In previous posts we have stressed that discrepancies between models and observations force scientists to re-examine the foundations of both the modelling and the interpretation of the data. So it has been for the apparent discrepancies between the Microwave Sounding Unit (MSU) lower tropospheric temperature records (MSU 2LT), radiosonde records and the climate models that try to simulate the climate of the last few decades. Three papers this week in Science Express, Mears et al, Santer et al (on which I’m a co-author) and Sherwood et al show that the discrepancy has been mostly resolved – in favour of the models.
It is worth encapsulating exactly what the problems have been and why they have taken so long to resolve. The MSU records are derived from a series of satellites that have been in orbit since late 1978. Each satellite has had different calibration problems (due to orbital decay, sensor issues etc.) and stringing them together has been fraught with difficulty. Different groups have made different decisions about how to do this and this has lead to quite some differences in MSU products particularly between the UAH group (Spencer and Christy) and the RSS group (Wentz, Mears and colleagues) . The differences have been mostly seen in the trends, rather than the monthly or interannual variability, and so have been more difficult to validate. Incidentally, it is a clear sign of ‘cherry-picking’ when people only report their favorite one of the groups’ trends instead of the range.
There have been three principle MSU products: Channel 4, Channel 2 and the 2LT records. MSU-4 is a record of lower stratospheric temperatures, MSU-2 is mainly mid-troposphere combined with a significant chunk of the lower stratosphere, and MSU-2LT is an attempt to use more viewing angles to try remove the stratospheric influence from MSU-2 and leave a lower-tropospheric record. (Recent upgrades to newer satellite instruments with more channels have lead to the 2LT record being renamed the TLT record).
The disagreement with the models related mainly to the MSU 2LT record. Models do quite well at matching the history of MSU-4 (whose variability is a function mainly of ozone depletion and volcanic aerosol effects), and models also match the lack of significant trend in MSU-2 (which is affected by stratospheric cooling and tropospheric warming which cancel out to some degree) (i.e Hansen et al 2002). So the problem has been principally with MSU 2LT, which despite a strong surface temperature trend did not seem to have been warming very much – while models and basic physics predict that it should be warming at a slightly larger rate than the surface.
In the first Science Express paper, Mears et al produce a new assessment of the MSU 2LT record and show that one of the corrections applied to the UAH MSU 2LT record had been applied incorrectly, significantly underplaying the trend in the data. This mistake has been acknowledged by the UAH team who have already updated their data (version 5.2) so that it includes the fix. This correction (related to the drift in crossing times at the equator) mainly affects the tropics, and was most important for one particular satellite (NOAA-11). Interestingly, Fu and Johansen (2005) singled out this same satellite and this same correction as being the source of divergence between the different records, though without being able to say exactly what the problem was. The fix leads to an increase of about 50% in the UAH global mean trend (0.086 to 0.12 deg/decade). The new RSS version of the 2LT record still shows a higher trend (0.19 deg/decade), with the difference being due to the methodology used to splice the different satellites.
In a related paper, Santer et al compare the surface/lower-troposphere coupled tropical variability at different timescales in the data and in model simulations performed for the new IPCC assessment. At monthly timescales (which should not be affected by trends in the model or possible drifts or calibration problems in the satellites or radiosondes) there is a very good match. In both models and data there is the expected enhancement of the variability in the lower-troposhere (based simply on the expected changes in the moist adiabatic lapse rate as the surface temperature changes). The models have large differences in their tropical variability (which depends on their represenation of El Nino-like processes in the Pacific) but the results all fall on a line, indicating that the lower tropospheric amplification is robust across a multitude of cloud and moist convective parameterisations.
At longer (decadal) time scales, the models still show very similar results (which makes sense since we anticipate that the tropical atmospheric physics involved in the trend should be similar to the physics involved at the monthly and interannual timescales). However, the original UAH 2LT data show very anomalous behaviour, while the new RSS 2LT product (including the latest correction) fits neatly within the range of model results, indicating that this is probably physically more consistent than the original UAH data.
