Talk:Satellite temperature measurement/Archive 1

Archive 1

Archive 2

Page creation

(William M. Connolley 19:57, 21 Aug 2003 (UTC)) I've moved here text from the hist t page and the gw page that were about the sat t record. Both pages were starting to accrete similar material and the sensible place to put it was here.

This page gets to have some balloon stuff on it too, since its closely linked.

Other trends

from the Mears et al version is +0.131 °C/decade, and from Vinnikov and Grody, +0.22°C to 0.26°C per decade [1].

Mears has no source, and V+G link requires "authorization".

Added Mears source (#1 hit from google, you lazy chap) and V+G via (gasp) Schneider... Besides which, I've just noticed that Mears *IS REFERENCED LOWER DOWN IN THE ARTICLE ANYWAY*. Oh good grief. V+G was referenced as a Science publication (2003) anyway so you have no cause for complaint at all. Especially as you keep inserting dodgy rubbish from Singer which has no better authority that self publication and newspapers.

Also, this is the first I've heard of anyone challenging Christy's claim that the satellite record shows hardly any warming. Where'd they get 0.13 degrees?

(William M. Connolley 21:26, 17 Aug 2004 (UTC)) Well, if wikipedia is restricted only to things that you've heard of, its in trouble.

And I heard somewhere that the .22 to .26 is too recent for Christy to even respond. Why is that going in an encyclopedia article?

Why should it be taken out just cos you don't like it?

If you put back Vinnikov, please do this:

• supply a working on-line link [2], or at least a reference to a peer-reviewed journal that could be checked in a good university library
• mention that Christy hasn't had time to respond yet [WMC: since when did Christy get an automatic right to rebut other peoples work? What happened to Spencer? Why haven't you read the trash on techcentral?]
• explain WHY Vinnikov gets such different results. [WMC: why not explain why S+C get different results?]
  • What does this mean? "To accurately retrieve the climatic trend, we combined the satellite data with an analytic model of temperature that contains three different time scales: a linear trend and functions that define the seasonal and diurnal cycles." [WMC: it means that you read the sciencemag article that you said was hard to get :-) ]

--Uncle Ed 17:52, 16 Aug 2004 (UTC)

An article on the Vinnikov-Christy difference

The differing conclusions come from the different ways the information was analyzed.

In his calculations, Christy adjusts satellite readings for inaccuracies he says are introduced by the fact that the satellite instrument is heated up by the sun.

"They did not remove this effect," said Christy, who was in Washington this week to attend a National Academy of Sciences meeting on remote sensing.

Notice the casual intrusion of NAS into this sentence to make you think that maybe NAS has approved or had some connection to C's statement.

"They allowed it to remain in the data and it corrupted all of their calculations, like a computer virus."

In an interview, Grody said he and Vinnikov were not convinced that Christy's adjustments should have been made at all.

He said that even if the data should be adjusted to eliminate the effect of the sun, he is not convinced Christy's approach is correct.

Christy said the temperature readings obtained by his adjustments are borne out by temperature readings collected by weather balloons.

Grody, a research physicist at the NOAA center in Camp Springs, Md., said he and Vinnikov used information provided by the manufacturer of the heat-detecting instruments to eliminate the effect of the sun's heat from the measurements.

(Sorry, I forgot to include a web link for this... --Uncle Ed 18:55, 16 Aug 2004 (UTC))

(William M. Connolley 21:33, 17 Aug 2004 (UTC)) S+C always say that only their record is validated by balloons but I have yet to meet anyone who believes this. C has sounded off in the press: when he gets a paper out, like V+G did, people will start taking him seriously. He has to fight off another recent attack too: Fu et al (http://cgi.cse.unsw.edu.au/~lambert/cgi-bin/blog/science/gwarming.html - see the end piece by Scott Church which I am told is mostly correct).

Church seems to conclude "it is not at all clear that the trends we're getting are even meaningful yet". Do you know where there is a cogent response to the diurnal criticism put forward at [3] that Singer cited at sepp.org?--Silverback 05:43, 30 Nov 2004 (UTC)

(William M. Connolley 09:17, 30 Nov 2004 (UTC)) The article you cite is so grotesquely biased I couldn't bring myself to read it all to find the bit you mention. Could you quote it here?

There is a larger analysis there, too large to quote, but hopefully this paragraph where it begins will allow you to skip to the relevant portion. The "double peak" this notes would seem to be a red flag, at least it is portrayed as such. "Yet another flanking maneuver (this time in the pages of Science) fell apart before its weight could be added to the fracas. Konstantin Vinnikov and Norman Grody employ their own statistical scheme to recalibrate the satellite data and account for changes in the satellites’ orbital drift. It ignores the work by both UAH and RSS, and summarily dismisses the weather balloon records. Their lower atmosphere temperature trend is about 50 percent greater than the surface measurements. Within the same analysis they determine there to be a diurnal cycle of temperature in the lower atmosphere that manifests a double peak (at 11 a.m. and at 9 p.m.) with a local minimum in between the two at about 3:30 p.m. Twin peaks, maybe; a single dip lynches the concept because it is a near physical impossibility."--Silverback 11:02, 30 Nov 2004 (UTC)

(William M. Connolley 18:58, 30 Nov 2004 (UTC)) Yep, thats part of the hopelessly biased bit. Describing V+G as another flanking manoeuvre is absurd. Asserting that V+G fell apart is also absurd. All they mean is, they dislike it. But... to return to your original point... no, I don't know about the diurnal stuff. Sorry.

--Silverback 20:37, 30 Nov 2004 (UTC) What you describe as bias, appears to be expressions of frustration at the things that they see as allowed to survive peer review. If the analysis presented is correct, then the peer review has been poor, especially for a reanalysis of data which purports to correct problems with previous analyses, the standards should be higher then, and should include strong attempts to reconcile the previous literature with the new results. I try to ignore the denigrating commentary on both sides, I find pressure for a premature consensus biased as well, and bound to inflame passions when coercive, command and control, political actions that have hundreds of billions of dollars of economic impact are being advocated. Agreement between parameterized models with quite different details in the physical processes is not confirmation which justifies a consensus.--Silverback 20:37, 30 Nov 2004 (UTC)

Recalculation of S+C/UAH series

Should the article be edited to show that S&C May 2005 show a sudden jump to:

[quote="http://climate.uah.edu/may2005.htm"] Global Temperature Report: May 2005

Global temperature trend since Nov. 16, 1978: +0.12 C per decade [/quote]

and

[quote] Notes on data released June 10, 2005:

UAH is reprocessing the complete global temperature dataset to include a new correction, according to Dr. John Christy, director of the Earth System Science Center (ESSC) at The University of Alabama in Huntsville.

"The April and May 2005 results include that new correction," Christy said. "We expect to have the complete dataset available in time for the June Global Temperature Report." [/quote]

or is it best to wait for the recalculation?

I've already added a note to the page about that change. At the moment the note is v brief, because I'm assuming that they will be explaining themselves pretty soon - it would be very odd if they didn't - this is a major change.

Also, I've removed the picture: its now obsolete and misleading. In fact it was always misleading, because it only had the S+C data on it, but the S+C data on it is now wrong, although we don't yet know why. William M. Connolley 2005-06-29 16:32:21 (UTC).

Don't you always insist upon retaining published material? Why are you now removing information, particularly before you know the details? (SEWilco 29 June 2005 16:42 (UTC))

The UAH series in that graph is now known to be wrong. There is no question about this - S+C say so on their web site (though they phrase it differently, as in the process of recalculation). Given that the trend has gone up by 50%, the graph is now misleading/wrong. There is no justification for retaining data that the originators admit is wrong. OTOH, the links to the S+C website still exist. BTW, please go off to GWC and justify your unjustified revert, if you would be so kind. We're supposed to have given up unexplained reverts.

Scientists Find Errors, Satellite Data Now Matches Balloon Data [4]. In a nutshell, the satellites were drifting in orbit and increasingly reporting night-time temperatures as daytime data. Once the drifts were detected and measured, and the data adjusted, the satellite data now measures a temperature rise consistent with balloon data.

Is this the error discussed above? Simesa 15:08, 12 August 2005 (UTC)

Yes. Dragons flight 15:13, August 12, 2005 (UTC)

Precision of data in table showing effects of choosing different periods

There seems uncertainty ranging from 0.12 to 0.26. Is it appropriate to show 7 decimal places???? (The UAH only show 2 dp in the reports I have seen.)

30June 13:59UTC

Fair point. I've knocked off a few dp's. William M. Connolley 2005-06-30 15:00:45 (UTC).

Figure caption

On the caption of the new figure showing the RSS MSU data it says that the trend is 0.0129 oC/decade. Shouldn't this be 0.129 oC/decade?

