The AGW climate feedback discussion

[Xnn]

Have found some comment on Lindzen and Choi's paper.

Dr Roy Spencer:

It is not clear to me just what the Lindzen and Choi results mean in the context of long-term feedbacks (and thus climate sensitivity). I’ve been sitting on the above analysis for weeks since (1) I am not completely comfortable with their averaging of the satellite data, (2) I get such different results for feedback parameters than they got; and (3) it is not clear whether their analysis of AMIP model output really does relate to feedbacks in those models, especially since my analysis (as yet unpublished) of the more realistic CMIP models gives very different results.

Chris Colose comments that Lindzen tends to use outdated ERBE data.

So, maybe Lindzen has not found evidence of a negative feedback after all.
It is a very strange paper, especially where he states that the models ought to have higher sensitivities.

Andre;

In the Northern Hemisphere, seasonal snow can cover over half of the land. So, while the Southern Hemisphere has more permanent ice, the albedo feedback is stronger in the Northern Hemisphere. The hemispheres are not symmetrical with respect to snow/ice albedo.

 

[Andre]

http://chriscolose.wordpress.com/2009/03/31/lindzen-on-climate-feedback/#more-429" that Lindzen tends to use outdated ERBE data.

Now Chris states (see link in that post):

This has been documented in Wong et al 2006, Journal of Climate, a paper not even mentioned by Lindzen.

Maybe I am permitted to demonstrate the problems of using non peer reviewed blogs, since http://www.drroyspencer.com/Lindzen-and-Choi-GRL-2009.pdf state:

The observed data used in this study are the 16-year (1985–1999) monthly record of the sea surface temperatures (SSTs) from the National Centers for Environmental Prediction, and the Earth radiation budget from the Earth Radiation Budget Experiment (ERBE) [Barkstrom, 1984] nonscanner edition 3 dataset. Note that this data were recently altitude corrected and are acknowledged to be stable long-term climate dataset based on broadband flux measurements [Wong et al., 2006]....

references:
Wong, T., et al. (2006), Reexamination of the observed decadal variability of the Earth radiation budget using altitude-corrected ERBE/ERBS nonscanner WFOV data, J. Clim., 19, 4028– 4040, doi:10.1175/JCLI3838.1.

 

[Andre]

In the Northern Hemisphere, seasonal snow can cover over half of the land. So, while the Southern Hemisphere has more permanent ice, the albedo feedback is stronger in the Northern Hemisphere. The hemispheres are not symmetrical with respect to snow/ice albedo.

Even if the snow can cover half of the land, -if I interpret that as down to 45 degrees lattitude (halfway the equator) then there is still 82% more area below that lattitude.

Other factors considering snow-ice feedback, of which I don't know, if the models did account for it:

1: Snow usually coincides with winter - low solar angles and Arctic winter, no sun. So there is not a lot of insolation to be albedo-reflected in the first place.

2: Temperature of the surface is also dependent on temperatures subsurface, although the heat tranfer rate due to conduction is low, one can assume that the temperature of the layers of the first few feet of rock/soil, do play a role in the annual cycle. Snow is an excellent isolator (think of iglos) and soil covered with snow is practiclally not losing much more heat, whereas barren soil continues to radiate out and continues to cool. So maybe that the lower atmospheric temperatures above a snow cover due to more reflection is offset partly/substantially by the retention of heat underneath the snow. Note also that this can be considerable comparing the mass of the soil with the mass of the atmosphere. Therefore in the subsequent warming after the winter, a warmer insolated soil may offset the positive feedback to some extend.

 

[Trakar]

Now Chris states (see link in that post):



Maybe I am permitted to demonstrate the problems of using non peer reviewed blogs, since http://www.drroyspencer.com/Lindzen-and-Choi-GRL-2009.pdf state:

Indeed, it would seem, that there are a lot of problems regardless of venue. You refer to the Lindzen Choi paper published in July 2009, Chris's comments were made in March of 2009, and if I'm not mistaken, from the actual wording, appear to be in reference to comments Lindzen made on Anthony Watt's non-peer reviewed blog (http://wattsupwiththat.com/2009/03/30/lindzen-on-negative-climate-feedback/).

So far, I've just been skimming through PF's Climate change discussions, trying to get a feel for where understandings exist and the types of arguments that are prevelent here before deciding whether or not to enjoin such discussions, hopefully, this type of mistake/misrepresentation is not overly rampant!?