One additional piece of evidence that has been discussed frequently was the claim that the trends in UAH MSU 2LT closely matched those of the radiosonde (balloon) network (Christy et al, 2003). Since the UAH team have acknowledged the error in their analysis, the apparent match to the radiosondes now seems to have been fortuitous. This may partly be due to the coverage of sondes used in that analysis being biased to the high latitudes (since the effect of the error was principally in the tropics), or it may be because of undetected biases in the radiosonde network itself. In the third paper this week, Sherwood et al report on an apparent bias in the daytime readings of these radiosondes which, again, appears to have suppressed the trends in the data sets (Steve discusses this more fully in an accompanying piece).
It will not have escaped the notice of keen observers that the satellite/model discrepancy has been used extensively in certain circles to cast doubt on the models, surface temperature record and our understanding of basic physics. Some recent examples for instance, used the UAH 2LT record absolutely uncritically (despite the fact that there have been many previous revisions, and that other analyses give very different results). Recently, one of these authors was quoted as saying:
… as long as weather satellites show that the atmosphere is not warming, I cannot put much faith into theoretical computer models that claim to represent the atmosphere but contradict what the atmosphere tells us.
Since the satellites now clearly show that the atmosphere is warming at around the rate predicted by the models, we will report on his no-doubt imminent proclamation of a new found faith in models as soon as we hear of it…
Presumably there will still be some effort to make as much as possible of the remaining difference between UAH and RSS. Is there anything shaky about the splicing method used by either team?
The link to the accompanying piece in “(Steve discusses this more fully in an accompanying piece)” appears to be broken…
I’m a bit confused about what has been resolved and what remains open. You note that at decadal time scales, the new RSS data is consistent with the models, while the old UAH data is not; what about the new UAH data? Is it still not consistent, or just not yet analysed?
Similarly, do the trends in the new UAH data still match the radiosonde data, or is that not yet analysed? Perhaps this is in the “accompanying piece”?
Thanks!
Pleased to see that this particular ghost has been laid to rest. One question though; is it true that physical measurements from weather balloons show the same apparent discrepency between models and observations?
I seem to remember that this was one of the claims of a documentary made by the Danish state television company (Danmarks Radio) recently:
http://www.dr.dk/dokument/co-produktion/dommedag_aflyst.shtm (Judgement Day Cancelled)
“There’s something wrong with the theory of the anthropogenic greenhouse! Leading climate researchers say that neither the earths temperature nor sea levels are rising as predicted.” Some of the usual ‘international’ and ‘home grown’ sceptics where wheeled out (Professor John Christy, Jurgen-Peder Steffensen (Niels Bohr Institute – http://politiken.dk/VisArtikel.sasp?PageID=388617 ), Dr. Niels-Axel Morner…etc.. All depressingly predictable, but this is the standard fare of the Danish media – unfortunately the people are given what they want to hear.
Funny title, BTW. Who sez scientists aren’t funny?
I’ve always found this website useful in deciphering the debate; it is a little dated (late 2003), but it gives good insight on the concerns and thought processes of those deeply involved in the MSU analyses.
Keep up the good work folks,
Dano
Your charset doesn’t allow my little signature – that post with the link (curr. #4) was mine.
D
Forget my question in 4! I’ve read the ‘Climate Science for Dummies’ reports over on Climateark….
Hot off the electronic press: Spencer’s spin is at http://www.techcentralstation.com/081105RS.html . An RC response would be great. The splicing question I asked about in #1 above seems even more like the critical issue. I’d really like to learn more about this, so any comment or pointers would be very much appreciated.
It may be of interest to note that the latest UAH TLT v5.2 still shows the odd annual
cycle over the Antarctic which I pointed out in a GRL report.