— Preceding unsigned comment added by Brian Jackson (talk • contribs) 16:43, 30 June 2005‎ (UTC)

Suggested Addition

It would be useful if someone could better explain why and how the satellite data was adjusted. As the article stands it looks to me as though people were concerned that the satellite data was not supporting their hypothesis so instead of changing their hypothesis, they simply changed the data. I'm not saying that this is what happened, but the current explanation sure paraphrases as "the data didn't show the warming we expected so we revised the data upwards". —Preceding unsigned comment added by Troyusrex (talk • contribs) 19:07, 28 May 2008 (UTC)

You're suggesting we try to explain how come Spencer and Christy managed to extract a cooling trend for so long from a dataset that when properly examined shows warming? Well we could... William M. Connolley (talk) 22:33, 28 May 2008 (UTC)

A few changes to consider

Can someone consider a few changes:

1. Mears et al is quoted as .133 but is linked to [5] which shows .134 for TMT and .192 for TLT. Is TLT a better comparision to S&C UAH .123 trend?

You may be right, but I am not familiar enough with the issue to provide a definitive answer right now. Dragons flight

2. Can the effect of cherrypicking table that uses 5.1 be updated to 5.2? [6] contains some figures. Are these good enough (There seems a discrepancy between .116 and .123)? Even if not, it is clear the analysis has not only been completed but also published.

I have updated the table with my own calculation. Publishing the figures here would require the permission of the author, potentially doable, but unlikely to happen today. The diffence between 0.116 and 0.123 is the difference between Dec 1978-Dec 2004 and Dec 1978-Jul 2005. Dragons flight

3. Radiosondes - would it be sensible to include something along the lines of Michael Tobis's comments on Real Climate

"Neither the sonde programs nor the MSU units were designed to detect long-term trends. They are enormously useful in other applications. The design and deployment of these instruments should not be criticized on the grounds that they are not especially useful for purposes for which they were not designed.

The fact that the NCEP reanalysis is implicitly calibrated to a drifting (biased) instrumental record is something I had not heard discussed previously, though. It seems this should be a matter of some concern in studies of the long-term record. "

Can we also add the finding that there was a trend in solar heating bias of -.16K per decade (from Real Climate again) which is enough to make it compatable with surface warming, (however there could be other issues with this data)?

All of that seems appropriate, though one will have to think about where and how is the best place to add it. Incidentally, you can always take up a keyboard and start making these edits to the article yourself. Just go to the page and click on the "Edit this page" button and go to it. Though maybe I should warn you, once you start you might just keep coming back to do it over and over again. Dragons flight

Thanks for the invitation. However I thought it might be too contentious and too far from my areas of expertise to consider doing it without getting other opinions first.

4. The article references NOAA-9 as the problem satellite. The recent issue seems to be with NOAA-11 (see [7]). Does this need correcting or adding?

They are all problem satellites, depending on what your problem of the week is ;-). However we probably should mention that NOAA-11 played a significant problem in the recently corrected snafu. Dragons flight 15:39, August 13, 2005 (UTC)

crandles 13:09

junkscience

I removed the anons addition of junkscience. For one thing, its junk. For another, its a top-level link and doesn't address the satellite record. William M. Connolley 17:02, 25 November 2005 (UTC).

Corrected introduction

The previous version of the introductory paragraphs implied that the +0.123 °C/decade UAH trend has been increased by the +0.035 °C/decade correction- actually the +0.123 (updated to +0.129) figure is from v5.2 and already includes this correction.

Also, regarding the second paragraph, I'm pretty sure that the correction was applied only to the LT dataset, and not the T2 channel (which is why the latter is still on version 5.1). So I've changed the discussion to reflect this. Brian Jackson 11:01, 15 March 2006 (UTC)

New Satellite Data

Climate Audit has a plotted graph of some new MSU data on their site:

Spencer & Christy 6.0

The differences between the Northern and Southern Hemispheres are interesting.

--JSleeper 23:27, 30 September 2006 (UTC)

UAH v5.2 etc.

I'm out of my element with this. Can someone fix the article to account for v5.2 clearly in both the summary and in the "Discussion of the satellite temperature record" section. Also, the 5.1 data link is broken. The "table" doesn't make any sense to me, and it doesn't include the years that the paragraph says it does. I don't think it's useful in its current form. I'd be willing to help generate a graphic if there is a point to be illustrated here. Mishlai 23:07, 18 February 2007 (UTC)

What a joke

"Previously reported discrepancies between the amount of warming near the surface and higher in the atmosphere have been used to challenge the reliability of climate models and the reality of humaninduced global warming. Specifically, surface data showed substantial global-average warming, while early versions of satellite and radiosonde data showed little or no warming above the surface. This significant discrepancy no longer exists because errors in the satellite and radiosonde data have been identified and corrected. New data sets have also been developed that do not show such discrepancies."

It's funny, this is from the executive summary of the CCSP report. The report also reports that from 1979 to 2004, the discrepancy exists. How can anyone trust these people?67.141.235.203 20:48, 21 May 2007 (UTC)

It remains slightly unclear if there is a problem...

Though we clearly have a problem here. KDP: this [8] is wrong: the problem is between the various obs (inc the MSU) and the models/sfc obs.

Anon: please don't modify direct quotes as you did here [9].

As to how the problem should be described: given that the report concludes These results could arise either because “real world” amplification effects on short and long time scales are controlled by different physical mechanisms, and models fail to capture such behavior; or because non-climatic influences remaining in some or all of the observed tropospheric data sets lead to biased long-term trends; or a combination of these factors. The new evidence in this Report favors the second explanation the conclusion appears to be that it isn't a serious problem. "potential serious inconsistency" appears to be POV, as it could just as easily be "potentially non-serious incons"

William M. Connolley 21:22, 21 May 2007 (UTC)

The report itself says "potentially serious inconsistency" (p 11). Raymond Arritt 21:25, 21 May 2007 (UTC)

Sorry, didn't realize I was changing a quote on that lower part, thought it was a paraphrase which I was adding quoted info to. And WMC please don't add misleading information such as: "An inconsistency still exists between UAH and RSS in the tropics, the area in which tropospheric amplification should be seen, UAH shows more warming on the surface than the troposphere, and RSS shows the opposite. " Thanks.67.141.235.203 21:29, 21 May 2007 (UTC)

What are you talking about, old fruit? William M. Connolley 21:59, 21 May 2007 (UTC)

I was referencing your quote which is misleading, please review read the paper and you will see.65.12.145.148 22:21, 21 May 2007 (UTC)

I believe that he was referring to my edit (which is explained below) --Kim D. Petersen 22:08, 21 May 2007 (UTC)

Hmm - the report itself says (page 63 [10]):

Compared to the global trends, the tropical trends show even more spread among data sets, particularly in the lower stratosphere16. The result of the greater spread is that the range of plausible values for the difference in trends between the surface and troposphere is larger than that for the globe as a whole. Similar to the global case, in the tropics the UAH satellite plus the two radiosonde data sets (RATPAC and HadAT2) suggest more warming at the surface than in the troposphere, while the opposite conclusion is reached based on the other two satellite products (RSS and UMd). Resolution of this issue would seem to be of paramount importance in the interpretation of observed climate change central to this Report.

I may have misinterpreted this - if so its my fault entirely. --Kim D. Petersen 22:05, 21 May 2007 (UTC)

As i read it, in this particular quote (which was already there): "(r)esolution of this issue would seem to be of paramount importance in the interpretation of observed climate change central to this Report". They are talking about differences between the products, and not between models and observation - and that was the part that i was trying to highlight. --Kim D. Petersen 22:10, 21 May 2007 (UTC)

Hmm, it seems to be ambiguous in its language. As I understand it, "the issue" is primarily the difference between 1=(sfc obs + model results/basic theory) and 2=(tropospheric temperatures). It doesn't help that there are multiple substantially different versions of 2. Note that the para from the report above is forgetting the problems with the sondes William M. Connolley 22:24, 21 May 2007 (UTC)

I vote since the paper says "potentially serious inconsistency", let's use that wording over the made-up (and POV) modifier WMC came up with.65.12.145.148 22:25, 21 May 2007 (UTC)

I agree, go with what the official report says rather than what Will says. Even the IPCC calls it substantial. ~ UBeR 01:21, 22 May 2007 (UTC)

I made a couple of corrections to the page. It is confusing because there are two issues. But pick any data set you wish, and the discrepancy between observations and the models are significant. All but one are 'very significant'. Saying "we've figured out the discrepancy and it's not as bad as before the reanalyis is fine". BUT, the models are still way off. And if the energy is not in the trop.(observations), the trop can't be heating the surface, period.65.12.145.148 01:36, 22 May 2007 (UTC)

updating of the figures.

