamics, heating is a process, it has symbol Q. It’s different from the internal energy U which is proprtional to temperature.(http://en.wikipedia.org/wiki/Heat)

The source of heat at the surface is incoming sunlight. You may recall I said: “You need enough heat flux to get through to the surface to drive convection, but if 5% of the light was sufficient to do that,…”

That’s because they have none for themselves.

I respect Eamon, for his brave commitment to stay in his homeland amidst unspeakable violence and terrorism despite the threats to him and his family,

and for the simple fact that he simply stated his name.

da ta dah ta da

blah blah blah blah blog blah

It isn’t what I entend to do with it it is what you DON’T do with your own neame – which is to use it to sign what you write and and stand up for it like a man or worman or anything else but a weenie

But the more interesting question is, why do you care what my real name is? What do you intend to do with it?