“Evidence of possible sea-ice influence on Microwave Sounding Unit
tropospheric temperature trends in polar regions”
GEOPHYSICAL RESEARCH LETTERS, VOL. 30, NO. 20, 2040,
doi:10.1029/2003GL017938, 2003
Also, the UAH middle troposphere result (TMT) does not show the reversal in temperature
which I found in the TLT product. Here’s the MT data:
http://vortex.nsstc.uah.edu/data/msu/t2/tmtdayacz7998_5.1
There may still be a need for more work on the UAH TLT analysis, IMHO.
A comparison of the new MSU data and the surface based temperature record of Jones and Moberg is available here:
Wikipedia:Satellite_Temperatures.png
If one limits analysis to the interval Jan 1983-Dec 2004, both trends are very similar: 0.182°C/decade and 0.163°C/decade for the surface and satellite data respectively.
However, there still is an unexplained discrepancy with the very earliest satellite data showing values that are noticably warmer than the ground data. If those first 4 years are included, a larger discrepancy remains in the trends, showing 0.170°C/decade and 0.116°C/decade for the surface and satellite data respectively. The UAH group has acknowledged this discrepancy in the early data (personal communication), but offers no specific explanation for it.
Ref: Comment #10
When I was working on my paper (see #9), I first looked at the zonal anomaly data on a yearly basis. There was a rather strong warming seen in the UAH TLT over the high latitudes in the SH in 1980-81. As I recall, that is the reason I focused my analysis on the Antarctic region.
There is a brief period in the satellite record in 1980-81 when there is only one MSU in orbit. That was on NOAA 6, with an afternoon LECT. The lack of an AM satellite during this brief period may have slightly biased the data. Or, there may have been a higher latitude regional warming that year??
A discussion of the merging procedure can be found in:
Mears, Schabel, Wentz, “A Reanalysis of the MSU Channel 2 Tropospheric Temperature Record” Journal of Climate, Volume 16, pg. 3650-3664, November, 2003.
http://www.ssmi.com/papers/msu/A_Reanalysis_of_the_MSU_Channel_2_Tropospheric_Temperature_Record.pdf
Gavin, just to be clear. Your paper looks at 2lt v5.1, not v5.2? So v5.2 may be consistent with the models, although no-one’s tested it yet?
[Response: Yes. Only v5.1 was available for that analysis, but you would need to contact Ben Santer for what difference v5.2 may make. -gavin]
Eric (#9), have you tried looking at the new RSS dataset for anomalies in the annual cycle?
Reply to Rick Watkins #4
Judgement Day Cancelled
Most of the information presented in “Judgement Day Cancelled” is picked from The Oregon Petitions and other sceptical homepages and is in contents similar to the editorial in WSJ “Kyoto By Degrees” (6/21/05). I suppose that the producer of this TV broadcast and WSJ have done the utmost to present the best arguments against climate change.
The best discussion of the arguments presented in “Judgement Day Cancelled” and WSJ can be found in RealClimate:
https://www.realclimate.org/index.php?p=167
I have tried to do my best to discuss the “Judgment Day Cancelled” in Danish language from my limited knowledge about climate science. Danish speaking readers of RealClimate will be able to find my arguments on
http://www.dr.dk/dokument/design2003/debat.shtm?&gbCmd=showThread&pID=AE185AC3F6200E29C1256F580049DF71&gbStart=1&gbnull=y#GB2
Some of the postings are in English language.
When this TV-program is broadcasted on some other TV station worldwide, please use the RealClimate as the best source.
I have been very pleased about RealClimate.
So version 5.2 of the satellite data still shows a 7.5 year cooling trend is in progress. Perhaps Gaia has already acted to correct the so-called harmful warming.
Re: Comment 14
You’re certainly a prolific contributor to the DR debate site! I’d like to make a more lengthy reply when I have time. My ‘home ground’ would be
http://vejret.tv2.dk/debat/ where I attempt some information/news dissemination (largely, but hopefully not entirely ignored…!)