An Anon made some updates of the MSU and RSS data - but i've found some discrepancy [11] says what the Anon changed to 0.184°K/decade - but on the same page there is a paper about the changes from v2.1 to v3 [12] stating the old value 0.208°K/decade - now my question is: Which one is correct? Or is it both - depending on what timeseries average you choose? --Kim D. Petersen 04:55, 30 May 2007 (UTC)

Hmm, tricky. Its hard to tell what they are using - I think I will email them William M. Connolley 08:18, 30 May 2007 (UTC)

All data on their page is now from 3.0. The change paper is interesting - the tropical stuff looks important William M. Connolley 10:27, 31 May 2007 (UTC)

I'm going to delete this:

An alternative adjustment introduced by Fu et al [5] finds trends (up to the end of 2004) of +0.19 °C/decade when applied to the RSS data set [6]. A less regularly updated analysis is that of Vinnikov and Grody with +0.22°C to +0.26°C per decade (Oct. 03) [7], [8].

Using just the T2 channel (which include significant contributions from the stratosphere, which has cooled), Mears et al of Remote Sensing Systems (RSS) find (through February 2006) a trend of +0.135 °C/decade [9]. Spencer and Christy of the University of Alabama in Huntsville (UAH), find a smaller trend of +0.054 °C/decade [10].

It's dated, irrelavant and more confusing than informative.67.141.235.203 13:50, 30 May 2007 (UTC)

The two numbers given (RSS & UAH) are comparable and regularly updated.67.141.235.203 13:52, 30 May 2007 (UTC)

Please don't delete them. The information is far from irrelevant. If its confusing, then it needs to be further explained, not removed William M. Connolley 14:26, 30 May 2007 (UTC)

I'd like to second William's request not to delete the data. -Friendly Neighbour 14:38, 30 May 2007 (UTC)

The data presented here is about 3 years old now. When is it going to be updated? Periander6 (talk) 19:46, 11 June 2008 (UTC)

Ref request

This [13] needs a ref, please William M. Connolley 16:43, 30 May 2007 (UTC)

Hegerl, Gabrielle. Influence of Patterns of Climate Variability on the Difference between Satellite and Surface Temperature Trends. J. Clim 2002.

Hansen, J. Climate forcings in Goddard Space Studies SI2000 Simulations J. Geophys. Res. 2002 —Preceding unsigned comment added by ? (talk • contribs)

Or [14] William M. Connolley 10:57, 31 May 2007 (UTC)

Thanks for the refer. It's neat that tropical amplification shows up in shorter time scales, but nothing more. Tropospheric temp varies alot, so over short periods, it will vary alot(thats what the Models say). More importantly Models project significant amplification in the tropics on decadal time scales. That means over long periods of time it trends a good bit higher than the surface temp. No observations show this. Don't get hung up on the "day to day" vagaries.67.141.235.203 15:38, 31 May 2007 (UTC)

None? Rss is 0.2 oC/dec in the tropics [15]. What do you think the sfc is? William M. Connolley 11:27, 4 June 2007 (UTC)

yes, None. 1.54X .17 > .2 oC/dec. Are you comfortable with a Se of that magnitude. (incidently until a few years ago the amplification was modeled closer to 1.8, but I wouldn't expect you to know that.)

Nor believe it, without evidence. And why have you picked 0.17? [16] says 0.12 or 0.14, and 0.12 * 1.5 < 0.2. Note that 1.54 is only one value and is over-precise William M. Connolley 19:46, 4 June 2007 (UTC)

just for fun, [[17]]. I picked 1.7 because it was the number used on this page. I used 1.54 because it is a published scaling factor, and used in practice. It seems you cherry picked numbers that suited you. If you are maintaining that observations and model results agree regarding tropical amplification, you are being unrealistic. And as this is fundamental to the physical possibility of enhanced greenhouse warming, I don't see how anyone can have much confidence that it is occurring.67.141.235.203 20:18, 4 June 2007 (UTC)

Note that when the observations are lousy (as is the case here), there's a conceptual difference between "models agree with the real world" and "models agree with observations." Raymond Arritt 20:24, 4 June 2007 (UTC)

RA, I understand the distinction, but we don't know which is accurate. There is great uncertainty, too much to be 90% certain, in my opinion.67.141.235.203 21:02, 4 June 2007 (UTC)

0.17 is global, we're talking about the tropics, so please drop the stupid accusations of cherry picking. 1.54 is just one study, there are plenty of others more recent. And I don't see why you're quoting that RC post, which hardly supports your POV, unless The instrument problems uncovered by these papers indicate that there is no longer any compelling reason to conclude that anything strange has happened to lapse rates suits you William M. Connolley 20:29, 4 June 2007 (UTC)

More this. "The troposphere---that part of the atmosphere that is involved in weather, about 85% by mass---is expected to warm at roughly the same rate as the surface. In the tropics, simple thermodynamics (as covered in many undergraduate meteorology courses) dictates that it should actually warm faster, up to about 1.8 times faster by the time you get to 12 km or so; at higher latitudes this ratio is affected by other factors and decreases, but does not fall very far below 1."

Also, why does the Global Troposphere heat more than the Tropical Troposphere? Polar Amplification, go for it there may be a Nobel in it for you....67.141.235.203 20:58, 4 June 2007 (UTC)

Jolly good, but an unsourced number (even in an RC post!) doesn't trump PR lit. So were still looking at 0.12 or 0.14 * 1.5, which isn't > 0.2. So, care to revise your earlier revision that *no* obs are consistent? Meanwhile, I'm surprised that you want me to teach you atmos phys - you seem to have been somewhat arrogant above. If you're rethinking, all to the good William M. Connolley 21:03, 4 June 2007 (UTC)

If you're comfortable hanging your hat on one set of cherry picked numbers(as you evidently are) by all means....feel free. If I repeat it enough, perhaps you will read it. "If you are maintaining that observations and model results agree regarding tropical amplification, you are being unrealistic. And as this is fundamental to the physical possibility of enhanced greenhouse warming, I don't see how anyone can have much confidence that it is occurring" (In spite of your one set of cherry picked numbers).67.141.235.203 21:19, 4 June 2007 (UTC)

You're missing the point, in your doomed efforts to defeat GW. We have the tropical sfc obs, of 0.12-0.14 oC/dec. We have the RSS series, with 0.2. This is perfectly compatible with the expected amplification, allowing for uncertainties. We also have the much-corrected UAH series, which isn't compatible. If all the obs and theory and models point one way, and one dataset points the other, which would you choose? William M. Connolley 21:33, 4 June 2007 (UTC)

Please read the paper you referenced, and the literature. No one but you is claiming that the evidence points YOUR way, quite the contrary. One set gets close (1.49 I believe) to its model and youre hanging your hat on it. Reread the literature to get my point, it's quite clear. Your cherry picking is mixing apples and oranges, resulting in a very bad fruit salad. ('Doomed effort to defeat GW', what a bizarre statement!)67.141.235.203 21:38, 4 June 2007 (UTC)

Nothing like a few inconvenient facts to end a 'debate'. It's interesting how far this debate goes, and in how many circles. Yet it's the very first & second laws of Thermodynamics that shoot so many holes in your AGW theory.65.12.145.148 00:29, 18 June 2007 (UTC)

If you're trying to convince us that you know anything about physics, it isn't working. Raymond Arritt 02:07, 18 June 2007 (UTC)

Nice personal attack, why not try speaking to the substance of the statement. You are right I don't know 'much' about physics. But what I do know, I don't ignore. Can you say the same?67.141.235.203 14:41, 18 June 2007 (UTC)

I have a degree in physics and I can tell you with authority that nothing in climatology (including "AGW", presumably "Anthropogenic Global Warming") breaks either first or second law of thermodynamics. If increasing concentrations of greenhouse gases did not increase the surface temperature, this could mean we missed a physics law (or more probably a process changing the radiative balance). I would say that rather expecting no global warming with increasing concentrations of greenhouse gases and decreasing (or even not stable) aerosol load, would imply changing the known laws of physics. If you do not know much about physics - as you admit - why do you think the physicists do not understand the laws of thermodynamics? --Friendly Neighbour 15:07, 18 June 2007 (UTC)

The GCMs programs are based on the laws of physics. The GCM projections do not match the observations (discrepancies are in size and direction of surface and tropospheric temp changes). We do not see tropical amplification, the global troposphere is warming greater than the tropical troposphere, THERE is the breakdown. So, either the models are wrong. Our understanding of the laws of physics is wrong. Or the observations are wrong. Or some combination of the three. If the models are wrong (it's not AGW causing increases in global surface temps), we should look elsewhere. If our understanding of physics is wrong, we should figure that out before we make these projections. And if the observations are wrong, we don't have the information necessary to make these projections. Or you can stick with your beliefs and ignore this.67.141.235.203 15:33, 18 June 2007 (UTC)

You said yourself (at least I assume it was you) that there are considerable errors in the satellite measurements. I agree that they are still large. The problem (for your kind of reasoning), is that such data are enough to be one more argument for a consistent and tested theory but certainly not enough to refute such a theory. To make it short: you would have to refute the null hypothesis of a warming trend with a significant statistical confidence (which would be difficult with low quality data) before you could start proving there is no warming trend. The burden of proof is certainly in your court because a "no global warming" hypothesis goes against the large data set we have globally as well as against the theory based on known laws of physics and against models using those laws. --Friendly Neighbour 16:11, 18 June 2007 (UTC)

