A Few Things Ill Considered

A layman's take on the science of Global Warming featuring a guide on How to Talk to a Climate Sceptic.

Monday, May 01, 2006

send this to... Digg it! | Technorati | Del.icio.us | Reddit | Furl | Spurl

Aerosols Should Mean More Warming in the South

(Part of the How to Talk to a Climate Sceptic guide)

This article has moved to ScienceBlogs

It has also been updated and this page is still here only to preserve the original comment thread. Please visit A Few Things Ill Considered there. You may also like to view Painting With Water, Coby Beck's original fine art photography.

Labels:

posted by coby @ 5/01/2006 09:57:00 AM  

13 Comments:

  • At May 01, 2006 1:51 PM, Anonymous Anonymous said…

    there is also the Ozone factor:
    http://tinyurl.com/ojch2

     

  • At May 29, 2006 4:26 PM, Anonymous Anonymous said…

    Hi Coby,

    Perhaps aerosols should not mean more warming in the south, for the reasons you have stated, but I think you’re missing a crucial point here: why did the SH also cool in the period 40s-70s? How can you have, as the AGW theory proposes, a CO2-induced global warming offset during that period by aerosols but the SH, where they play no significant role, also experiencing the same cooling as the NH?

     

  • At May 29, 2006 5:37 PM, Blogger coby said…

    This would require a bit more close examination of regional trends during mid 20th century, both model expectations and empirical observations to determine if there is any discerancy or not. But on general consideration I don't think you should expect a problem. If NH warming tends to dominate the global trend I don't see why NH cooling couldn't do the same. We observe less warming in the SH overall, perhaps there was also less mid-century cooling in the south.

    Yes, check here: http://data.giss.nasa.gov/gistemp/graphs/Fig.B_lrg.gif

     

  • At May 30, 2006 1:38 AM, Anonymous Anonymous said…

    Coby: those graphs are not hemispheric. I prefer the graphs and most especially the data files here: http://hadobs.metoffice.com/hadcrut3/diagnostics/
    It's quite obvious that the SH followed pretty much the same trend as the northern one during the 40s-late60s: temperatures below the average. In my view this is completely inconsistent with a CO2-induced global warming mitigated by aerosols. And it points at natural causes as the main drivers of the observed trends in both hemispheres. In any case, the last graph in your link (23.6S-90S) shows the same thing.

     

  • At May 30, 2006 7:49 AM, Blogger coby said…

    Mikel,

    Careful about the scale. The y-axis for the NH is 1.2oC high, for the SH it is only 1oC high so there is a bit of stretching there. Regardless, I think the trends are quite different. In the south warming resumed in 1950, rather than 1975 in the north. In the south, just a little bit more smoothing would remove that 15-20 year spike starting 1935ish and leave one continuous warming. In the N it is a 40 year change in trend.

    The SH in my graph shows a local high anomaly in 1945 and resumed general warming in 1950, like the hadcrut3 analysis.

    Why do you prefer a simple hemispheric division? The tropics behave differently in a lot of respects from the rest of the globe, it seems to make sense to divide it there. Also, "natural causes" sounds nice, but what exactly are you refering to?

     

  • At May 30, 2006 10:17 AM, Anonymous Anonymous said…

    ”Careful about the scale”

    Precisely my point when I say I prefer to look at the datafiles themselves (looking at those graphs with tremendous ups and downs is not the same as reading that the global temp has increased from 13.77C to 14.46C, a mere 5%, in the period 1976 to 2005, is it?). Looking at the Southern Hemisphere HadCRUT3 data file (non smoothed) you see that from 1945 to 1978 all years but 4 had negative anomalies in the SH, where positive anomalies didn’t really set in until the late 70s.

    ”In the south warming resumed in 1950, rather than 1975 in the north. In the south, just a little bit more smoothing would remove that 15-20 year spike starting 1935ish and leave one continuous warming”

    You leave me with the impression that you’re looking for a “good behaviour” of the data that simply isn’t there, Coby. Whether I look at the graphs or at the datafiles, what I actually find is more negative anomalies in the SH than in the NH for the period in question (more cooling) and even more cooling in the SH if I just concentrate on the southern extratropics, where we find some of the coldest years of the century. Also note that in this period AGW means that we should have seen warming in the south, not cooling, which is what we see. Look at the datafiles: temperatures went down in the SH after 1945 and didn’t start a steady increase until the late 70s. Is this not in direct disagreement with a CO2 induced global warming mitigated in the mid century by aerosols produced in the NH?

    ” "natural causes" sounds nice, but what exactly are you refering to?”

    Non-anthropogenic. And that is as much as I can say, really.

     

  • At May 30, 2006 1:56 PM, Blogger coby said…

    global temp has increased from 13.77C to 14.46C, a mere 5%

    or in Farenheit 56.79 to 58.03 = 2.2%
    in Kelvin 286.9 to 287.6 is only .24%

    It isn't clear to me there is much useful information in these numbers. Certainly trying to infer no danger from a low percentage figure is not very supportable. If I have one drink my blood alcohol content will rise to about .025% and I will feel it. 24 drinks will get me to .50% BAC and typically death. This is one example of many available.

