Merged Recent climate observations disagreement with projections

[Wangler]

[qimg]http://culturaenfichas.com.ar/georgieva-hansen.gif[/qimg]

Alec, I don't know if I am the only one, but I can't see your graphic. Could you repost, please?

 

[Dancing David]

Neither I nor anyone else has claimed that there is no natural climate variability. That's been stated numerous times, and I think that everyone else understands it.

Well that defeats the the deny, minimize and delay tactic doesn't it.

It is about saying your opponent is saying things that they aren't.

 

[mhaze]

Yes? I will tell you what nonsense is. Here you have Georgieva's figure 6 and the same climate normals developed by Hansen & Co. all together....

Well, climatic normals consist of 30 years of weather readings averaged together and presented decadally.

So I'm not really sure why you have presented a data series through 2009. Is it climatic normals? If so I don't see how that could be. It would appear to be just temperature data.

Which is irrelevant to this particular discussion. But take the averages if you like and present them and make a point.

Neither I nor anyone else has claimed that there is no natural climate variability. That's been stated numerous times, and I think that everyone else understands it.

So if you agree the temperature has multiple causative forcings and feedbacks, why do you insist that a correlated relationship between the solar aK index and global temperature can be defeated with either of these arguments:

1. T leads aK index
2. delta T leads delta aK index

Just curious.

Umm...no it doesn't. First of all the value you quote is for R^2. Secondly the value is not for the correlation between PDO and "world temperature." Thirdly it is not even correct for the correlation D'Aleo tried to make.....

I didn't get it from D'Aleo and have not read that stuff you point to.

Got it from UN sponsored research by Russia on long term cycles in fisheries. Will dig a bit.

It's remotely possible that you are right and it is R^2.

 

[aleCcowaN]

Well, climatic normals consist of 30 years of weather readings averaged together and presented decadally.

So I'm not really sure why you have presented a data series through 2009. Is it climatic normals? If so I don't see how that could be. It would appear to be just temperature data.

Which is irrelevant to this particular discussion. But take the averages if you like and present them and make a point.

Your typical argument "it is not, and if it is, it is irrelevant".

You had there the dataset I used -follow link in my post above-. I averaged the temperature for a 30-year period, what is the period selected and adopted by international agreement a couple of generations ago. The average value for 1979-2008 is represented on 1994, which is represented 8.5 years to the right of 1971-2000 Georgieva's figure.

Also, in Georgieva et al's figure the x-axis scale is decadal and so their values. You may ignore all the intermediate values in my figure. I left them as it make easier seeing both temperature series. You may make a quarrel of all this: do it if you please, I'll redraw it decadally and you'll get two temperature series even more similar.

Then: it is.

The next step is about relevancy. The question here is if it is relevant to your argument. Well, in a next post I will answer it. I'd like to re-read some of your posts an quote two or three that may fit, just to avoid having you escaping asymptotically into a new topic, as it is customary.

 

[mhaze]

Your typical argument "it is not, and if it is, it is irrelevant".

You had there the dataset I used -follow link in my post above-. I averaged the temperature for a 30-year period, what is the period selected and adopted by international agreement a couple of generations ago. The average value for 1979-2008 is represented on 1994, which is represented 8.5 years to the right of 1971-2000 Georgieva's figure......

So the left side is the first averaged value, the main body is the averages point to point, and the right hand side is a dotted line indicating the raw data past the 1994 average?

 

[aleCcowaN]

So the left side is the first averaged value, the main body is the averages point to point, and the right hand side is a dotted line indicating the raw data past the 1994 average?

All the values in blue share the same criteria and data source. Here the figure again, now in canonical climatic normals (based on Gregorian calendar. For Chinese, Roman, Babylonian, etc., go to the previous figure and select any one-each-ten set that fits -though years are still 365/366 days-)

I will come back to this later:

By the way, you claimed Georgieva's temperature were Hadley's, didn't you? Let's remember

The Georgieva article uses the well known and understood A/K solar indices, and the Hadley Center temperature data. All instrumental measurements have certain precisions. These are of course known to some degree of accuracy, and they are either stated, or referenced. Based on these facts, scientists proceed to look for relationships and draw conclusions.

