U.S. Dept. of Ag report: Climate change is damaging America . . . now

Would you happen to have a link for that? I would really like to know what observations and measurements are the basis for the phenomenon you claim is “documented.”

I have found that Table 8.2 in Chapter 8 of the latest IPCC (Working Group 1) report lists 23 models and has equilibrium climate sensitivity values for 19 of them. Here are the numbers I get using this: the average sensitivity is 3.21 with a standard deviation of 0.69 C. Hence, we would get a 1-sigma result of 3.21 C ± 0.16 C or a 1.65-sigma result of 3.21 C ± 0.26 C if we assume that the standard error was the correct thing to use for the uncertainty. Again, this is clearly much smaller than the range that the IPCC quotes for an uncertainty that is somewhere between 1-sigma and ~1.65-sigma.

Note if we assume that the standard deviation is a better measure, then the 1-sigma result is 3.21 C ± 0.69 C and a 1.65-sigma result is 3.21 C ± 1.14 C. So, clearly the IPCC statement of the equilibrium climate sensitivity being likely (66-90% chance) of being between 2 C and 4.5 C is much closer to what one gets if one assumes the standard deviation, not the standard error, provides a reasonable estimation of uncertainty. (In fact, the standard deviation is still small relative to the IPCC error estimate if you assume it is a 1-sigma result…but is pretty much right on the nose if you assume it is a 1.65-sigma result. This may be coincidence since I don’t think the IPCC explicitly came to this estimate by just looking at the spread in the models…I think they relied more on estimates derived from studies that look at the best estimates of climate sensitivity one gets from current climate or past climatic events.)

To claim that the IPCC would argue for using the standard error in comparing models to experiment is thus completely and utterly lacking of any foundation (even if one ignores the issue of the unforced climate variability). It is basically setting up a strawman of the highest order.

The sequence of events went like this:

Someone asked the IPCC to release their review comments. The Review Editor comments were also requested.

Despite their own policy, the IPCC refused.

So, a Freedom of Information Act request had to be filed. The IPCC finally agreed to release them. It was not, as your citation claims, a change initiated by the IPCC to be more open. They were forced to open up their deliberations, and simply put the best spin on it to make it seem like their idea … and their spin was successful, if your reaction is any indication.

How in the world you can twist this around to think that the IPCC are the good guys in this is beyond me. They tried to hide their review deliberations, despite their stated policy. We, the public who is paying for these shlubs to play scientist, had to use the FOIA to force them to open their deliberations up. You seem to think that’s all wonderful. Me, I think it sucks. They got caught with a shabbily run, poorly documented review process, and tried to hide it. They persisted until legally forced to reveal it. What’s wonderful about that?

There are a host of threads on this, and other IPCC oddities, located here. I stongly suggest that those of you who think jshore is correct about the integrity and honesty of the IPCC have a long, leisurely read of them …

Finally, I’ve haven’t a clue which “actual scientific point” you made that you are referring to, but if you’d like to repeat it, I’m more than happy to discuss it.

My best to you, and to everyone,

w.

I’m talking about the points that I discuss in posts #157 and 162 regarding the use of standard deviation vs standard error to represent the spread in model results when comparing to experimental data (e.g., as is done in the Douglass et al paper).

My best to you.

Well, here’s an excerpt from Chapter 7 of the IPCC Climate Change 2001 Working Group I Section 7.2.1.1

The fundamental numbers behind this are described on the page of Baird and Cann’s text I cited above:

From Environmental Chemistry, by Baird and Cann, 3rd Edition, page 185. I understand it is difficult to find some sources online, but you can search for the book in a library near you here. ILL works wonders, too.

I don’t get this - I have never heard the term “CAGW” before this thread and now you’re suggesting that I should define it? :confused: I suggested the dictionary definition of “catastrophe” because yours seems overly specific and counter-intuitive - for example, why is it only a catastrophe if deaths are weather-related? I would classify the 1918 Influenza epidemic as a catastrophe, although it does not meet your definition. I think you’ve placed unnecessary restrictions on the word “catastrophe” because you’re attaching it to AGW.