Re #15: This post is just plain trolling. As Manny is very aware, 1998 is the current record warmest year and a simple trend plotted from it to the present will always show a cooling trend… until the next record year. Somehow I doubt Manny will undergo a conversion experience when that happens. In any event, if current trends hold up this year is looking to be a strong second to 1998. If Manny is consistent with his methodolgy, at the end of this year he should drop 1998 (which will then be the eighth year) from his 7-year plot and show a warming trend from 1999 to 2005. I won’t hold my breath for that, though.
So version 5.2 of the satellite data still shows a 7.5 year cooling trend is in progress
Correct. It’s also worth pointing out that there has been no major volcanic eruption since Pinatubo in 1991. This obviously had a large cooling effect as has been mentioned in earlier articles on this site.
For example here
https://www.realclimate.org/index.php?p=100#more-100
which includes the comment
“In fact, the Hansen et al (1992) paper actually predicted the temperature impact of Pinatubo (around 0.5 deg C) prior to it being measured”
It might be interesting to speculate what the effect on the temperature record (any record) would be if Pinatubo had erupted in 2001, say, rather than 10 years earlier.
This article also from realclimate
https://www.realclimate.org/index.php?p=142#more-142
tells us that the effect of Pinatubo lasted for “up to 3 years”. So in our ‘what if’ scenario 1992-94 would have been warmer and 2002-04 would have been cooler. This would be fairly easy to model.
I’m going to suggest that if this scenario had actually happened the 7.5 year cooling trend that Manny talks about in #15 would actually have extended back 15 years to 1990. Nor would we need to concern ourselves which data set to use. If the CRU surface temperature record were used I’m fairly sure it would produce a slight cooling trend – try it and see!
If this is the case then the warming throughout the 1990s is simply an artifact of volcanic activity early in the decade.
If UAH’s acceptance of the change means they recalculated their estimates from 0.09 to 0.12 degrees/decade rather than 0.19 as per RSS, is this issue really laid to rest? It’s still a big discrepancy.
Re#18,
Yes, 0.12 is still way too low.
But since the adjusted RSS measurements fit the models and S&C’s 0.12 figure (with the “anomalous behavior” of UAH over long time scales) doesn’t, I think you’re supposed to accept RSS over S&C.
Of course, it is supposedly “a clear sign of ‘cherry-picking’ when people only report their favorite one of the groups’ trends instead of the range,” so maybe that’s not the case?
I also noticed that S&C’s adjustment was indicated as 50%. Some people have used this 50% as a sign of how flawed S&C’s data was, because 50% is huge, right? Forget about the fact that a small amount of a small amount is relatively huge. I noticed that the RSS correction was not presented either as deg C/decade nor as a percentage.
I’m also confused by this part: “However, the original UAH 2LT data show very anomalous behaviour, while the new RSS 2LT product (including the latest correction) fits neatly within the range of model results, indicating that this is probably physically more consistent than the original UAH data…”
Why compare the new, corrected RSS 2LT product with the original, uncorrected UAH 2LT data? Why not compare to the new, corrected UAH 2LT data?
#17 – It’s worth noting as well that 1998 was exceedingly warm because of a strong El Nino, another natural occurrence. So when you put cooling natural volcanic eruptions in 1982 and 1991 together with a warming natural strong El Nino in 1998 you get a slight upward tilt to the global temperature graph, but it’s entirely due to the dominace of natural events.
For Manny and John Finn
Yet more betting on climate with World Climate Report
Now I’d invite wagers on when the trend won’t be negative any more — which could change if there is a big volcanic eruption. If not, my money would be on this year’s data switching the trend (though I’m not sure how the “new” MSU data affects this).
Re #15: 1998 (the first year in Manny’s proposed cooling trend) was the warmest year on record. You don’t have to be a statistician to understand that a simple trend plotted with a record warm year as the start date will show a cooling trend… until the next record year. Such an approach is the essence of cherry-picking. While we’re on the subject, it’s worth noting that 2005 is on track to be the second warmest year, plus of course the years in between haven’t exactly been cool relative to the long-term trend.