[I will come back to this petite lesson of yours on epistemology and how you preach by example]

Where did you get that information?

I wonder why, if we are using climatic normals and 0 deviation is period 1961-1990 (or 1951-1980 in Hansen's), why on earth there's no zero value in Georgieva's graph. It looks like they used the three datasets available (including Hansen's) and averaged them.

 

[aleCcowaN]

Here the graph with lines to see the gaps:

Which climatic normal is 0 anomaly?

About their work: This quote:

In Figure 6 the long-term variations in global temperature are compared to the long-term variations in geomagnetic activity as expressed by the ak-index (Nevanlinna and Kataja 2003). The correlation between the two quantities is 0.85 with p<0.01 for the whole period studied...

So the sunspot number is not a good indicator of solar activity, and using the sunspot number leads to the under-estimation of the role of solar activity in the global warming.

Somebody knows where I can find Nevanlinna and Kataja's ak-index? I can't get it (just a pile of anti-GW advocacy). I mean the values of the index as used in figure 6.

I'm trying to check the correlation value but no data is provided in Giorgieva's paper but those plotted in figures.

I have other figures that show that using those climatic normals in the article could be more against the hypothesis than in its favor. In the end I would be surprised if a part of GW doesn't correlate with solar activity, as even T⁴ can be approximated to linear for 1K variation. Or was in doubt the fact that the Earth irradiates almost all the energy it gets from the Sun? (say 97% in a most furious GW process)

 

[Wangler]

Somebody knows where I can find Nevanlinna and Kataja's ak-index? I can't get it (just a pile of anti-GW advocacy). I mean the values of the index as used in figure 6.

Alec, I think I found it:

http://www.geo.fmi.fi/MAGN/K-index/ak-aa_index.txt

The file is ak-aa_index.txt.

Did you try Google? The above link was found in the second Google return on a search of "Nevanlinna and Kataja 2003". I didn't even have to follow the Google return link, as the http address was shown as part of the Google search return. Cut and paste, and I was there.

A plot of 30-year running averages from ak-aa_index.txt looks like this:




I think that this is Figure 6, in essence (minus the temperature data).

 

[mhaze]

also look at this site for explanations

http://www.ngdc.noaa.gov/stp/GEOMAG/kp_ap.html

I'm puzzled that a search would bring up lots of "Denier sites", since these indices have been being gathered for quite a while and are standard solar measurements.

I definitely think the naming is confusing - aa, ap, K, aK, etc.

.....By the way, you claimed Georgieva's temperature were Hadley's, didn't you? .....Where did you get that information?

I wonder why, if we are using climatic normals and 0 deviation is period 1961-1990 (or 1951-1980 in Hansen's), why on earth there's no zero value in Georgieva's graph. It looks like they used the three datasets available (including Hansen's) and averaged them.

The article says...

global temperature anomalies (deviations from the values in the base period 1961-90) from the Climatic Research Unit (Jones and Moberg 2003) are compared to R.

Jones runs the Hadley Center, if I recall correctly. Their baseline normal was 1961-1990.

FYI

I don't see any reason one could not or should not compare geomagetic solar databases to Hansen or other temperature databases. Sometimes the argument is made that one or another database is picked strictly because it better confirms the desired result. And the same applies to debates over which solar index to use, of course.

 

[fsol]

I didn't get it from D'Aleo and have not read that stuff you point to.

Got it from UN sponsored research by Russia on long term cycles in fisheries. Will dig a bit.

It's remotely possible that you are right and it is R^2.

Any luck with that at all?

 

[aleCcowaN]

Last figures:

Georgieva's cites this paper

Jones, P. D. and A. Moberg. 2003. Hemispheric and large-scale surface air temperature variations: An extensive revision and an update to 2001. J. Climate 16:206-223.