    I don't know why you prefer to look at individual years and anomalies, those are artifacts of chaotic inter annual variability (weather), it is the smoothed trend that tells us what is going on with the climate.

    Is this not in direct disagreement with a CO2 induced global warming mitigated in the mid century by aerosols produced in the NH?
    I still disagree with your exageration of what interuption in warming there was in the south, but we can set that aside. The NH and SH are not seperated by any barrier. Presumably cooling in the N influences temps in the S. I don't see any theory killing problem here.

    An appeal to natural causes requires two things to be convincing:
    - a measured phenomenon and a physical mechanism whereby climate is influenced
    - a way for CO2 concentration variations to have little or no climate influence

     

  • At May 31, 2006 3:26 AM, Anonymous Anonymous said…

    Coby,

    Excuse my pigheadedness, but I think we should be able to come to a better agreement on the main point at stake here.

    In fact, I don’t think I have much to argue about your initial post but it does relate directly to the famous issue of the 40s-70s global cooling, and the alleged aerosols explanation. This is where the problem arises. I may not be able to break any theory with my comments here but it should be possible to use logic and compare theories with data that are readily available.

    I still disagree with your exageration of what interuption in warming there was in the south

    Let’s try to look at the same data for a moment. In the HadCRUT3 webpage I gave there is actually a link to the Northern-Southern hemisphere difference: http://hadobs.metoffice.com/hadcrut3/diagnostics/hemispheric/difference/. Whichever graph or data file you look at in that page, you will see that in the period in question there is a positive NH-SH difference in their respective anomalies from the 1961-1990 average, meaning a greater cooling in the south than in the north. Furthermore, if we look at the SH non-smoothed annual datafile and accept that the years 1944-1945 were the beginning of a trend-shift for global temperatures, we see that in these years the southern hemispheric average temperature was about 14.02C and 13.96C, respectively (assuming a 14C value for the ’61-’90 SH average, I don’t have any better value but it doesn’t matter very much). These average temperatures were never reached again in the SH until the late 60s and then it cooled again:

    Would you agree that there was a noticeable cooling in the SH during this period, in spite of the CO2 buildup and the absence of a significant aerosol forcing in this area?

    If so, how would you reconcile this with global warming counteracted in the mid 20th century by NH aerosols? You have already proposed an influx of cold air from the NH to the SH but perhaps you’re willing to admit that this is a bit ad-hoc and could not possibly explain how the SH actually cooled more than the NH?

    Aside from all this, you have made some very valuable statistical remarks but a couple of precisions:

    1) I was not saying that a low x% increase makes any statistical sample irrelevant (although when x is close to the error margins it can be). My point is that looking at a table of absolute temperatures (or expanding the y axis in a graph to show these) can be more intuitive and in some instances express better the relative magnitude of the observed changes. I wonder why I still haven’t been able to find any such tables/graphs anywhere.

    2) From a logical point of view, I am in a very strong disagreement with the constraints you impose to an “appeal to natural causes” to be acceptable. But in fact, I’ll give you that I should have better said “unknown” (including anthropogenic) rather than “natural”. I only chose the second because I keep thinking that the early 20th century global warming is more than likely natural and so it seems logical to assume that the mid century global cooling, in the main, must have had a similar origin. Just my thinking.

     

  • At May 31, 2006 8:30 PM, Blogger coby said…

    Hi Mikel,

    "http://hadobs.metoffice.com/hadcrut3/diagnostics/hemispheric/difference/
    Whichever graph or data file you look at in that page, you will see that in the period in question there is a positive NH-SH difference"

    This is not a detailed enough description. The NH has warmed more than the SH for approxiamately 1920-70. Specifically, it gains ground rapidly starting 1920ish to 1935ish and kind of plateaus until 1962ish where it loses ground rapidly. Do we agree so far? Now cooling was from around 40 to 70 so it does not look very straightforward to describe how that all fits together.

    Here's my attempt:

    From around 1900 to 1920, the N and S warmed more or less in synch at which point the N took off, warming faster than the S for 15 years. Then the S started warming faster and/or the N slowed down. In 1940 until ~1960 they moved they more or less together (both cooling) though I see the difference closing up a little mostly due to one big dip in the diff chart which is consistent with some greater cooling in the N. Then in 1962ish and for ten years, the N cooled while the S warmed, closing the gap between N and S completely, and then some. In 1970ish, the N stopped cooling and began warming faster than the south and continued gaining ground until 2000.

    Does that seem reasonable?

    Personally, I think we are really pushing the limits of meaningful scrutiny here, anything more finely resolved in time than this is really getting into the chaotic interannual variability.

    "Would you agree that there was a noticeable cooling in the SH during this period, in spite of the CO2 buildup and the absence of a significant aerosol forcing in this area?"

    When I look at http://hadobs.metoffice.com/hadcrut3/diagnostics/hemispheric/southern/ I see a big spike around 1935 to 1948ish, the down part of which could be called a cooling trend. But it really does start warming again in by 1950.