Re #18: The idea is to subtract out the Pinatubo effect and look at the long-term trend. It’s really very hard to show a cooling trend over the last 15 years given 1998 and the number of near-record warm years in that period. One could of course posit a series of convenient volcanos to blot all of those out, but I don’t think that’s a useful exercise.
Re #19: UAH (Spencer and Christy) have a major credibility problem at present. They have admitted to a long-standing “algebra” error of major proportions, and their new number of 0.12C/decade looks to be based on some shaky assumptions. The recent set of Science papers (which collectively involved many climate scientists) constitute the first comprehensive check of the UAH data and conclusions. Note also that Spencer at least has made his very strong bias clear in a series of articles on Tech Central Station. For now, the RSS number of 0.19C/decade (nicely in agreement with the models and surface record) seems more credible.
Re #21: But if you subtract out the volcanos and ignore the 1998 el nino year, there’s still an up-trend. And regarding el nino years, is there some basis for thinking that because there’s an el nino involved there is no anthropogenic component? Why wouldn’t we expect already-warm el nino years to just get even warmer (as in 1998) from a combination of natural and anthropogenic causes (noting that the net effect of all natural causes is a slight cooling globally)? When you say “entirely due to the dominance of natural events” it seems like a self-cancelling phrase; if it’s entire, there’s nothing else to dominate.
Re #20: Part of the confusion (and Gavin, please correct me if I screw this up) may be from the distinction between the raw satellite data and the corrected data (the latter being where the debate is) sometime getting glossed over. The passage you quoted uses “data” and “product” in a confusing way; I’m pretty sure that both refer to corrected data. Anyway, the Science papers (with a built-in delay for peer review and editing) of necessity related to the UAH v5.1 (corrected) data rather than the UAH v5.2 (corrected) data which was produced in the last few months (and was not published in a journal, and so did not have the associated delay). I’m confident that the UAH v5.2 results (er, product, er, corrected data) are already being worked over throroughly, and probably S+C are returning the favor. Likely it will be a bit of a scramble since the outcome will influence the contents of the forthcoming IPCC Fourth Assessment Report.
“UAH (Spencer and Christy) have a major credibility problem at present.”
I haven’t seen any real attacks on their methodology (other than the corrections that have been made), just “look, they made a 50% error (actually 43%), and other peoples’ results fit our theory better, so don’t pay attention to S&C anymore.”
“They have admitted to a long-standing “algebra” error of major proportions”
It’s “major” because the trend is so low in magnitude in the first place. If you look at a plot of v5.1 vs v5.2, it certainly doesn’t appear major.
“and their new number of 0.12C/decade looks to be based on some shaky assumptions.”
Such as?
“The recent set of Science papers (which collectively involved many climate scientists) constitute the first comprehensive check of the UAH data and conclusions.”
With all of the attacks on the conflicting results of “the satellite record vs surface record” and the public availability of UAH data and results, I find this hard to believe. It seems to me the RSS folks have found errors in S&C in 2002 or even earlier that S&C had to address, so they clearly too a hard look at the UAH data and conclusions. As mentioned in the article above, Fu and Johanson clearly took a hard enough look at S&C’s work to suggest a problem with the data from one satellite, and Fu likely took a good look at the UAH data for a publication last year.
“Note also that Spencer at least has made his very strong bias clear in a series of articles on Tech Central Station.”
Are you asserting that Spencer is skewing the S&C results due to his bias? And how strong are the biases of other climate scientists?
“For now, the RSS number of 0.19C/decade (nicely in agreement with the models and surface record) seems more credible.”