No. The satellite record refutes the theory of AGW incorporated in the model, with regard to both amplification and regional variances. It's so fundamental (it goes back to the most basic laws of physics) it can certainly be used to refute the alarmist claims. You base your 'reasoning' on a common fallacy, mine is based on physical evidence.67.141.235.203 17:33, 18 June 2007 (UTC)

I'm not sure exactly what you think is broken. We have: theory and models say that the tropical trop trends should be greater than the sfc trends. We have sfc and MSU observations that show this is so. Where is the problem? William M. Connolley 18:00, 18 June 2007 (UTC)

Many problems. For instance, the troposphere in the higher latitudes should not warm faster than the tropical troposphere, or the surface for that matter. The observations that show tropical amplification show that. If the laws of phyisics are what we think, this is physically impossible.67.141.235.203 18:24, 18 June 2007 (UTC)

You concentrate on one poorly constrained data series to prove what? That the long time tropospheric data trends are probably similar in value to their standard deviations? This is possible. We need better data but we're making fast progress. Most of what we know on this subject was published on the latest 10 years. Or are you trying to prove that global warming is crap basing on this one uncertain trend and ignoring all the good quality data series? As I wrote above, this kind of reasoning will lead you nowhere. --Friendly Neighbour 18:57, 18 June 2007 (UTC)

BTW, what you wrote above does not parse. Lower troposphere should warm faster than surface due to latent heat flux. There is no reason I know why tropical troposphere should warm more than in higher latitudes. Ever heard about things like Hadley circulation? Or snow/ice caused positive feedbacks? --Friendly Neighbour 19:09, 18 June 2007 (UTC)

Yes, reality is that the higher latitudes have a greater temperature response than the tropics. This is a robust response that has been both modeled and repeatedly observed, and applies equally whether the climate change is of natural or anthropogenic origin. I'd be interested to know where .203 gets the idea that "troposphere in the higher latitudes should not warm faster than the tropical troposphere" given that it contradicts widely-known observational evidence. (Little Ice Age, anyone?) Possibly some law of physics or another. Raymond Arritt 19:17, 18 June 2007 (UTC)

This page. "Climate models predict that as the surface warms, so should the global troposphere; and that globally the troposphere should warm about 1.2x more than the surface and in the tropics the troposphere should warm about 1.5x more than the surface." Citations have already been provided. If you're trying to convince us that you know anything about arithmetic, it isn't working.67.141.235.203 20:00, 18 June 2007 (UTC)

Right. But why stop at LIA? During Last Glacial Maximum (about 20ka) the surface temperature of tropical ocean was 2-3 K lower than today while high latitude surface temperatures were 10-20 K lower thanks to the high positive feedback that glaciation creates. --Friendly Neighbour 19:50, 18 June 2007 (UTC)

"The troposphere---that part of the atmosphere that is involved in weather, about 85% by mass---is expected to warm at roughly the same rate as the surface. In the tropics, simple thermodynamics (as covered in many undergraduate meteorology courses) dictates that it should actually warm faster, up to about 1.8 times faster by the time you get to 12 km or so; at higher latitudes this ratio is affected by other factors and decreases, but does not fall very far below 1." cited above.67.141.235.203 20:11, 18 June 2007 (UTC)

(edit conflict and reset indent) Ah, I see now. Looks like .203 is misinterpreting "surface" to mean global mean surface temperature in both parts of the statement. In that light then indeed it's simply 1.2 < 1.5 But in this context "surface" means the surface beneath the local atmospheric column; i.e., "globally the troposphere..." and then "in the tropics the troposphere...". But in the tropics the surface doesn't warm up much, so we're taking 1.5x a relatively small number. On a global average the surface warms a lot more than in the tropics, so we're taking 1.2x a larger number. Raymond Arritt 20:18, 18 June 2007 (UTC)

I beleive you are incorrect. See the Sherwood quote. —The preceding unsigned comment was added by 67.141.235.203 (talk • contribs).

I'll add that the higher multiplier between troposphere and surface warming in the tropics comes from greater evaporation (latent heat flux in the physics jargon). Nothing in the text cited by 67.141.235.203 is about the difference of surface warming in and outside the tropics. He implied on June 4 that a "Polar Amplification" will get someone a Nobel. Apparently he has no idea that this is exactly the way Earth climate works :-) I have to ask my students tomorrow if they remember the fact from my lectures as I repeated it several times. It's really one of the most basic features of Earth climate. --Friendly Neighbour 20:40, 18 June 2007 (UTC)

By "Sherwood quote", do you mean this fragment of the article: The Sherwood et al. study [25], published August 2005, looked at solar heating issues and found a spurious trend of about -.16K per decade had been introduced into the record, asserting that this masked the true warming, particularly in the tropical regions [26].? If so, you did not understand it. It says nothing about the tropical surface trend being smaller than any extratropical one. Rather the other way, if such a small error masks it. For example Faraday Station in the Antarctic Peninsula has a trend of +0.56 K/decade (Vaughan et al. 2001). An error trend of -0.16 K/decade would certainly not mask it. --Friendly Neighbour 20:49, 18 June 2007 (UTC)

Tropical amplification is climte scientist jargon for the increase in tropical tropo. temp due to "latent heat flux". This amplification doesnt occur in the higher latitudes. (Also, according to greenhouse gas theory, the lower tropo. heats the surface, the surface doesn't heat the lower tropo.) All the jr. highs in our area have been out for weeks....67.141.235.203 20:52, 18 June 2007 (UTC)

Exactly! You start to learn. So why do you mix this effect with unrelated effect of increasing surface warming with increasing latitude? It is a well known and understood effect, as I've been trying to explain to you for several hours :-( --Friendly Neighbour 20:58, 18 June 2007 (UTC)

This Sherwood quote."The troposphere---that part of the atmosphere that is involved in weather, about 85% by mass---is expected to warm at roughly the same rate as the surface. In the tropics, simple thermodynamics (as covered in many undergraduate meteorology courses) dictates that it should actually warm faster, up to about 1.8 times faster by the time you get to 12 km or so; at higher latitudes this ratio is affected by other factors and decreases, but does not fall very far below 1." —The preceding unsigned comment was added by 67.141.235.203 (talk • contribs).

Sloooowly you are getting there. The amplification is larger in the tropics. But the warming being amplified, ie the sfc warming, is larger in the Arctic William M. Connolley 21:14, 18 June 2007 (UTC)

GW Theory says that the heat is mixed from the tropics. (Theory = 1.8 scaling factor in tropics vs. ~1 towards the poles, in higher latitudes the suface temp should increase 3-4X of the tropics) LOOK at the observations. This is not happening. These observations are inconsistent with the theory and simple physics. You claim to have a physics background, how can this be difficult? 67.141.235.203 21:09, 18 June 2007 (UTC)

There is no Global Warming theory. It's simply climate science. And what exactly is not happening? Do you claim that extra tropical (that is outside the tropics) warming is not larger than tropical? Than you simply do not know the literature. I gave you above an example from Antarctic Peninsula (about 5 K per century) which is an order of magnitude larger than the tropics. If you do not believe numbers, you may see it for yourself. For example if you do not see the extra tropical amplification in this 1884-2006 comparison, then I believe you are beyond my capabilities of helping you. Regards, --Friendly Neighbour 06:23, 19 June 2007 (UTC)

That's not what amplifcation is, it has to do with the amplification of heat increase in the troposphere relative to the surface, not simply an increase in surface. You should know that if you are teaching this subject.65.12.145.148 12:48, 19 June 2007 (UTC)

Are you really unable to understand written English? I tried to explain to you yesterday (at 20:58) that the troposphere warming amplification and high latitude amplifications are two unrelated phenomena which you yourself tried to treat as one. By the way, the two students I asked today both knew about the high-latitude amplification. I'm glad you learned about it as well. Only yesterday you claimed it's the other way.