You must have read too quickly - my cite was in that paragraph:

(Bolding added for emphasis)

You may not be able to access Climate Dynamics, and that’s OK. There are similar numbers in the IPCC’s Fourth Assessment, in Chapter 3, look at figure 3.10 on page 229.

In that figure, the stabilization level of between 440-485 ppm CO[sub]2[/sub] gives a global average temperature increase of 2.8-3.2°C.

Would you like to share your reasoning behind this?

Thanks, jshore.

For those who are not following the dialogue, there was a paper published recently by Douglass et al. The subject matter was the temperature rise in the tropical troposphere. The paper showed that the estimates of all of the climate model were much, much higher than those of the balloons and the satellites. A number of AGW supporters use this to argue that the data must be wrong.

The question under discussion was whether the correct metric for determining whether the models and the observations are statistically different was the standard error of the mean of the model results, or the standard deviation of the model results … thrilling stuff, I know, but bear with me.

The underlying question (not the statistics, but whether the data is correct, or the models are correct) is important because the AGW hypothesis requires increased heating of the tropical troposphere. That is to say, AGW theory predicts that if the tropical surface heats up, the tropical troposphere will heat up much more.

However, the data shows that, far from heating up much more, the tropical troposphere is heating up less than the surface.

For me, that’s enough difference right there. You can take all the fancy math you want to dispute it, but the simple fact remains. The AGW hypothesis requires that the tropical troposphere heat up much faster than the surface. The observational data (including two separate balloon datasets, and satellite data) show the opposite, that it is heating up slower than the surface.

The modelers response to this is “the observation must be wrong”. The difficulty with that argument is that both the RATPAC and the HadAT balloon data, as well as two groups interpretation of the satellite data (RSS and UAH) all disagree with the models. The only way this could happen is if something is wrong with two different sets of balloon data, and two different group’s interpretations of the satellite data, and all four groups errors are in the same direction (cooler) … which seems very unlikely.

This difference is starkly illustrated in the NAS report. It shows how the models give a hot spot in the tropical troposphere, while the data shows a cool spot. I really don’t care which method you use to calculate the difference between the models and the data, because models say heating, and observations say cooling. Both the NAS report and the Douglass study show this quite clearly.

jshore would have us believe that because a couple of the models are kinda close to the data, the models are correct … that is, as long as a couple of the models overlap the data, the models have included the data, and thus there is no problem with the models.

This is a curious point of view, because the more inaccurate the models are, the more likely it is that one or more of them will overlap the data … and thus the group of models will be declared to be correct, because they include the data.

However, to me, the idea that the more inaccurate the models are, the more likely it is that they are correct, seems … unusual. But as I say, the exact details of the math don’t matter to me. It is the fact that the signs of the changes are different, models increasing, and data decreasing.

This issue has raised a lot of heat in the blogosphere. Lubos Motl discusses it here, and shows the predicted hot spot versus the data here. David Douglass discusses it on Climate Audit here. It has also been discussed on RealClimate, which you’ll have to find for yourself, since I wouldn’t give those censorious slimeballs the time of day. In addition to censoring anything that they disagree with, in this instance Gavin Schmidt attacked Douglass’s scientific integrity on RC, then didn’t have the balls to either admit his mistake or apologize when Douglass called him on it … typical of the charming fellow. You can read that interchange here, starting at post 79.

I encourage y’all to 1) read the Douglass paper, and 2) read the comments, and 3) make up your own minds whether the models actually are representing the situation correctly.

w.

What’s wrong with the realClimate coverage? Is it that they link to other scientific [url=http://ams.allenpress.com/perlserv/?request=get-abstract&doi=10.1175%2F2008JCLI1929.1]papers(well, abstracts, so far) rather than just blog discussions?

well, I buggered that up, didn’t I. That should read scientific papers

Thank you. Apparently, to you, “documented” means that it’s possible to set up a computer simulation consistent with the claimed phenomena.