You have it here

No climate normal in that paper, but referenced Hadley's. I took Hadley standard series 2 -as it is suggested in that article- and got this:

I also plotted data from crut3

Thank you a lot Wangler for those ak-index data. I'll come back later to the finding-not-finding part, but for now I made my own calculations and plottings and it fits almost exactly with your figure and Giorgieva's (I took 1858-1880 period instead of 1851-1880 for the first point and repeated 1999 value for year 2000):

Then I used good'ol Pearson to get some correlations. Here are the series:

Then I conclude the r=0.85 with p<0.01 may be pretty right. I will get back later with my own conclusions (I have a bunch of students now for consultation -final exams next week) so I like to be seated calm in a quiet environment and write those conclusions (now I can't help bursting into laughter each time I take a look at the paper and all figures) [Sorry for my grammar and spelling, but being in a hurry plus laughter...]

 

[aleCcowaN]

Finally:

The first time I saw Georgieva's paper I thought "yet another paper that proves the Sun is heating the Earth" as in these times of people trying to climb in the social scale through education many people have cravings for inventing gunpowder every time they get inspired by watching Hiroshima's atomic mushroom. But I tried to keep my mind open as something interesting might be there, or maybe it was a good paper that has felt into wrong hands.

I argued that business of temperature series to get mhaze admitting that global temperature scales are not intrinsically wrong. I think that is done here:

Jones runs the Hadley Center, if I recall correctly. Their baseline normal was 1961-1990.

FYI

I don't see any reason one could not or should not compare geomagetic solar databases to Hansen or other temperature databases.

Similarly, we discuss Georgieva 2005. It is based on publicly available data sets. They have to some degree of understanding, ranges of precision stated. Those are the premises on which the work is done. Not that the data sets are perfect, but that they are within the stated ranges. You can disagree with the premises, the method, the conclusions, etc.

No matter he continues arguing about the datasets:

Nonsense.

Go look up yourself the estimated precision of the Hansen database, and then that of Hadley Center. They are what they are. There is a significant difference. It doesn't just go away and cannot be attributed to a poster on an internet forum with whom you disagree. Attribution of the assertion of precision is to the respective scientists.

You argue the point between the Hadley Center numbers and the Hansen numbers, if you wish to get somewhere. But I take it you have now abandoned trying to defend the Hansen number set? That's smart. That is what even MacDoc recommended, as I recall.

I repeat:

Hadley 2 / GISS : r= 0,995
Hadley 3 / GISS : r= 0,994

then, it's difficult sustaining that those temperature sets are "significantly different" but any of them is good enough to present a suspected causation when aK-aa index is involved (no matter r decreases to 0,954). In fact, once is accepted the link between solar activity an earth temperature -was that in doubt?- the work can be taken and presented as a proof that temperature sets are intrinsically right.

But, again to Georgieva's. Does the paper explain how concrete deltas in solar activity are translated in delta-T? Not at all, as we can see if we change values in the temperature set, by multiplying it by 2 or dividing it by 2: r stays constant (the matter always was not the sun giving energy but the earth not giving it back into space at the usual rate). The following table and figure show that constant r value.

Two or three forum members tried to argue with mhaze about this, but he wouldn't listen. Correlation doesn't mean causation. Contemporary correlation not even points what causes what else. Many examples could be given: it probably gives a r= 0.8 when one compares number of UFO sightings and human capacity to heavily damage the earth (measured in kilotons of nuclear power) all as a proof that aliens visit us to avoid our destruction (even today there are less UFO sightings than 20 years ago because nuclear weaponry tends to decline). You can also correlate number of crop circles with guided missiles and get good values.

But it can be argued that aK index correlates with real, well measured temperature anomalies (not the double, not a half). Well, this in my next post.