So what is the next climate connected problem you have? However, try to be quick because someone will tell us sooner or later that Wikipedia is not the right venue for your education. --Friendly Neighbour 13:06, 19 June 2007 (UTC)

They are connected. If the surface temp increases one degree at the equator, we'd expect the surface temp near the poles to increase 3-4x that. An increase of 1 temp at the equator should correspond with an increase in the equatorial troposphere of approx. 1.8 (1 X 1.8) degrees. The increase in the troposphere near the pole should be about 3-4 degrees (3-4 X ~1). Are you claiming that the observations show this?67.141.235.203 14:01, 19 June 2007 (UTC)

No, they are not connected. Their physics is completely different. One is latent flux related, the other global circulation and albedo dependent. If you tried to read what I write (for your sake only!), you would know than not only 3-4 K/century but even over 5 K/century is recorded in the some polar zones (like Antarctic Peninsula). In other Polar regions no warming is recorded (mostly East Antarctica) but this is no surprise if you know that high latitude climate is mostly circulation controlled (because of low solar irradiation). This makes the phenomenon much more patchy than it would have been if solar irradiation and greenhouse gases were the only forcings. In the Arctic, anthropogenic pollution (aerosol) is complicating the situation even more (and good data are lacking for most Arctic Ocean) but the recent fast rate of sea ice decrease suggests the temperature increase must be quite high comparing to the tropics. One more example: over Western Greenland, the one decade of 1991-2000 have seen significantly statistical temperature rises of 2-4 K (Box 2002). Once more: this is per decade, not per century! This is much more than the tropical 0.13 K/decade surface trend (Christy et al. 2007).

Any more problems? --Friendly Neighbour 14:59, 19 June 2007 (UTC)

Are you claiming that the models don't model(roughly) what I said in my last post? (Either the parameters are wrong or modeled forcings are WAY off.)yes, yes, it's the best we have & we're 90% sure.....67.141.235.203 15:09, 19 June 2007 (UTC)

I have no idea what you are claiming unless it is that polar warming is no faster than the tropical one. If this is what you are trying to convey, than no, you are wrong. Both the models and measurements show much faster high-latitude warming. If you are claiming something else, please state it in a non contradictory way (your previous efforts failed miserably to state what you actually believe unless you are able to believe in several mutually contradicting ideas at the same time). Good luck! --Friendly Neighbour 15:19, 19 June 2007 (UTC)

From the info on the page, I misinterpretated the modeled increase in tropospheric temperature to be relative to the global surface temperature. Upon correction by Raymond, I compared the modeled increases in latitudinal tropospheric temperature to their respective modeled latitudinal surface temperature increase and the observations. The discrepancy is greater than my previous incorrect comparison. If the info on this page is correct, the troposphere in the higher latitudes should increase in temperature at 1.5 to 2X that of the tropics. If you are aware of any observations that show this please refer me to them.67.141.235.203 15:31, 19 June 2007 (UTC)

See e.g. here, specifically Fig. 3.6(C) or here, specifically Figs 3 and 4. Raymond Arritt 15:37, 19 June 2007 (UTC)

Dear 67.141.235.203, every day and sometimes in every comment you are claiming something different. A good example:

1. For instance, the troposphere in the higher latitudes should not warm faster than the tropical troposphere, or the surface for that matter (yesterday at 18:24)
2. the troposphere in the higher latitudes should increase in temperature at 1.5 to 2X that of the tropics (today at 15:31)

Are you changing your views every day or are you simply completely confused? --Friendly Neighbour 15:48, 19 June 2007 (UTC)

What about the southern pole?67.141.235.203 15:50, 19 June 2007 (UTC)

It's like talking to the deaf. I explained the reason today at 14:59. --Friendly Neighbour 15:52, 19 June 2007 (UTC)

Your point? The models are horribly inaccurate? Finally we agree.67.141.235.203 12:51, 20 June 2007 (UTC)

No, my point is that any purely thermodynamic model will not work for the polar areas because circulation (atmospheric but also oceanic, especially in the Arctic Sea) is the dominant forcing, anyway. Of course, nobody tries to use such oversimplifications. However, coupled models do give satisfactory results in the polar areas. Any more questions?--Friendly Neighbour 15:09, 20 June 2007 (UTC)

More of the US CCSP report quoted.

I really hate to mess with wikipedia articles on controversial subjects, but the US CCSP quote appeared to give the impression that all model/temperature problems had been resolved with the UAH 5.2v correction. I quoted a bit out of Chapter 5 to help clear this up. 65.91.20.162 15:15, 5 June 2007 (UTC)

Could you interpret that for me?Rod Serling 2001 03:54, 6 July 2007 (UTC)

NPOV?

Using the T2 channel (which include significant contributions from the stratosphere, which has cooled), Mears et al of Remote Sensing Systems (RSS) find (through February 2006) a trend of +0.135 °C/decade[9]. Spencer and Christy of the University of Alabama in Huntsville (UAH), find a smaller trend of +0.054 °C/decade

Isn't this statement kind of taking sides between UAH and RSS regarding which channels properly represent the troposphere and how they should be weighed? I think we should remove the part in parenthesis. 65.91.20.162 15:17, 29 August 2007 (UTC)

No. That T2 includes the strat, and that the strat has cooled, is not controversial at all. And its not taking sides anyway, since these values are for both RSS and UAH using T2 William M. Connolley 15:46, 29 August 2007 (UTC)

Satellite temperature table WP:OR?

Where did this table come from? At best, it's unsourced. At worst, it's original research. Oren0 (talk) 04:12, 10 January 2008 (UTC)

"The problems with the length of the MSU record is shown by the table below, which shows the UAH TLT (lower tropospheric) global trend (°C/decade) beginning with Dec 1978 and ending with December of the year shown." Dragons flight (talk) 06:22, 10 January 2008 (UTC)

Satellite temperature table

I see they corrected their data. Thus the inconsistency from 12/07 to 1/08. A 'heads up' would have been nice.

Version 3.1 Channel TLT - January, 2008

TLT 3.1 corrects a processing inconsistency in TLT 3.0: the production code changed between processing AMSU years 1998-2006 and year 2007. For TLT 3.1, all AMSU data have been reprocessed for full version consistency. The effect on TLT decadal trend was minor. —Preceding unsigned comment added by 67.141.235.203 (talk) 16:50, 6 February 2008 (UTC)

Comparison of Tropical Temp Trends with model predictions

This article conflicts with this published paper. —Preceding unsigned comment added by 67.141.235.203 (talk) 17:21, 7 March 2008 (UTC)

Could you be a bit less precise, please? William M. Connolley (talk) 17:41, 7 March 2008 (UTC)

I'm not sure if you're new to Wikipedia, one of the rules is that you should always be civil. Sarcasm is not acceptable. [[18]] —Preceding unsigned comment added by 67.141.235.203 (talk) 19:12, 7 March 2008 (UTC)

WP:SARCASM? I've seen the paper. Can't say I like it. But where is the conflict? William M. Connolley (talk) 19:44, 7 March 2008 (UTC)

The CCSP SAP 1.1 Executive Summary states:

"Previously reported discrepancies between the amount of warming near the surface and higher in the atmosphere have been used to challenge the reliability of climate models and the reality of humaninduced global warming. Specifically, surface data showed substantial global-average warming, while early versions of satellite and radiosonde data showed little or no warming above the surface. This significant discrepancy no longer exists because errors in the satellite and radiosonde data have been identified and corrected. New data sets have also been developed that do not show such discrepancies." The IPCC Fourth Assessment Report Summary for Policymakers states:

"New analyses of balloon-borne and satellite measurements of lower- and mid-tropospheric temperature show warming rates that are similar to those of the surface temperature record and are consistent within their respective uncertainties, largely reconciling a discrepancy noted in the TAR."

Sorry that you don't like it. =) too bad.67.141.235.203 (talk) 19:57, 7 March 2008 (UTC)

OK, well since those two assessments do indeed say that, what are you suggesting? William M. Connolley (talk) 21:01, 7 March 2008 (UTC)

I'm not sure if you don't understand or if you didn't read the article, but you see this statement "Previously reported discrepancies between the amount of warming near the surface and higher in the atmosphere have been used to challenge the reliability of climate models and the reality of human induced global warming." followed by this statement "New data sets have also been developed that do not show such discrepancies." is contradicted by this statement from the referenced paper "Models are very consistent, as this article demonstrates, in showing a significant difference between surface and tropospheric trends, with tropospheric temperature trends warming faster than the surface. What is new in this article is the determination of a very robust estimate of the magnitude of the model trends at each atmospheric layer. These are compared with several equally robust updated estimates of trends from observations which disagree with trends from the models." —Preceding unsigned comment added by 65.12.145.148 (talk) 02:07, 8 March 2008 (UTC)

I've read your paper. But I'm not at all sure its reliable. Thats only my opinion, of course. The models show more warming aloft; so does RSS; and the sondes have been shown to have a probable cool bias William M. Connolley (talk) 15:49, 8 March 2008 (UTC)

Published research or some guy on the internets opinon, which should we weigh heavier?? —Preceding unsigned comment added by 65.12.145.148 (talk) 01:45, 9 March 2008 (UTC)

The "some guy" happens to be a former climate modeller - but asides that - if you read the CCSP report more thoroughly, you'd find that both satellites, balloon and models find more warming aloft. That is the major conclusion of the report. There is a small discrepancy over the tropics (only) that (the report) concludes is most likely an error in the sat/balloon sets. --Kim D. Petersen (talk) 03:20, 9 March 2008 (UTC)

maybe you too did not read the referenced paper. if you do you will see it is in conflict as noted above. are you claiming that we we should take the offhand opinion of one "former climate modeller" over a peer reviewed paper? —Preceding unsigned comment added by 65.12.145.148 (talk) 04:46, 9 March 2008 (UTC)