To me, “documented” means (in this context) verified by actual observations and measurements of the real world.

Here’s what your link says about that:


No I’m not. I mentioned “CAGW” as a possible interpretation of what you meant by “AGW” in your question to me. If my interpretation is incorrect, please feel free to supply your own explanation.

Actually I typed too quickly. By “says you,” I was trying to say that I don’t accept your claim as some sort of established scientific fact.

Sure. In the absence of a lot of positive feedback, it’s pretty clear that an increase of CO2 to 450 ppm will not warm the earth enough to have significant negative effects. Instead, it would be a warming more on the order of what was experienced between 1850 and 2000.

So the critical question is whether or not there is a lot of positive feedback. There’s no real evidence to suggest that there is a lot of positive feedback, and the natural assumption, in a system that has been somewhat stable for a looooooong time, is that negative feedbacks predominate.

Indeed, if there were a lot of positive feedback at work, one would expect much more warming to date than has occurred. Alarmists get around this problem by postulating that the extra heat is hidden in the oceans, like some monster under the bed waiting to jump out and attack us. Or by hypothesizing that other human emissions are masking to extra warming. Both of these hypotheses have the feel of epicycles to them. And in any event, there is little or no evidence to support these hypotheses either.

dual post, removed …

If you don’t know the answer to this question at this point in the dialogue, MrDibble, you may be beyond assistance. However, for others who might have just joined the discussion, the problem with RC is that they heavily censor any opposing scientific questions and comments, while pretending that they are a scientific site. This allows them to further the illusion of AGW consensus, and to pretend that they are being responsive to comments, while carefully avoiding any difficult (or interesting) scientific questions. It is cowardly and underhanded.

But you knew that …

w.

Up to this point, you haven’t said anything attrociously wrong although the statement, “A number of AGW supporters use this to argue that the data must be wrong” is a poor summary of the reasons why the AGW supporters suspect this (see below for an explanation of why they do) and the claim that “the estimates of all of the climate model were much, much higher than those of the balloons and the satellites” is incorrect, as it has not yet been established that the difference is statistically-significant and there are also issues of which data sets are used (e.g., for the RAOBCORE re-analysis of the balloon data, do you use the latest version that shows much more warming in the tropical atmopshere or the earlier version that Douglass et al. used).

This statement is untrue or, at best, incredibly deceptive, as you well know since you have been told it a number of times. The prediction that the warming will be magnified in the tropical atmosphere relative to the surface is not a specific consequence of the warming being due to the mechanism of greenhouse gases. It would be true if the warming were due to solar or presumably any other mechanism. And, it is not only true of the slow warming that has been occurring over the last few decades (which most scientists attribute primarily to greenhouse gases) but also to temperature fluctuations that are seen in the tropical atmosphere on the timescales of months to a few years years. It is a general consequence of what is called “moist adiabatic lapse rate theory”.

Hmmm…So, you no longer believe it necessary to determine if a discrepancy is statistically-significant? It seems strange how much statistical significance matters to you in some circumstances but not in others. Until the analysis of statistical significance is done correctly, it cannot be determined whether an important difference remains. In particular, the Douglass paper had three main faults that invalidate its conclusions on this:

(1) It used the models’ standard error instead of standard deviation to look at the issue of statistical significance.

(2) Even before computing the standard error, they averaged over different runs of the same model, which is an incorrect procedure because it tends to average out the unforced variability, which will not be averaged out in the real world.

(3) They did not use the latest version from the RAOBCORE re-analysis project for the balloon data. They make some justifications for this, but given that the more current versions show substantially more warming in the tropical atmosphere, it seems worrisome that their result relies quite critically on choosing an older version of a data set where the source of the data set now has a more recent version that those producing this data set believe to be more correct.

This does not mean Douglass et al.'s conclusions are necessarily wrong but it does mean that we don’t have any particular reason to believe them to be correct given the very serious errors in their analysis. It may be that a correct analysis done in a smarter way than Douglass did (e.g., by normalizing everything by the temperature trend at the surface, as you yourself did in one plot over at ClimateAudit) confirms that there is a statistically-significant difference between data and models, at least for some of the data sets. However, this still needs to be demonstrated.