 

[Andre]

Note that Xnn stated:

Have found some comment on Lindzen and Choi's paper.

http://www.drroyspencer.com/2009/11/some-comments-on-the-lindzen-and-choi-2009-feedback-study/" :

It is not clear to me just what the Lindzen and Choi results mean in the context of long-term feedbacks (and thus climate sensitivity). I’ve been sitting on the above analysis for weeks since (1) I am not completely comfortable with their averaging of the satellite data, (2) I get such different results for feedback parameters than they got; and (3) it is not clear whether their analysis of AMIP model output really does relate to feedbacks in those models, especially since my analysis (as yet unpublished) of the more realistic CMIP models gives very different results.

http://chriscolose.wordpress.com/2009/03/31/lindzen-on-climate-feedback/#more-429" that Lindzen tends to use outdated ERBE data.

So, maybe Lindzen has not found evidence of a negative feedback after all.
It is a very strange paper, especially where he states that the models ought to have higher sensitivities.

Implying that all comments were on Lindzen and Choi 2009. However we see that this paper does indeed implement the corrections of Wong et al 2006, but moreover the other source which may not be linked to, has implemented these corrections.

Finally note that Spencer in his blog, to which one may not link, may agree with Lindzen and Choi that the climate is much less sensitive than the climate models suggest.

 

[mheslep]

Have found some comment on Lindzen and Choi's paper.

Dr Roy Spencer:

The comment is from Spencer's blog. Even though he's apparently a fine scientist, you might have labeled that comment as such.

 

[sylas]

Implying that all comments were on Lindzen and Choi 2009. However we see that this paper does indeed implement the corrections of Wong et al 2006, but moreover the other source which may not be linked to, has implemented these corrections.

Finally note that Spencer in his blog, to which one may not link, may agree with Lindzen and Choi that the climate is much less sensitive than the climate models suggest.

Quite right. Chris' comments to which Xnn has referred were with respect to an earlier version of Lindzen's work, before it was published. I think you should presume an honest mistake.

Chris' comments were correct at the time, and by the time Lindzen published those particular problems were fixed. So in fact, the problem was not with being a blog per se; it was rather a case of keeping track of what was being referred to.

Dr Spencer, like Lindzen, is one of a very few working climate scientists who argues for a very low climate sensitivity, which makes his criticisms of Lindzen's paper that much more credible. He is evidently basing his criticisms on the real merits of the method itself, without being led into spurious agreement merely because he like the conclusion.

Like Spencer, I want to be completely sure I am being fair to the paper on its own merits, and not merely disagreeing because I don't like the conclusion. I want to understand the paper, and see how it obtains the result and just the method on its own merits.

A comparison paper would be:

• Gregory, J.M. & Forster P.M. de F (2006) The Climate Sensitivity and Its Components Diagnosed from Earth Radiation Budget Data, in J. of Climate Vol 19, Iss 1, pp 39-52

This is a widely cited and much more detailed paper which ALSO uses the ERBE data, and infers a net positive feedback. There are some interesting similarities and differences with the two papers. That Lindzen does not cite this paper is strange; it is very influential and widely cited prior work doing precisely what he has attempted; constrain sensitivity using ERBE data.

I will be posting more on this, but I really want to be sure I understand both papers before I comment further. I want to uphold the high standard shown by Roy Spencer, in sticking to arguments on their own merits and not just because of preconceptions about the right answer.

By the way, the ice feedback matter is not all that fundamental. The expectation of models is of positive feedback in the tropics, so just looking at the tropics is fine. If ERBE really does show a negative feedback in the tropics as Lindzen and Choi suggest, this is revolutionary. Forster has some challenges for the models as well, but not so sweeping; and obtains a positive feedback effect from the same data. I want to understand why the two papers are so different, despite using the same data.

When I post, I will be sticking simply to peer-reviewed sources.

Cheers -- sylas

 

[Xnn]

Finally note that Spencer in his blog, to which one may not link, may agree with Lindzen and Choi that the climate is much less sensitive than the climate models suggest.

Andre;

Thanks for pointing out that Lindzen used the current data in his 2009 paper. At least we can rely on the data. However, I've received another comment regarding the paper at Real Climate.

There is no good reason to expect that tropical oceans are a good proxy for assessing global climate sensitivity. There is already a large amount of water vapor in the tropical atmosphere and no sea ice or seasonal snow. Rising water vapor, melting sea ice and seasonal snow are the primary postive feedbacks of climate models that he compares against.

Next, look at the data charts in the Lindzen paper. SST shows no signficant trend over the period (global warming is most pronounced in the Arctic) So, he picks El Nino/La Nina events as "data points". However, wait a minute, we know that these events are not due to CO2 fluctuations. So, why should we expect climate models to model these as minature global warming/cooling events?

