2011-12-15 15:08:49Foster and Rahmstorf Measure the Global Warming Signal
dana1981
Dana Nuccitelli
dana1981@yahoo...
71.137.110.252

Post on tamino's paper:

Foster and Rahmstorf Measure the Global Warming Signal

2011-12-15 15:17:25
KR

k-ryan@comcast...
69.138.165.234

I think I would include the graphs of volcanic, solar, and El Nino activity, as contributors/components of the global temperature record. This doesn't feel supported (having read the initial and final Tamino posts on the subject) without the various components shown, while _with them_ it is very clear.

Otherwise I forsee objections like "Duh, how did you fake that? I don't see that trend in the data!!!"

2011-12-15 15:19:04
KR

k-ryan@comcast...
69.138.165.234

I hate to say it ('cause I know how much work it would be), but another animation might be worthwhile.

Show the temperature record, then successively remove the various components, leaving the linear trend increase. Might be very effective...

2011-12-15 15:35:20
Sphaerica

Bob@Lacatena...
76.28.5.93

Looks good to me.  You might add Figure 7 from the paper.

Also, the scaling on the copied table-images makes the fonts so huge that it's a bit unsettling.  You might try setting the width smaller (say 200, but you have to play with it.) and using align=center to center the smaller (but still legible) image.

2011-12-15 16:06:35
dana1981
Dana Nuccitelli
dana1981@yahoo...
71.137.110.252

KR, you just can't get enough of those animated GIFs, can you? :-)

Actually I like the idea.  I've sent Figure 7 off to Riccardo to see if he can digitize it for me.  If so, I'll try to make the suggested animation.  I'll also add Figure 7 to the post.

I'll also need to update a bunch of rebuttals - it's the sun, it's volcanoes, it's ENSO (I think we have one for ENSO...), global warming stopped...any others?

2011-12-16 01:38:20
Sphaerica

Bob@Lacatena...
76.28.5.93

Dana,

"It's internal variability" probably as well, although maybe for internal variability deniers mean "any mystical, magical force that can't be isolated and measured": http://www.skepticalscience.com/internal-variability.htm

Also "Warming stopped in 1998 (2002, whatever"):

http://www.skepticalscience.com/global-cooling-january-2007-to-january-2008.htm

http://www.skepticalscience.com/global-warming-stopped-in-1998.htm

http://www.skepticalscience.com/global-cooling.htm

2011-12-16 01:43:24
Sphaerica

Bob@Lacatena...
76.28.5.93

Oh, and I like the animated chart idea as well, but when Tamino first did this on his blog I asked him to do just that (not animated, but to remove each component one by one for the cumulative effect).  His response was:

Response: I agree that removing factors one at a time makes a more compelling narrative. But it's a statistically inferior approach, because different factors can correlate (or anticorrelate) with each other. In such a case, the best approach (statistically) is to remove them all simultaneously by multiple regression.

But you could at a minimum transition from a temp graph to the adjusted graph.

To do it one at a time, though, you'd need to get him to cooperate and provide you with the intermediate graphs.

2011-12-16 03:10:17
dana1981
Dana Nuccitelli
dana1981@yahoo...
64.129.227.4

Hmm internal variability is a little borderline, because they didn't investigate every single source of internal variability.  In fact ENSO is really the only one they filtered out (volcanic and solar are both external forcings).  The cooling myth would be applicable though.

Even if it's "statistically inferior", it doesn't hurt to show an animation with each effect removed one at a time.  I agree if you're trying to do a multiple regression, it makes sense to do them all at once.  But if you're just showing a graphic of what it would look like to remove each (after the regression is already done), then there's nothing wrong with removing each one at a time.

I've sent their Figure 7 to Riccardo to see if he can digitize it.  If so, then I can use it to remove each effect from the temperature data one at a time, and create a GIF out of it.

2011-12-16 05:29:59
Sphaerica

Bob@Lacatena...
76.28.5.93

It might be easier/quicker and more accurate to just ask Tamino for the raw data (and the resulting numbers used to generate his graph).