And, again, the prediction that tropical atmosphere heats up faster than the surface is not specific to the greenhouse gas mechanism of the AGW hypothesis, despite how many times you try to incorrectly imply that this is the case.

First of all, it is not clear that all of these data sets have statistically-significant discrepancy with the models, given the errors in Douglass et al.'s analysis (although some of them probably do). Second of all, it is not all that improbable that there are errors that go mainly in the same direction. For the balloon data, there is a known reason why the data would tend to be biased toward cooling…namely that there have been significant changes to instrumentation that have tended to result in more shielding of the temperature sensor from the sun over time for the daytime observations.

Third of all, there is a very good reason to believe that the data may be wrong while the models are right that has not been discussed here at all. Namely, it is because the data and the models agree very well in showing this magnification as you go up in the tropical atmosphere when one looks at temperature fluctuations that occur on timescales of months to a few years. It is only when one looks at the “fluctuations” on timescales of a decade or more that there is any discrepancy and this discrepancy is caused by the fact that the models continue to behave in the same way as they did for the shorter timescales whereas the data deviates from this. [I put “fluctuations” in quotes because over these decadal timescales, what we are looking at are not really fluctuations up and down anymore but simply an upward trend.] So, there are two options:

(1) The very basic physics involving moist adiabatic lapse rates that seems to dominate the observed structure of the fluctuations on timescales of months to years breaks down due to some unknown mechanism when one goes to the longest timescales of decades. I.e., the data is right and the models fail to capture some new effect that only comes in at very long timescales compared to most relevant atmospheric processes that seem to control the temperature structure in the tropical atmosphere, which seems sort of bizarre…particularly without even any hypothesis that I know of to explain how this could be the case.

OR

(2) The basic physics continues to hold and, to the extent that the data disagrees with this, this is due to errors in (at least some of) the data sets. And, in fact, there are good reasons to believe that, while the balloon and satellite data are accurate for looking at temperature fluctuations over a timescale of months to a few years, they are not so accurate in detecting slow underlying trends over multidecadal timescales. This has to do with the fact that neither the satellite nor the balloon data was designed for use as a longterm climate monitoring system. And, as I noted above, problems have been identified with the balloon data that would tend to produce a gradual cooling artifact over time. For the satellite data, there are also various issues having to due with drifts of individual sensors, decay of satellite orbits and shifts of time-of-day of measurements, and replacement of older satellites with newer ones and the need to splice the data from the different satellites together.

I make no specific claims as to whether the models and data are in statistically-significant disagreement because the Douglass et al. paper is too flawed to decide that one way or the other. (One might be able to make some conclusions from the U.S. Climate Change Science Program report [which I assume is what you meant to say instead of NAS report], but that report was issued before the latest versions of the RAOBCORE reanalysis of the balloon data and before [I believe] a correction to one of the satellite data analyses.)

Since RealClimate is the only place where the science on this seems to be being discussed close to correctly, it is rather important to read what they have to say…and it is available here and here. Note particularly the figure in that link that clearly shows how magnification of the surface temperature trend higher in the troposphere is a prediction of the climate models whether the cause of the warming is greenhouse gases or in an increase in solar forcing.

My best to you, as always.

jshore, thanks for your reply.

You are correct that the warming trend in the troposphere is a feature of the models, both with and without GHGs being considered. However, the warming is much greater (about double) with the GHGs being considered. Thus, I’m not sure why you say:

What I actually said was:

Please don’t read into that more than I said. There are, in fact, other hypotheses which do not require the increased warming, as you know very well since we have discussed them before. Pretending that we have not discussed them does not become you.

Regarding statistical significance, since it was (and remains) so contentious, I discussed the difference in direction (along with a look at the data and the models). When all of the models trend up and the data trends down, the exact statistics are of less significance.