RSS, Fu, S&C, etc, clearly have differences of opinion on how to best process the data. Just because one way of processing best matches a desired result doesn’t necessarily make it the most credible. S&C have the first/longest-standing LT analysis. It has undergone a number of corrections/adjustments/revisions, and it has probably had far-more scrutiny than the rest of the other LT analyses combined. I find that more credible than an analysis which came-out in the last year or two and has been met with applause as opposed to scrutiny.
Up until now, S&C were able to closely match radiosonde measurements. With the S&C adjustment and apparently some balloon adjustments, we’ll have to see if that remains true. I’m certainly curious.
I too find S&C more credible than the analyses which have come out in the last year or two and have been met with applause as opposed to scrutiny. I have posted a brief paper with an alternative explanation of why the models and data differ. It can be seen at http://www.abmcdonald.freeserve.co.uk/brief.htm . Comments would be welcome to al@abmcdonald.freeserve.co.uk especially as I am planning a revised and expanded version.
Re: #10 One must point out this last comment at the referenced URL.
Note: In the above figure, there is still a significant discrepancy between the very earliest satellite measurements and the ground based measurements at that time. For this reason only the interval 1983-2005 was used in calculating each trend. Including the earliest years of the satellite record produces trends of 0.170°C/decade and 0.116°C/decade for the surface and satellite data respectively. The origin of this discrepancy is unclear.
Theory says that the troposphere should warm more rapidly than the surface. If we consider the period of warming from the mid to late ’60s then a warmer troposphere in 1979 should not be surprising. That this tendency does not continue suggests strongly that the surface warming trend is overstated. We seem to have a corrected sat. trend of .12 degrees/decade vs an instrument record of at least .16 degrees/decade, so even after the satellite correction, the results are still inconsistent with theory. Now it is time to reexamine the surface record. It is very likely that global warming in the last century, based on the averaged surface instrument record is overstated by 25%. Murray
Re #25: Excuse while I adjust to the concept of gaining credibility via a process of having to re-do your work after someone else points out an obvious error showing that you failed to thoroughly check your work over a period of many years. Of course there is at least one more shoe left to drop on this issue, so we shall indeed see.
I do find it fascinating that Spencer’s Tech Central Station articles began at about the same time (early last year) it must have become clear to him that he and Christy were in a bit of trouble. His recent article branching out from climate science into a defense of creationism seemed to me to be a burning of the last bridge between him and the scientific mainstream. I find it hard to imagine he didn’t understand the implications of taking that step.
Just to clarify, when I used the word “comprehensive” I was referring to the RSS having been the only one other than S+C to do a complete independent analysis. I may be wrong, but I don’t think Fu and/or Johanson did that; rather, they critiqued S+C.
My understanding of the remaining discrepancy between S+C and RSS is that it’s a question of how the data between the various satellites are merged. I find RSS more credible on this because that type of (very tricky) analysis of satellite data is their stock in trade, whereas for S+C it is very much not. Have a look at the RSS website to get a better sense of what they do.
Finally, it’s not clear to me how much scrutiny S+C really had since that error remained for so long. Not enough, anyway.
Re #27: Or we could imagine that it’s really hard to get accurate data from satellites. You might try emailing the RSS team with that question about the older satellite data; apparently they often respond. Lots of people have tried to re-examine the surface data in an effort to resolve the (formerly apparently larger) discrepancy in favor of the satellites, but these efforts have been successfully refuted. The most ludicrous was by some variant or another of the M&M team to ascribe a bias in the surface record to economic differences.
Also, of course you know that the current numbers are not inconsistent with theory since the difference falls within the predicted error range.
“Excuse while I adjust to the concept of gaining credibility via a process of having to re-do your work after someone else points out an obvious error showing that you failed to thoroughly check your work over a period of many years.”
It clearly wasn’t that “obvious” of an error, or it would’ve been discovered long before now. It may have been “based upon simple algebra,” but that doesn’t mean it was “obvious.” It’s also unclear how much of the change from 0.09 to 0.12 was due to the error and how much was due to the new diurnal adjustment technique S&C were going to adopt in the first place based on the AMSUs.