No, I'm suggesting we should take the CCSP summary rather more seriously than one paper written to prove a point by skeptics. Which is exactly what we are doing William M. Connolley (talk) 17:05, 9 March 2008 (UTC)

Any legitimate science welcomes skepticism. If valid, it improves the science. If invalid, it is refuted. Refusing to acknowledge it is bad scienceMzmadmike (talk) 23:00, 1 November 2009 (UTC)

You are probably unaware but one scientist co-authored both. It's almost like you're cherrypicking the research. —Preceding unsigned comment added by 65.12.145.148 (talk) 20:40, 9 March 2008 (UTC)

Christy, yes. So... you're saying Christy is contradicting himself? I'm unsure what your point is. Oh, and don't be afraid to sign your posts William M. Connolley (talk) 20:55, 9 March 2008 (UTC)

Wikiproject Earth Proposal

Hello i have recently proposed the Wikiproject Earth. This Wikiproject`s scope includes this article. This wikiproject will overview the continents, oceans, atsmophere and global warming Please Voice your opinion by clicking anywhere on this comment except for my name. --Iwilleditu^{Talk :)Contributions} —Preceding comment was added at 15:39, 30 March 2008 (UTC)

Request to update the graph

Could someone update the graph to include data since 2006? Dan Pangburn (talk) 11:01, 12 June 2008 (UTC)

Yes, if anyone knows how to update a graph here is a more current one. http://www.nationalpost.com/893554.bin I'm not holding my breath.67.141.235.203 (talk) 18:20, 20 October 2008 (UTC)

Aqua

The Aqua sattelite was launched to end the need for questionable adjustments. Since it has been launched, the measured temperature has been falling, not rising. This may be real, or [incivility deleted - WMC] 124.148.69.18 (talk) 20:01, 20 November 2008 (UTC)

You're wrong William M. Connolley (talk) 21:03, 20 November 2008 (UTC)

Now that NOAA has read the graph for you, you might believe it. [19]67.141.235.203 (talk) 14:29, 3 March 2009 (UTC)

Update the graph

The graph starts in a reconstruction of what is probably a cool period except that we do not really have globa

124.148.69.18 (talk) 20:01, 20 November 2008 (UTC)

It's really pathetic that the graph on this article is outdated by four years. Would someone please update it?67.141.235.203 (talk) 15:19, 8 December 2008 (UTC)

It's not 4 years out of date. [20] Not that much different. I couldn't figure out how to upload it to Commons so that it complied with the original license. And even if I did, I'm not sure I'd use it — the original is a lot prettier. Atmoz (talk) 18:31, 8 December 2008 (UTC)

We must have different understandings of the term "out of date".67.141.235.203 (talk) 19:27, 8 December 2008 (UTC)

Latest Global Temperatures

Every month John Christy and Roy Spencer of UAH post an update to the temperature record on this website: [21]

Remote Sensing Systems also posts their satellite measurements on this website: [22]

I would like to see this incorporated into the article. There have been some complaints about updating the main graphic for the article, and perhaps since it's such a nice composition it should remain somewhere, but I think this article would be better served if the lead focused on graphs and/or summaries of the measurements of the two sets of data. Right now it is too focused on trends.

I also think the discussion of trends in the lead should be demoted to the body of the article. For one, the subject of the article is satellite temperature measurements, not satellite temperature trends. Accordingly, discussion about the trends should be a subsection of the article while the main thrust of it, and the lead section, should only address the measurements themselves. Perhaps a single sentence in the lead summarizing the trends would be appropriate, but no more than that.

For two, all the various citations of different trend calculations from Fu, Vinnikov, Hurrell, etc. are quite overbearing to be included in a summary. I am not a climate expert, nor are most of the folks who visit this page (presumably experts have more expert sources), so comparisons of varying calculations of trends from different scientists is inappropriate for the lead section.

Thoughts?

Zoomwsu (talk) 02:32, 10 June 2009 (UTC)

I would like to see this incorporated into the article - not sure what you mean by this. Do you mean the data (which is) or a more prominent link to the trends?

I also think the discussion of trends in the lead should be demoted to the body of the article - I'm rather doubtful. There are probably a few people with an abstract interest in the measurement and how it is done, but common sense (and a check on the edit history) will show you that most people are interested primarily in the trends William M. Connolley (talk) 07:32, 10 June 2009 (UTC)

Update the Graphic!

Come on folks. It is now half way through 2009 -- time to get it updated. SunSw0rd (talk) 13:01, 21 June 2009 (UTC)

Like: how about this one? [UAH_glob_temp.svg] SunSw0rd (talk) 13:09, 21 June 2009 (UTC)

It is missing RSS William M. Connolley (talk) 13:54, 21 June 2009 (UTC)

The data in UAH_glob_temp.svg (created about 8 months ago) is from the Atmospheric Science Department at the University of Alabama in Huntsville and it shows quite clearly that 1998 was some kind of anomaly. It also shows that the southern hemisphere might be returning to the pre-anomaly temperature patterns. (We won't know for another few years.) I am not sure why we can not show a more recent plot unless RSS data is included. Perhaps a second image with just UAH data can be added until someone develops a combined image. Q Science (talk) 05:35, 23 June 2009 (UTC)

As it happens RSS and UAH track each other very very closely. For example, see woodfortrees plot of RSS and UAH. Therefore I see no problem with adding the second image (already in wiki, in the public domain). Any objections from anyone? (A second point -- as both RSS and UAH are publicly funded, the data itself is in the public domain, though any image made from that data would not be unless so released.) SunSw0rd (talk) 17:58, 24 June 2009 (UTC)

Based upon no objections, I have added the wikipedia public domain satellite image which contains the data through 2008. Ideally we should have an updated image showing both RSS and UAH updated to the present date but one is not available. As the data sets are public domain I guess anyone could load them into Excel and generate a chart graphic and make it public domain and upload it to wikipedia. But since RSS and UAH are highly congruent I guess this is fine for now. SunSw0rd (talk) 16:05, 30 July 2009 (UTC)

You have an objection. As far as i can see the trendlines are polynominals (which give the wrong results at the endpoint), and since the trendlines aren't in the dataset.... --Kim D. Petersen (talk) 19:54, 30 July 2009 (UTC)

I am not sure of the objection? There are no trendlines at the endpoint. And may I point out that the liner trendlines in the existing image of RSS and UAH are also not in the dataset? If your objection to trendlines is that they are not in the datasets, then both images must be deleted. Please clarify your objection. SunSw0rd (talk) 17:25, 31 July 2009 (UTC)

Perhaps you would prefer we add this image? RSS and UAH graphic January of 1980 through June of 2009 This is based right on the raw data from RSS and UAH. There are no trendlines in this image. It is just the pure raw monthly temperature anomalies from the satellite sites, public domain data, public domain image. Objections here? SunSw0rd (talk) 18:51, 31 July 2009 (UTC)

I just plotted the data myself and your fig. is an accurate portrayal of the actual global data. I am deeply disturbed the the figure in the article seems to have some trend added to it. I suggest that it be removed ASAP and replaced with the correct data? Cheers MarkC (talk) 13:40, 5 August 2009 (UTC)

William Connolley reverted the RSS and UAH satellite comparisons asserting it was a "pointless dupl image". However it is not a pointless duplicate image because in fact it does not congruently map to the image provided by Dragons_flight. The released to public domain image I posted is directly based on the public domain data from RSS (link provided on image page) and UAH data (link provided on image page) -- whereas there is no evidence provided from where or how the image that Robert A. Rohde (aka Dragons_flight) was generated. As the images do NOT in fact map accurately to one another, and as the image I provided IS directly from the RSS and UAH data with no manipulation or processing whatsoever (including no calculated trendlines) I am reverting Mr. Connolley's revert -- AND -- if it is reverted again I will formally request moderation. SunSw0rd (talk) 13:44, 5 August 2009 (UTC)

I removed the graph because it is of abysmally poor quality. The fact that you didn't even label the time axis with years does not inspire confidence in your skills to produce scientific graphs. I suggest that you read a wonderful book by Edward Tufte, entitled "The Visual Display of Quantitative Information." Short Brigade Harvester Boris (talk) 14:26, 5 August 2009 (UTC)

Short Brigade Harvester Boris -- thank you for your feedback. Time axis now labeled with years. Grey background now replaced with white. Please note -- I am applying no "processing" to data from RSS and UAH public domain source -- no smoothing, no averaging, no trendlines, no nothing. Goal is to ensure this is as it actually is published. SunSw0rd (talk) 15:29, 5 August 2009 (UTC)