The claim that RealClimate is the “only place where the science on this seems to be being discussed close to correctly” is, unfortunately, far from true. There have been a number of statisticians who have disagreed with RealClimate. Some of them have even tried to discuss it on RealClimate … but of course, some of them have been censored. Is it your contention that science being “discussed close to correctly” includes censorship, or is that just a side benefit? Many scientists who disagree with Gavin and his minions don’t even bother trying to post there, Gavin has covertly thrown too many of their contributions in the trash can. I sincerely hope you don’t think that such scientific censorship is acceptable.

What you really are saying is that RC is the only site that agrees with jshore … and I would admit that you 100% are right about that, it stands pretty much alone in that regard, lots of other folks disagree. As someone who generally worships at the altar of “consensus”, why are you so opposed to the consensus all of a sudden?

w.

(PS - as regular readers know, I hold that consensus is meaningless in science. To paraphrase Michael Crichton, when you talk of consensus, you are not talking about science. I’m just twitting jshore a bit, because this time he’s complaining that the consensus is against him and the only site where science is discussed “close to correctly” is the home of rampant censorship, the one and only [Un]RealClimate.)

Just noticed that both those links are the same. I meant one of them to be to here and this is the link with the figure that I was talking about.

I lost you here. What do you mean it is double? Do you mean the amplification factor in the tropical atmosphere between the surface and up at, say, the 300mB level is double what it is without GHGs or what? And where do you get this number from?

There are other hypotheses that do not require the warming to increase as you go up in the tropical atmosphere? Can you expand on what these are again as I seem to have missed them?

Your claims here are so vaguely and unclearly stated I can’t even figure out what it is that you are saying anymore.

Whether the trend is up or down is sensitive to these data-quality issues…as the various versions of the RAOBCORE re-analysis show. I don’t think the direction of the trend gets you out of having to show statistical significance. (Although, as I noted, I think that the right way to get the tightest constraint is the method that you had of normalizing by the surface trend. But, you still have to then do the statistics correctly.)

Who are these statisticians exactly? You claimed that Briggs disagreed but linked to a thread where Briggs did not really address the statistical issue of whether Douglass et al. were right…at least that I could see.

Well, in this case, I think RC agreeing with me is good because I’m right. I have given you quite a bit of evidence and argument for this…and you are the one here who is not answering these arguments but instead appealing to authorities, although they are rather dubious and ephemeral authorities from what I can tell.

And, RC only stands alone in this regard when you lump it into the other company that you lump it into. Other sites, like Tamino’s blog, the Deltoid blog, William Connelly’s blog, etc. do (I am pretty sure) agree with RC on this point. But, at any rate, I don’t think a consensus of the blogging community carries as much weight as a consensus of the scientific community.

By the way, in other climate news, the U.S. National Academy of Sciences and the analogous bodies in Brazil, Canada, China, France, Germany, India, Italy, Japan, Mexico, Russia, South Africa, and the U.K. issued a new stronger and more specific statement [PDF file] on climate change today, that says in part:

Quick answer, out of time …

AR4, Chapter 4, Figure 9.1. Typical numbers, 300 mb, solar only, ~ 0.4°C/century, with GHGs, ~0.8. These numbers are typical for the models.

Gotta run, past midnite, traveling for about ten days.

Best to all.

w.

Color me confused. This figure shows the hindcast values from climate model simulations for the last century showing that the IPCC estimates GHGs to have had a dominant warming factor and solar to have shown less warming over that time. This is true both at the surface and at 300mb.

This has nothing to do with the structure of the warming in the tropics (i.e., the prediction that it increases as you go up in the troposphere), a prediction which is what you originally wrongly claimed was a unique signal of warming due to AGW when in fact it is the structure expected regardless of the warming mechanism.

The best to you on your travels!