2011-12-16 06:29:16
Alex C

coultera@umich...
67.194.38.116

>>>then there's nothing wrong with removing each one at a time.

I wonder if Tamino meant that the animation wouldn't show a clean-cut transition.  After removing so-and-so effect for instance, you might end up exasperating the "flat" trend.  There are other ways that might make it less visually appealing, as contrasted to an all-in-one approach.  I like the concept though.

I agree too that getting the data would probably be quicker.  The graphics should be a piece of cake to compile with the component data applied to the records.

2011-12-16 07:21:27
dana1981
Dana Nuccitelli
dana1981@yahoo...
64.129.227.4

We'll see what it looks like.  I'll email tamino and see if he can give me the necessary data.

Actually, I don't have his email address (anyone got it?).  I'll leave a comment on the Open Mind post to see if he can send it.

2011-12-16 08:55:16
Sphaerica

Bob@Lacatena...
76.28.5.93

Alex C,

At the time, I wasn't suggesting Tamino make an animated graph, just a stepwise progression in a series of separate graphs.

On the other hand, that request did suggest a change to the methodology (which is what he objected to), when really I was only talking about presentation.  In theory, he should be able to use his original method to get his regression coefficients, but then to generate step-by-step changing graphs along the way.

I've never done or thoroughly looked into multiple regression, though, so I can't be sure it really pans out this way.

2011-12-16 10:16:18
dana1981
Dana Nuccitelli
dana1981@yahoo...
64.129.227.4

tamino has made the data available on his website.  I'll see if I can make an animation tonight.

2011-12-16 14:10:11
dana1981
Dana Nuccitelli
dana1981@yahoo...
71.137.110.252

Okay, I just removed all three factors in one step, because tamino didn't break out the three factors in his data, and it's probably better to keep it simple anyway (just a before and after shot).  See Figure 3 in the post.  Let me know what you think.

2011-12-16 14:27:16Comment
Robert Way

robert_way19@hotmail...
142.163.168.141

Didn't Kevin C comment that Tamino made a mistake?

2011-12-16 14:29:19
Sphaerica

Bob@Lacatena...
76.28.5.93

Awesome.  Good to go.

2011-12-16 16:17:30
dana1981
Dana Nuccitelli
dana1981@yahoo...
71.137.110.252

Rob - Kevin had some criticism about the paper.  Can't recall exactly what, or find the comment at the moment.

2011-12-16 22:57:32
Kevin C

cowtan@ysbl.york.ac...
144.32.72.165

You should go with the article, because it is an accurate reflection of the peer-reviewed literature. Even if a) I am right about the problem and b) the effect is big enough to notice, then Tamino's analysis is a useful stepping stone.

The only thing I would avoid doing is making the case in a way which contradicts other peer-reviewed research, e.g. Kaufmann 2011. Either you can explain the recent trends from solar+volcanoes+enso alone, or you need aerosols too. Both can't be right. At the moment we don't know which is true. Figuring in CERES and OHC and Santer is tricker still - I don't know enough to even attempt this. The big danger is that we start behaving like the deniers and producing Gish-gallops of contradictorary excuses for the lack of recent warming, which just looks desparate.

The article as it stands avoids all of these pitfalls, so it gets my go ahead.

[FWIW Tamino has now posted data and code, so you could do the progressive contributions. As I read Tamino's response to Sphaerica, his objection is to the statistical validity of a regression using only some of the terms, not to the graphical presentation of different combinations of terms from the complete regression.]

2011-12-17 01:50:01
Sphaerica

Bob@Lacatena...
76.28.5.93

Kevin,

Is there a pointer to/explaination of yur questions/criticisms?  I'd like to read it.

Also... have you suggested to Tamino that he figure out a way to add aerosols as a fourth factor (or perhaps, instead and more generally, change in albedo)?  I think that would be hugely useful, although maybe difficult.

[Update -- I just posted that request/suggestion to Tamino on his latest code-release blog post.]

2011-12-17 01:52:41
KR

k-ryan@comcast...
216.185.0.2

Excellent.