***…Mears & Wentz were additionally able to demonstrate to us, privately, that there is an error that arises from our implementation of the UAH technique. This very convincing demonstration, which is based upon simple algebra and was discovered too late to make it into their published report, made it obvious to us that the UAH diurnal correction method had a bias that needed to be corrected.
Since we (UAH) had already been working on a new diurnal adjustment technique, based upon the newer and more powerful AMSUs that have been flying since 1998, we rushed our new method to completion recently, and implemented new corrections. As a result, the UAH global temperature trends for the period 1979 to the present have increased from +0.09 to +0.12 deg. C/decade — still below the RSS estimate of +0.19 deg. C/decade…***
BTW, you might not want to bite the heads off of people who have had long-standing, published errors. They are all over the place, and I imagine most of them are a lot more “obvious” than accounting for diurnal corrections regarding satellite data.
“I find RSS more credible on this because that type of (very tricky) analysis of satellite data is their stock in trade, whereas for S+C it is very much not.”
Are you sure you don’t find RSS more credible in large part due to their higher temperature trend?
“Finally, it’s not clear to me how much scrutiny S+C really had since that error remained for so long. Not enough, anyway.”
Exactly how much scrutiny has the RSS methodology undergone? S&C update their results monthly (I can’t find a comparable data set on RSS, with LT results only available thru Dec 2004) and have had publications concerning their results, methods, adjustments, etc, every few years since 1990. That alone permits for more scrutiny. The fact that people have been dying to find fault with the UAH results ever since S&C started publishing them has resulted in even more scrutiny.
I don’t quite understand your logic. S&C aren’t credible because of a long-standing “obvious” error, and S&C’s methodology and results hadn’t been scrutinized enough because this “obvious” error had not been found. Yet RSS states that one of their goals is “to provide a complete and independent analysis as a check of these [satellite temperature] important results.” RSS folks like Wentz and Mears have had no less than 4 publications dedicated to this end, beginning as early as 1997. If you find RSS so credible, and if this is “their stock in trade,” then why has it taken them 8+ yrs to find the long-standing, “obvious” error, and how can you imply S&C’s works haven’t been scrutinized heavily if one of RSS’s goals is to check UAH?
The fact of the matter is that S&C and RSS use slightly different methodologies to interpret the satellite data, and this produces different results. S&C also saw the benefits of correlating their results to radiosonde measurements as verification, and RSS doesn’t. This isn’t the first time RSS has suggested a correction that S&C have adopted, but I also don’t see RSS making specific criticisms of/finding flaws with the methods S&C use, either – just differing opinions on what is better.
[Response: To clarify, the ‘personal communication’ from RSS to S+C concerned a sign error in how the diurnal drift correction had been applied in the S+C 2LT product. This was not a ‘bias’, this was simply a mistake. This had not come to light before because RSS had not previously attempted to do a 2LT retrieval, believing (correctly) that uncertainties in how to mesh the different satellites would affect this product more than MSU-2 or MSU-4 (which they had already worked on). The current differences between S+C 2LT (v5.2) and RSS 2LT trends are large precisely because of this.
With respect to the radiosonde comparison, it was clearly not sufficient to alert S+C to the error in their retrieval – most probably because the error mostly affected the tropics, while most of the radiosondes are in the mid to high latitudes. The fact that the reported match is still supposed to be as good (according to Christy) re-inforces the point that it is not a good discriminant of the at least some aspects of the retrieval. Plus, given the new results from Sherwood et al and the probablity that the radiosondes are biased low, the likelihood of a good match of the corrected sonde trends and the RSS 2LT product is quite high – I definitely would like to see that comparison though. -gavin]
“Water crisis looms as Himalayan glaciers melt”:
http://www.cnn.com/2005/TECH/science/09/09/himalayan.glaciers.reut/index.html