In 1991 Mount Pinatubo in the Philippines erupted, releasing large amounts of sulfur dioxide which combined with water in the atmosphere to drop the global average temperature. This created a natural experiment (a natural experiment is a condition where some variable is changed independent of human influence and scientists can then study it - it’s especially valuable in a field such as climatology where large-scale designed experiments are impossible) where you can compare the conditions in the atmosphere before the eruption (warmer) with the conditions 2-3 years post-eruption (cooler) and again with >3 years post-eruption (warmer.) One of the handy pieces of information that came out of the Pinatubo studies is the change in water vapor in the atmosphere, measured by NASA’s Water Vapor Project and the Television Infrared Observation Satellite. As global temperatures cooled, the amount of water vapor fell, and as global temperatures rose again, the amount of water vapor climbed. Even better, when compared to computer models (cue the creepy music for folks who believe computer models are as useful as a Pop-Tart abacus) the rise and fall tracked that predicted by models.

From: Soden et al. 2002. Global cooling after the eruption of Mount Pinatubo: a test of climate feedback by water vapor. Science v. 296(5568) p. 727.
In addition, it has also been more recently shown that you can identify the signature of current atmospheric water vapor to global warming, most likely from increased greenhouse gases (GHGs):

From: Santer et al. 2007. Identification of human-induced changes in atmospheric moisture content. Proceedings of the National Academy of Sciences. v. 104 n.39: 15248-15253.
It’s not just in the last year, though, here’s a work from 3 years ago that discussed the detected trends in atmospheric water vapor:

From: Soden et al. 2005. The radiative signature of upper tropospheric moistening. Science, v. 310 (5749) p. 841-844.
So there is evidence to suggest that there is a water vapor feedback. Satellite measurements, in particular, document this. The part I referred to as well-documented is the greenhouse effect of water vapor, which has been known since the experiments of John Tyndall in the 1860s. It was not clear from your post which part of the science you disagreed with.
You also quote the IPCC with reference to water vapor, but I think you misunderstand what you have quoted.

(Bolding mine)

I’ve highlighted where I think you have misunderstood: seasonal evidence is indeed insufficient to demonstrate long-term water vapor feedback, but the three papers I quoted above do not depend on seasonal changes or cycles.

How much warming do you expect if positive feedback is occurring? I don’t see a cite here.

“Fact” is an odd word to choose here - obviously there are no “facts” in the future - science is not the process of using a crystal ball or exhibiting Revealed Truth[sup]TM[/sup]. If I parse your meaning correctly, you are saying that although I have presented evidence to support my position on likely temperature increases and I have seen none of yours, your position is unchanged. Would you like to present any evidence? It could make this a more interesting debate.
Note to intention: enjoy your trip!

wevets, it’s unfortunate that you have decided not to define “AGW.” I believe that a lot of confusion results from the failure to clearly define this sort of term. In any event, if you won’t explain what you mean by “AGW,” then I cannot really answer your earlier question.

Laugh as much as you like, but it’s a real problem. Essentially, the paper you cite shows that you can take a simulation which assumes amplification and fit it to a 4 year temperature record. Then, if you take the assumption out and change nothing else, the simulation no longer fits. To me, this doesn’t show much at all.

Let me ask you this: Do you believe that climate simulations are advanced enough to predict temperatures over a 5-year time period reasonably accurately?

I will try to come up with a number and a cite for you.

That’s incorrect. For example, I can predict with a good degree of confidence that if you heat gold up to 1064 degrees celcius, it will melt. i.e. it’s reasonable to say “The melting point of gold is 1064 degrees celsius.”

My objection was simply that you were presenting your view as if you were discussing something settled like the melting point of gold.

I’ve already explained why I think the sort of amplified warming necessary to produce signifiant negative effects is unlikely. In any event, it’s worth emphasizing that you have the burden of proof. (If you subscribe to the hypothesis I think you do.)

The only evidence I have seen for such a hypothesis is that it’s possible to construct a simulation consistent with the hypothesis and consistent with past temperatures. Sorry, but that’s not very persuasive.

Let me repeat my earlier question: Do you believe that climate simulations are advanced enough to predict temperatures over a 5-year time period reasonably accurately?