 

[aleCcowaN]

Let's see now how good is related aK-index with temperature anomalies. In this table I played with numbers:

Some results (remember it is R² and not R):

Georgieva's paper gives 0.9045

I invented a CO₂ coefficient (B) to adjust temperatures by a supposed increment in CO₂ that follows roughly the real one. Using 1851-1880 period as a base and playing with it we obtain similar values of 0.8816 to 0.9007, even when we suppose CO₂ decreases temperature -as many say by thinking in dry ice (my goshness!)-.

When I play with the hypothesis that CO₂ increases temperature but has a 30 or 50 year delay, value sustains and we even get a better r. This is all completely arbitrary and capricious, and it has the only one intention of testing the sensibility of what is inferred from the paper: the conclusion is pretty insensible to any consideration other than the earth gets most of the heat from the sun.

Why does all this happen? Because we are comparing just variations in the margin of everything: an indirect measure for variation in solar activity -as if the sun got turned off when K index is 0- and temperature anomaly less than one degree in a 290K world governed by Stefan-Boltzmann T⁴.

Two or three forum members tried to argue this line of thinking with mHaze, but he wouldn't listen.

But this is not all of it because there is more to say about the paper and epistemology, and again to aK and Hadley's temperatures. I can tell in advance that thin haired apes are good to match two patterns and some of them can calculate some correlations. But we must remember

Another hint!

Solar physics has high scientific standards. You will not find silly and continual statistics, math and conceptual errors common in solar physics, like you do in "climate science".

Oh, yeah! they even get clean briefs twice a day ...

 

[mhaze]

Well now that is all very interesting.

Is some of it relevant to my original post on this matter, post 911?

.....we don't need to hear any more nonsense about "Solar can't have influenced or contributed to GW during the last 50 years, so it must have been man's effects".

I thought for sure you were going to ramble your way to some conclusion.

Or do you simply agree with my comment in #911?

....again to Georgieva's. Does the paper explain how concrete deltas in solar activity are translated in delta-T? Not at all, as we can see if we change values in the temperature set, by multiplying it by 2 or dividing it by 2: r stays constant ....

And please tell us what you conclude from this, if anything.

 

[aleCcowaN]

Well now that is all very interesting.

Is some of it relevant to my original post on this matter, post 911?

.....we don't need to hear any more nonsense about "Solar can't have influenced or contributed to GW during the last 50 years, so it must have been man's effects".

I thought for sure you were going to ramble your way to some conclusion.

Or do you simply agree with my comment in #911?

It's quite logical you are impatient ... like a child. No one needs that long explanations here but you -maybe one or two more-. Your quote "Solar can't have influenced or contributed to GW during the last 50 years, so it must have been man's effects" is just a strawman. Everybody know that the sun heats the earth and also know that variations in solar activity are accountable for 20 to 30% of net global warming during 20th century. This figure

based on "Combinations of Natural and Anthropogenic Forcings in Twentieth-Century Climate" GERALD A. MEEHL, WARREN M. WASHINGTON, CASPAR M. AMMANN, JULIE M. ARBLASTER, T. M. L. WIGLEY, AND CLAUDIA TEBALDI

show no important increase in that effect during the second half of 20th century but still it has a positive slope during all that century, then it correlates quite well with the total temperature change (r=0,8 or even better, specially if you can prolong the series and introduce calm 19th century values). This also provides the same good correlation values to any indicator of solar activity. Some of them will simply do better than others -let's throw up spots and try magnetism-. You only need both series (natural solar globalwarming and total anthropogenic+natural global warming) to be similar and voilà! you can always argue that the first one is the real cause of the second one -when you abandon correlations and move on other measurements that feint proves useless-. But it seems you can't understand it, as your next question seems to show

And please tell us what you conclude from this, if anything.