It still looks amateurish. Compare with the clean, clear appearance of the existing graph. (Hint: Excel's defaults almost always produce lousy graphs. Whatever Excel wants to do, you should usually do the opposite. And please do read Tufte's book -- the lessons you learn from it are worth the time.) Short Brigade Harvester Boris (talk) 15:49, 5 August 2009 (UTC)

Yes I understand Excel limitations however -- I want to ensure that image is not subject to "processing" which means that anyone else can duplicate and validate it -- a problem with the current Dragons_flight image is that the RSS and UAH trends in that image do not appear to correspond to the raw monthly data. In any case the current version has increased font sizes and bolded values. I do not want to require that specialized image processing software be used considering sensitivity among some quarters regarding this topic. Thank you, SunSw0rd (talk) 15:56, 5 August 2009 (UTC)

User Atmoz reverted new graphic stating: "there is no information in this graph not present in the other graph in lead". However this is not quite accurate on two counts. First the other graphic only had data through 2005 but was suddenly updated on July 31st 2009 after I posted the new public domain image. It now has data extending to present -- but look at the raw RSS and UAH data (from the links). The Dragons_flight graphic is NOT accurate for RSS and UAH monthly global temperature anomalies -- I don't know what the processing was that was used to generate the image but it is inaccurate. If anything based on this change the original image should be deleted. Second as Kim Petersen pointed out (above) trend lines are not part of the actual data set -- I deliberately did not include any trend lines in the new image for this reason (and it is a valid one). Thus I am reverting back. Please note -- due to the problems notes with the original (Dragons_flight) image if one must be deleted the new image is the one that accurately reflects the current RSS and UAH public domain data sets without any image manipulation or unknown effects processing. Thank you. SunSw0rd (talk) 19:16, 5 August 2009 (UTC)

I assume DF offsets the satellite data by a constant so that the anomaly from the satellites is taken with respect to the same years as the surface measurements. There is nothing wrong or shady with this. You can do that with anomalies. Where they are anomalous from is entirely arbitrary. There is nothing wrong with adding a linear trend line to the graph. Plenty of sources do that (both RSS and UAH actually do). KDPs objection was to adding higher order polynomial fits. Your graph contains nothing that the DFs image does not. It contains less information. It is redundant. It is not needed. Plus, the label on your dependent axis is misleading. -Atmoz (talk) 20:39, 5 August 2009 (UTC)

There is nothing wrong with a linear trendline (as in dragonflights). The reason i objected to the other graph was, as Atmoz quite correctly states, that it was based on polynominals. If you really really object to a trend-line without a source (which is foolish for a linear trend), then both UAH and RSS calculate a trend-line which can be sourced (its already in the article).

As for the new graph, i agree with Boris that it looks amateurish (sorry), and with Atmoz that it doesn't provide any information that isn't in the original. --Kim D. Petersen (talk) 20:51, 5 August 2009 (UTC)

The trend line was fitted by whom and by what method? Also, it is not correct to alter the anomaly offset of the actual data without stating that this has been done. The anomaly is calculated from the data minus the average of that data in some period. Since the satellite data does not exist for the period when the surface anomaly was calculated, this is unwarranted data manipulation. Therefore the real data should be shown and I have therefore reverted WMC removal pf this useful graph. I'm sure the author will in time try to make it prettier, but for me it contains the needed data without the distraction of undefined trendlines. Cheers MarkC (talk) 23:30, 5 August 2009 (UTC)

Linear regression. Duh. Altering the offset makes no difference to anything. Saying it does proves your incompetence. This graph is entirely useless. -Atmoz (talk) 23:45, 5 August 2009 (UTC)

Your insulting reply does nothing more than reveal your ignorance of model fitting methods. There are many ways to calculate a linear regression, depending on how residuals are treated, they may be minimized by least squares or other weights. Also, adding an offset does change the regression line because it alters the intercept. I suggest you go study some data analysis text before insulting those who actually know something about data analysis? Cheers MarkC (talk) 23:53, 5 August 2009 (UTC)

Without saying, it's OLS. -Atmoz (talk) 01:38, 6 August 2009 (UTC)

Then that should be in the legend. Will you add it? Also, I ask you to reconsider that the new figure does add value as it shows concordance between two data sets, and if it were extended to show the NH and SH data that would be an even better addition would it not? The rv by WMC in the middle of this discussion is quite uncalled for I think -but then I can't revert it now can I? So how about adding more WP:VER data to the figure? Another important point is that the sat data does not give the same temperatures as the land measurements and that fact is deliberately hidden i the old figure by unstated adjustment of the anomaly offset. This very poor science. Cheers MarkC (talk) 10:46, 6 August 2009 (UTC)

+++

RSS has just released the July 2009 anomaly -- jumped up to .410°C. When UAH releases I will update the image and, if the related public domain data source links from RSS and UAH are new files, I will update the file links on the image page. Please note that changes to source image will be automatically reflected on page when viewed (standard HTTP effect) therefore no page edits will be required to show the up to data global temperature anomalies from RSS and UAH. I will update and upload the image monthly as new data is released each month. Regards to all, SunSw0rd (talk) 20:12, 5 August 2009 (UTC)

Do you have access to any decent plotting software (e.g., tools in MATLAB, Kaleidagraph)? Something like that would produce a very nice fig. There might be some nice free plotting software available, but as a loyal subject of my lord, I don't buy my own software and therefore have never looked for the freebies. Awickert (talk) 11:58, 6 August 2009 (UTC)

You could use R. -- Originalwana (talk) 13:08, 6 August 2009 (UTC)

+++

The fundamental issues here that I see are these. (1) The current (lead) graphic has apparently been processed in such a way that it is inaccurate. For example -- look closely at the period between 1980 and 1985. In the raw data the high anomaly for RSS is 0.271 (March of 1983), and the high anomaly for UAH is 0.245 (also March 1983). Now look at the Dragons_flight image. It references for example the same UAH data set here but the image shows rather than 0.245 something that appears to be just less than 0.4 at that point in the graph. The reason for the discrepancy is not explained. In fact at multiple points the lead image does not appear to map to the data set referenced. I am open to an explanation as to what is happening here.

(2) A second difference relates to the RSS source files. Dragons_flight asserts that the source data for RSS is here, whereas the source file for the public domain image I uploaded is here. The difference is notable, and the distinction appears in two ways. First, the Dragons_flight source file is suffixed _v03_0.txt, whereas the one I reference is suffixed _v03_2.txt. The one I reference maps very congruently to UAH, whereas the version Dragons_flight references not only does not map to RSS, but it does not appear to map to the Dragons_flight image either. Again, compare the month of March for 1983 for a common reference point. As best as I can determine, no column in the RSS data file referenced by Dragons_flight maps to the Dragons_flight image.

It is for these reasons that I generated a public domain image from the public domain referenced data sets. Anyone can look at the image I provided and month by month validate that it accurately corresponds to the column one global temperature anomalies for both data files for both RSS and UAH. There have been comments that the image I posted is "amateurish". I would assert that the image as it exists now (see [here]) is very clear, easy to read, and easy to visually map the displayed data points to the corresponding data points in the referenced text files.

Please provide an explanation as to the discrepancies in the current "lead image" with the referenced data files for that image. Thank you. SunSw0rd (talk) 13:44, 6 August 2009 (UTC)

Well??? SunSw0rd (talk) 18:31, 7 August 2009 (UTC)

You will have to ask dragonflight - frankly the sat record has had so many errors, that it hardly matters what version is used, or how old it is... chances are that it will change rather drastically within 3-6 months, because of an error in the sats or in the algoritms.

Thats also the reason that i (still) do not think that your graph has anything new to show, and i'm also still of the impression that your graph is too amateurish to be used as an alternative or extra. --Kim D. Petersen (talk) 21:06, 7 August 2009 (UTC)

As for cleaning up the graph, (a) remove the horizontal gridlines and (b) remove the markers at each data point, leaving only the curves. Short Brigade Harvester Boris (talk) 21:18, 7 August 2009 (UTC)

I'll make a plot and throw it up based on the two links above, using the "global" data for UAH and the -82.5/82.5 data from RSS (I assume that those are latitude, so that would be the most global one). If someone can verify that those are indeed degrees of latitude, I'll create a new graphic and upload it. Awickert (talk) 02:05, 8 August 2009 (UTC)

I used this RSS link here and this UAH link here as my sources. RSS only goes from 70N to 70S, whereas UAH goes 84N to 84S -- although the reference I found went from 70 to 82.5. Thus the UAH plot should swing a little "wider" compared to RSS -- but they track each other pretty closely. SunSw0rd (talk) 23:21, 8 August 2009 (UTC)

Regarding Short Brigade Harvester Boris -- I personally like the horizontal markers at the 0.2 degree level -- as it is possible to get a more accurate view of discrepancies from the data tables if any. In addition, I don't really care for trend lines for an article dedicated to the satellite temperature measurements -- because they are not related to the topic. SunSw0rd (talk) 23:21, 8 August 2009 (UTC)