Kevin C - You may well be right about the aerosols, and I would love to see this work extended with an aerosol proxy term as well. But given a limited number of components for attribution, all of the data will get attributed to those factors used (hence Tamino's objections to regressions using fewer terms). Adding an aerosol proxy would result in a different factorization - the question would be whether the statistics of that factorization would be an improvement on the statistics shown by F&R. Alternatively, it may be that the Asian aerosol contribution is actually smaller than we might think; there are uncertainties there too, after all.

I think this is a very interesting bit of current science, and definitely worth talking about - regardless of where future developments from this work go in detail.

2011-12-17 02:26:46
Kevin C

cowtan@ysbl.york.ac...
144.32.72.165

Sorry, I'm afraid I really don't have anything more to say until I have some real results to show. I only commented because Robert brought my name up. Work and family pressures are preventing me from working on this at the moment. Hopefully next term I can do a bit more.

2011-12-17 02:51:13
dana1981
Dana Nuccitelli
dana1981@yahoo...
64.129.227.4

Bear in mind that tamino's trend is 0.17°C/decade, whereas the model-expected trend is 0.2°C/decade.  So it's possible that in addition to the exogeneous effects considered by tamino, aerosols are still causing additional cooling.  In other words, this may just be a partial answer, with Kauffman being another part.

2011-12-17 03:22:57
Sphaerica

Bob@Lacatena...
76.28.5.93

Dana,

It would also be interesting to see the study done again not on the individual temperature series, but rather each series plus ocean heat content (i.e. a measure of the total accumulated energy of the system, rather than just the atmospheric reflection of it).

But that's neither here nor there in respect to your post.  I'm eager to see it go up, and to see what comments you get.  I already had to bite my tongue when SirNubwub asked a question to which this was a good answer (scaddenp cheated anyway and referenced the paper directly) to his question about a pause in global warming.

2011-12-17 04:28:03
dana1981
Dana Nuccitelli
dana1981@yahoo...
64.129.227.4

I'll probably publish this one Monday morning.

2011-12-17 05:56:13Tilt!
John Hartz
John Hartz
john.hartz@hotmail...
98.122.98.161

Dana,

The information contained in the pop-up boxes of most of the graphics do not square with the written description of the graphic. I suspect this is because the graphic was cut & paste and retains the pop-up message from the original source.

For eaxmple, if you run the cursor over Figure 1, "Fig. 5" shows up in the pop-up box. 

PS-- I've noticed this to the case on prior articles that you have published.

2011-12-17 06:05:20Too much "inside baseball"
John Hartz
John Hartz
john.hartz@hotmail...
98.122.98.161

Dana,

You have not spelled out what the acronymns used in figures and tables stand for.

You have also not told the reader what a tmeperature anomaly is and what "baseline" was used to compute the ones shown in your figures and tables.

2011-12-17 06:07:13Final recommendations
John Hartz
John Hartz
john.hartz@hotmail...
98.122.98.161

The article needs both a good introductory paragraph and a good summary paragraph a la Huber and Knutti Quantify Man-Made Global Warming.

.

2011-12-17 08:02:11
John Hartz
John Hartz
john.hartz@hotmail...
98.122.98.161

How do the findings of Foster & Rahmstorf compare to the findings of Huber & Knutti?

Should there be a crosswalk between this article and  Huber and Knutti Quantify Man-Made Global Warming?

 

2011-12-17 12:49:16
dana1981
Dana Nuccitelli
dana1981@yahoo...
71.137.110.252
It's got an intro, John. And the pop-up labels are the figure/table numbers in the paper itself. The acronyms are just the temperature series that everyone reading the post will be familiar with. I'll see if I can make a connection to Huber and Knutti.
2011-12-17 18:24:59
Riccardo

riccardoreitano@tiscali...
2.33.129.55

I think that these kind of short and simple reporting of the science has the strongest impact on the average reader. Good job.

Typo: first paragraph after the table "0.141 and 0.149 0.157 °C"

2011-12-18 02:32:14
dana1981
Dana Nuccitelli
dana1981@yahoo...
71.137.110.252
Oh yeah whoops, I looked at the wrong column there. Thanks Riccardo.