    "I keep thinking that the early 20th century global warming is more than likely natural and so it seems logical to assume that the mid century global cooling, in the main, must have had a similar origin."

    Well, I think it is pretty reasonable to say that most of the early 20th century warming was natural. And a portion of the mid century cooling was natural too. I like this graph to compare the various major forcings:
    http://en.wikipedia.org/wiki/Image:Climate_Change_Attribution.png

    Sorry to take so long to respond, but I did want to give it the consideration it needed! Thanks for the thoughtful discussion.

     

  • At June 01, 2006 11:47 AM, Anonymous Anonymous said…

    Hey Coby,

    Thanks for your detailed answer. It’s actually so detailed that at times I find it difficult to follow and, as you say, one gets the occasional impression of being scrutinizing statistical noise:-)

    I don’t quite agree with this: ”In 1940 until ~1960 they moved they more or less together (both cooling) though I see the difference closing up a little mostly due to one big dip in the diff chart which is consistent with some greater cooling in the N”. If you look at the smoothed graph or at the black line (best estimate) of the annual graph you clearly see a positive (NH) – (SH) anomaly value for the whole early 20s late 60s period. This basically means that during all of this period when it warmed it warmed more in the north and when it cooled it cooled more in the south (or there was positive anomaly in the north and a negative one in the south in the same year, which also gave a positive difference between both anomalies).

    Let’s see if I can sketch a coarser and perhaps more useful summary that we both could agree on:

    1- The average temperature trends for both hemispheres in the 20th century are similar: GW towards the first part of the century, GC in the middle and GW again in the end, with a dominant GW for the overall period.
    2- As the HadCRUT3 hemispheric difference graphs show, the mid century cooling was more pronounced in the south (it had larger negative anomalies relative to its 61-90 average). This relative (and I guess absolute) colder conditions in the south than in the north persisted throughout the whole 40s-late 60s period.
    3- In the mid century it’s possible to see a small relaxation of the cooling towards the 50s in the south. However, this relaxation is even greater in the north, to the point of making the anomaly positive for a short period. Hence if it’s fair to call this period one of global cooling for the NH and for the world as a whole, it’s even more fair to consider it like that for the SH.

    Now, how would all of this fit with the idea of GHGs-induced global warming mitigated by aerosols (global dimming)?

    Point 1 above suggests to me that both hemispheres have been responding to similar forcings throughout the whole century. Point 2 seems to discard completely the idea of mid-century global cooling being caused by aerosols, which didn’t affect much to the south and yet it experienced a more pronounced cooling.

    As for your alternative explanation in a previous post, I am trying to think of a world subject to a warming force due to increased GHGs since the 19th century…this warming didn’t materialize around the mid-century due to large quantities of aerosols emitted in the NH…somehow this NH cooling trespassed the equatorial belt and affected the SH, which not only didn’t warm but cooled even more than the NH…doesn’t sound very plausible, does it?

    My conclusion so far: the major cause of the global mid-century cooling was not aerosols.

    PS- We had a talk about that wiki page when discussing the possible origins of the early 20th century warming (which I didn’t find very convincing), remember? In this case, I’m afraid it lacks a hemispheric analysis.

     

  • At June 01, 2006 2:16 PM, Blogger coby said…

    Mikel,

    You have said a couple of things that I think indicate an understanding of the diff chart different from mine.

    "If you look at the smoothed graph or at the black line (best estimate) of the annual graph you clearly see a positive (NH) – (SH) anomaly value for the whole early 20s late 60s period. This basically means that during all of this period when it warmed it warmed more in the north and when it cooled it cooled more in the south "

    I think that it is the slope of the graph that indicates differing trends and even though there was a high N-S anomaly for the 30s-50s (+/-) the fact that it is basically flat for that period means there was no difference in what ever was going on, be it warming, cooling or stasis. That is how I arrived at my description.

    It is when that line is moving that there is an indication of differing hemispheric trends. Upwards indicates N is gaining ground (either it is warming more or cooling less), downwards indicates S is gaining ground (warming more or cooling less). If that is indeed the understanding you have as well then I will look harder again, if it isn't we should resolve this issue.

     

  • At June 01, 2006 3:40 PM, Anonymous Anonymous said…

    Yes, I also had the impression that we were interpreting those curves differently.

    I think that it all becomes clear if you look at the data. For any given year, say 2005, the NH anomaly is +0.643 and the SH anomaly is +0.280. Thus the NH-SH difference is +0.364, a positive value that represents a point in the curve above the x axis. A flat curve means that the NH-SH difference is stable and the slope just indicates the speed at which both anomalies converge or diverge, depending on whether the curve moves towards the x axis (where the difference is 0) or away from it. A bit complicated perhaps. As I said, looking at the real figures is often more intuitive than graphs.

     

  • At June 01, 2006 3:58 PM, Anonymous Anonymous said…

    A little ammendment to my explanation above: the x axis in these graphs does not represent 0. The 0 value is represented by a horizontal dotted line in the middle of the graph.

     

Post a Comment

<< Home