By asking that you show an alarming, extraordinary, even terminal, lack of mathematical grounds. I suppose both you don't understand and prefer not to understand. If you understood, please explain what the r=0,85 with p=0,01 (not "without") means in Georgieva's and what is the hypothesis tested that way in the paper -the keypoint of your use of it-. It's impossible to take your assertions seriously when you claim "relevant", "not relevant" and a lot of other things without showing any traceable criteria on your part.

When you are presented that your precious r-value keeps constant when we introduced changes in one series and also keeps approximately constant when we introduce another non-linear variation and time-gap variation, and even it decreases to 0.85 when introduced a linear variation from 1 to 2 as a multiplying factor, you should keep yourself and revise your very foundations -you may still find your Eldorado of no-warming, but this is not at all the path to do it-.

But I have to came back to Georgieva's and both variables compared and their intrinsic problems -the motive of my laughter- (this tomorrow)

 

[BenBurch]

MHaze, you are arguing with an actual scientist, you know... Might behoove you to learn some of the science.

 

[mhaze]

It's quite logical you are impatient ... like a child. No one needs that long explanations here but you -maybe one or two more-. Your quote "Solar can't have influenced or contributed to GW during the last 50 years, so it must have been man's effects" is just a strawman. Everybody know that the sun heats the earth and also know that variations in solar activity are accountable for 20 to 30% of net global warming during 20th century. This figure

show no important increase in that effect during the second half of 20th century... .....

Meehl's work (an earlier article) is the single reliance for climate sensitivity used in the IPCC documents. The article that you have referenced the graph from is here.

Regarding your colorful graph, may I direct you back to the subject header of the OP (Original post,thread title).

"Recent climate observations disagree with projections".

One may well wonder why, in a discourse in which facts are presented showing some problems with models' predictions, you would err by presenting a graph of one such ensemble (group) of model runs as "proof". Also, in briefly perusing the paper I did not see clearly what measure of solar that Meehl used. TSI, likely. But that isn't the same as what Georgieva discusses, is it?

Briefly: In a discussion about why recent climate observations disagree with projections, you offer as a rebuttal said projections.

...By asking that you show an alarming, extraordinary, even terminal, lack of mathematical grounds.......

I simply asked what you thought of it.

 

[Pixel42]

One may well wonder why, in a discourse in which facts are presented showing some problems with models' predictions, you would err by presenting a graph of one such ensemble (group) of model runs as "proof".

Ignore the line that shows the model projections and look at the rest. The other lines are the ones which make Alec's point - that greenhouse gases are the only forcing which correlate reasonably well with the measured temperature rise of the last 50 years - as I'm sure you understand full well.

 

[mhaze]

Ignore the line that shows the model projections and look at the rest. The other lines are the ones which make Alec's point - that greenhouse gases are the only forcing which correlate reasonably well with the measured temperature rise of the last 50 years - as I'm sure you understand full well.

Good question. But no, actually the subject of this thread is that measured and modeled "don't agree".

The top line that you refer me to may be considered the sum of forcings that best models presumed temperature. The other lines are individual forcings one by one. "Sum of forcings" is combined effect as they are not strictly additive. Sum of forcings may be either more or less than strict additive sum.

Meehl (2003 date I think, so his paper was written in 2001-2002 or so) looked at differing combinations of forcings, and based on those, made a conclusion.

The question asked by the OP is whether if the last decade or so of roughly level temperatures were input, what would the conclusion of the model be?

Areas which are suspect

• presumed strong CO2 forcing
• presumption that solar effects are limited to total solar irradiance (TSI).

But there could be others.

 

[Geckko]

Ignore the line that shows the model projections and look at the rest. The other lines are the ones which make Alec's point - that greenhouse gases are the only forcing which correlate reasonably well with the measured temperature rise of the last 50 years - as I'm sure you understand full well.

Be careful here.

The chart shows what the models attribute to various factors.

There is in fact no empircal study (just data, no models), that validate that blue line. It is an artefact of the model construction process (i.e. an Xppm rise in atmospheric CO2 concentration will drive a Y degrees C increase in mean global temperatures).