For someone who is specifically looking at the article for the purpose of determining whether there are any discrepancies from the data table, the horizontal gridlines perhaps are useful. For most other users they're just clutter. As for trend lines, RSS and UAH disagree with you -- they make a point of reporting the slope of the trend line and continually updating it. Short Brigade Harvester Boris (talk) 00:06, 9 August 2009 (UTC)

Thanks for the info. I will make the fig, sans horizontal gridlines (for presentation), sans trendlines (personal bias towards presenting data without trendlines, and to avoid the "wait a minute! this trendline is a scam!" thing), and with links to the data sources (so anyone with interest and a few minutes can create their own plot). I should have it up for your viewing pleasure tomorrow. Awickert (talk) 01:53, 9 August 2009 (UTC)

Shoot - I didn't do it and won't be around. I hope we can live without it for another week. Please ping me if I don't do it then. Awickert (talk) 03:37, 11 August 2009 (UTC)

(outdent) Finally made it. Just a couple issues I have to fix: somehow little lines ended up underneath the degree symbols, and the lines for the data are too narrow IMO. It also doesn't have the complementary traditional measurements that Dragons Flight has in their image. But before I address those issues, is it basically acceptable in your collective opinions or not? Awickert (talk) 19:38, 18 August 2009 (UTC)

I took the liberty and changed a few details in the graph. Hopefully this is an improvement. However, it seems I missed the discussion about horizontal gridlines. My opinion is that thin grey gridlines in the background are less distracting and therefore acceptable. Dragons_flight sometimes uses grey dashed gridlines. Btw, was the aim of this graph to mimic Dragons_flight's style? If so, a few other things may need to be changed. SPLETTE :] ^{How's my driving?} 20:33, 18 August 2009 (UTC)

This looks good to me. I don't think it should have comparison to anything but RSS and UAH -- that is the topic of the article. I think the style is fine. SunSw0rd (talk) 14:52, 19 August 2009 (UTC)

Thanks for the changes, Splette, I think they are a definite improvement. I sort of tried to follow Dragons_flight's guidelines, but it's not that important to me. So what I'd like feedback on is:

1. Would faint gridlines be an improvement?
2. Would it be good to include traditionally-measured temperature? (SunSw0rd thinks it unnecessary, but I think it would be useful for comparison.)
3. Should I make the lines representing the data be thicker?
4. Should I use different shades of red and blue to make it colorblind-friendly? (I'm leaning strongly towards yes on this one.)

Awickert (talk) 17:58, 19 August 2009 (UTC)

I vote yes on thicker lines and different colors. (Also, I think the tiny dashes at each data point are not needed.) Thanks. SPLETTE :] ^{How's my driving?} 18:03, 19 August 2009 (UTC)

I'll need to make a little bounding box to maintain a margin, but otherwise, how does this look? Thanks for letting me know about the dashes, I forgot to turn those off and I didn't see them. Awickert (talk) 16:03, 23 August 2009 (UTC)

Looking good. Suggest the axis labels be a bit larger. You could take off the "year" annotation on the x-axis to make room for them (it's fairly obvious that these are years). The "December 1978 to July 2009" annotation could also be removed. Short Brigade Harvester Boris (talk) 17:01, 23 August 2009 (UTC)

Thanks for the suggestions; did everything except the dates because someone always says "why is your graph out of date", and then we could say, "it's written that it's out of date, go fix it". Awickert (talk) 07:10, 31 August 2009 (UTC)

break

(outdent) It looks good. I looked at the file history for your image and it shows a lot of good work incorporating the suggested changes. It appears good to go. SunSw0rd (talk) 15:53, 2 September 2009 (UTC)

Thanks for the note and the ping at my talk; I missed this here. I'll add it into the article now. Awickert (talk) 20:45, 10 September 2009 (UTC)

Not sure where to put this, but this section seems fine. The UAH data is not global as it's indicated in the figure. I can't recall what it is offhand, and cannot find it on a quick glance at their website. IIRC it covers less area than RSS. -Atmoz (talk) 04:20, 15 September 2009 (UTC)

It was their "global" data from their site. I'll see if I can track it down. Awickert (talk) 05:17, 15 September 2009 (UTC)

That darn figure

I will create a new figure if the community wants one. I am putting a brief questionnaire after this. I will answer for myself first: Awickert (talk) 23:22, 14 September 2009 (UTC)

Please choose:

(a) Do you want a new fig or DF's one? (if you answer DF, you're done)

Weakly for new for basic update (new data, new algorithms) Awickert (talk) 23:22, 14 September 2009 (UTC)

Weakly yes. --Kim D. Petersen (talk) 23:39, 14 September 2009 (UTC)

Not sure what you mean by "new algorithms," but in general would prefer the new fig. Short Brigade Harvester Boris (talk) 23:54, 14 September 2009 (UTC)

SS pointed out that the values now were different than those that DF published, and someone said that they came up with new ways to compute the temperatures from the raw satellite data every so often and that that could explain the discrepancy (or at least I think someone did, can't find it now). Awickert (talk) 04:55, 15 September 2009 (UTC)

DF offset the sat anoms so they would align with the suf anoms. Nothing wrong with that. Any change made by RSS or UAH in their processing would be minimal compared to that. -Atmoz (talk) 17:55, 17 September 2009 (UTC)

Thanks for the info. I've stricken the "algorithms" part. I'll be out until Monday, but I'll remake the fig then. Awickert (talk) 17:59, 17 September 2009 (UTC)

Don't care. If new figure, it needs a trendline. Raw data is too difficult for some people to interpret. (Cooling since 1998, etc.) -Atmoz (talk) 04:44, 15 September 2009 (UTC)

(b) Do you want

(i) A trendline (please comment on what kind if you do)

Not necessary IMO, but not necessarily bad if it doesn't get in the way of the data. Would support a faint trendline that wouldn't obscure the data too much and/or the slope of a linear regression in the caption. Awickert (talk) 23:22, 14 September 2009 (UTC)

Trends are what people are looking for, not raw data or raw plotted data. Its in the dataset (ie. precalculated). So all in all it should display the trends. --Kim D. Petersen (talk) 23:39, 14 September 2009 (UTC)

Yes, but be sure to use the slopes reported by RSS and UAH. Otherwise our usual friends will hound you endlessly about the method used to compute the trendlines. Short Brigade Harvester Boris (talk) 23:54, 14 September 2009 (UTC)

Yes. Per above comment. -Atmoz (talk) 04:44, 15 September 2009 (UTC)

(ii) Conventional data for comparison

Strongly yes. Awickert (talk) 23:22, 14 September 2009 (UTC)

Define "conventional data"? If you mean the ground set, as a comparison, then yes - it would be nice. --Kim D. Petersen (talk) 23:39, 14 September 2009 (UTC)

Ground / Radiosonde / whatever DF used. Don't know what it is yet, so just said "conventional". Awickert (talk) 04:55, 15 September 2009 (UTC)

A strict comparison might be hard to do. The satellite data don't include the poles, but reported global and hemispheric average surface temperature data include the polar regions. You could figure it out from the basic gridded data but that would be a nuisance. Maybe zonal means are available somewhere. Short Brigade Harvester Boris (talk) 23:54, 14 September 2009 (UTC)

Yes. But difficult to do properly. The different data have different latitudinal boundaries, surface data has large holes in coverage, satellites and surface stations measure different things, so it's not right to directly compare them. Prefer to have older image done right than a newer image done wrong. (AFAIK, the DF image is done wrong.) Would also prefer this get done in the literature to avoid accusations of original research. -Atmoz (talk) 04:44, 15 September 2009 (UTC)

Thanks for the note, I have the latitudes for which RSS and UAH are given, and so I will see if I can average zonal means (weighting properly, of course, for their differing areas, just in case someone is listening). Atmoz is indeed right that published numbers will earn less animosity so I'll look there first. Awickert (talk) 04:55, 15 September 2009 (UTC)

Dude. It doesn't matter what you do. They aren't going to let you post the new graphic. No matter how many hoops you jump through. SunSw0rd (talk) 13:33, 22 September 2009 (UTC)

Error Bars

Several years ago, there were error bars on the data. I looked through the article history, but it doesn't appear to go back far enough. Should we get these restored? I recall them being pretty large and thusly, perhaps germane.

I know Cristy originally supplied error bars, although I have no idea how they've changed since his earlier versions. 66.195.102.82 (talk) 21:45, 17 September 2009 (UTC)

AFAIK, the current data supplied by UAH and RSS does not contain error bars. If you can find them somewhere, then we can discuss adding them to the graph. -Atmoz (talk) 21:56, 17 September 2009 (UTC)

I seem to recall error bars being mentioned in his actual papers on the matter, not the raw data feeds. Unfortunately, the only cite I have at the moment is from personal correspondence, and it is hopelessly out of date, to boot. 66.195.102.82 (talk) 18:18, 23 September 2009 (UTC)