What stands out most prominently in the ERBE data is the Mt Pinatubo erruption of 1992/93. A large fluctuation in SW with a smaller response in the LW. 2 of the models (GDFL CM2.1 and GISS-ER) appear to model that fairly well, but the rest miss it entirely. So, I'm left wondering if the other climate models had the aerosal data for Mt Pinatubo.

Finally, there is a gap in the ERBE SW data in 1993 and a shift thereafter. I'm not sure what that is all about. If it is valid, then there appears to be a trend in the data. If not, then it's just poor data. Confirmation is needed to ensure we understand what is happening since it may be significant.


Conclusion (based on Lindzens data):

None of the models model El Nino/La Nina events very well.
2 models (GDFL CM2.1 and GISS-ER) do a fair job of modeling Mt Pinatubo.
The remaining models either had no Pinatubo data or did a lousy job.
There maybe a trend in SW flux over the tropics.

 

[sylas]

... The expectation of models is of positive feedback in the tropics, so just looking at the tropics is fine. If ERBE really does show a negative feedback in the tropics as Lindzen and Choi suggest, this is revolutionary. Forster has some challenges for the models as well, but not so sweeping; and obtains a positive feedback effect from the same data. I want to understand why the two papers are so different, despite using the same data.

When I post, I will be sticking simply to peer-reviewed sources.

I have been working on this matter, but events have overtaken me. Which is just as well, given as this whole topic is about to be shut down and I am not yet ready to give a proper response of my own.

There has just come out a formal published response to Lindzen and Choi 2009. It is:

• Trenberth, K. E., J. T. Fasullo, Chris O’Dell, and T. Wong, (2010): Relationships between tropical sea surface temperature and top-of-atmosphere radiation, in Geophys. Res. Lett., 37, doi:10.1029/2009GL042314, in press. (Or try this link, which fails now but might work better in future: http://dx.doi.org/10.1029/2009GL042314, I am not sure.)

It will require a subscription to get the paper. The conclusion is concisely summarized:

As shown here, the approach taken by LC09 is flawed, and its results are seriously in error. The LC09 choice of dates has distorted their results and underscores the defective nature of their analysis. Incidentally, LC09 incorrectly computed the climate sensitivity by not allowing for the Planck function in their feedback parameter. For their slope of -4.5 W m^-2 K^-1 and using the correct equations (Section 1), LC09 should obtain a feedback parameter and climate sensitivity of -0.125 and 0.82 K, respectively, rather than their values of -1.1 and 0.5 K. In contrast, the case 4 (Table 1) results yield a positive feedback parameter of 0.6 and a climate sensitivity of 2.3 K. Moreover LC09 failed to account for the forcings in estimating sensitivity.[/color]​

It is worth noting that having errors or defective analysis in a published paper is not particularly unusual. The original paper does have errors, and Lindzen has already acknowledged some of them. He is likely to write a revised paper taking criticism into account, and we shall have to wait and see what he provides... just as we had to wait for a few months to see a formal published paper describing the errors in the original work. This is a normal part of the work of science, and a reminder that one should be very cautious about jumping on a new paper that has just been published as definitive. The discussion will continue.

For all that the paper was well outside the conclusions of the vast majority of work on this subject, and for all that the numbers as in the original 2009 paper are certainly wrong as given, the work itself is not immediately trivially incorrect. It was in some respects rather subtle, and it continues to be useful to have working scientists making unusual proposals, and putting them up for wider consideration by the scientific community.

Note that the corrections to mathematical errors in the original paper restore a positive feedback, in line with all other research, and a feedback value of 2.3 that is low but within the range of the standard IPCC estimates of something from 2 to 4.5. In my own response that I had been working on recently, this was something I had been trying to sort out as well, so it is good to have it confirmed by more expert commenters.

Bottom line... we are right back with positive feedback from the ERBE data, though with some suggestions that the value may be a bit below what is indicated by other methods. This is right back in line with Gregory and Forster (2006) which I had cited previously; they obtained a value of 2.3 +/- 1.4 for climate sensitivity as well.

Cheers -- sylas

 

[mheslep]

as this whole topic is about to be shut down ...

By PF mentors? Why?

 

[mheslep]

link?

 

[sylas]

By PF mentors? Why?

See [thread=367803]New GW/CC Policy[/thread] in Forum Feedbacks and Announcements.

Personally I am opposed to this decision, but I recognize that there's been a lot of thought gone into it and so I'm working with the situation. If you want to express thoughts on the matter, it should be in that thread, I think.

Cheers -- sylas

PS. I have fixed the link with something that should work better for the time being, but it still can't give the full paper without subscription. I've left the doi link as a possible link to try in the future, when the relevant journal issue comes out and it no longer appears in the "in press" link. Thanks for pointing this out.