Many readers will remember our critique of a paper by Douglass et al on tropical tropospheric temperature trends late last year, and the discussion of the ongoing revisions to the observational datasets. Some will recall that the Douglass et al paper was trumpeted around the blogosphere as the definitive proof that models had it all wrong.
At the time, our criticism was itself criticised because our counterpoints had not been submitted to a peer-reviewed journal. However, this was a little unfair (and possibly a little disingenuous) because a group of us had in fact submitted a much better argued paper making the same principal points. Of course, the peer-review process takes much longer than writing a blog post and so it has taken until today to appear on the journal website.
The new 17-author paper (accessible pdf) (lead by Ben Santer), does a much better job of comparing the various trends in atmospheric datasets with the models and is very careful to take account of systematic uncertainties in all aspects of that comparison (unlike Douglass et al). The bottom line is that while there is remaining uncertainty in the tropical trends over the last 30 years, there is no clear discrepancy between what the models expect and the observations. There is a fact sheet available which explains the result in relatively simple terms.
Additionally, the paper explores the statistical properties of the test used by Douglass et al and finds some very odd results. Namely, that their test should nominally inadvertently reject a match 1 time out 20 (i.e. for a 5% significance), actually rejects valid comparisons 16 times out of 20! And curiously, the more data you have, the worse the test performs (figure 5 in the paper). The other aspect discussed in the paper is the importance of dealing with systematic errors in the data sets. These are essentially the same points that were made in our original blog post, but are now much more comprehensively shown. The data sources are now completely up-to-date and a much wider range of sources is addressed – not only the different satellite products, but also the different analyses of the radiosonde data.
The bottom line is best encapsulated by the summary figure 6 from the paper:
The grey band is the real 2-sigma spread of the models (while the yellow band is the spread allowed for in the flawed Douglass et al test). The other lines are the different estimates from the data. The uncertainties in both preclude any claim of some obvious discrepancy – a result you can only get by cherry-picking what data to use and erroneously downplaying the expected spread in the simulations.
Taking a slightly larger view, I think this example shows quite effectively how blogs can play a constructive role in moving science forward (something that we discussed a while ago). Given the egregiousness of the error in this particular paper (which was obvious to many people at the time), having the initial blog posting up very quickly alerted the community to the problems even if it wasn’t a comprehensive analysis. The time in-between the original paper coming out and this new analysis was almost 10 months. The resulting paper is of course much better than any blog post could have been and in fact moves significantly beyond a simple rebuttal. This clearly demonstrates that there is no conflict between the peer-review process and the blogosphere. A proper paper definitely takes more time and gives generally a better result than a blog post, but the latter can get the essential points out very quickly and can save other people from wasting their time.
Sorry I should have been more precise in my previous post and given UAH SH land temperature anomalies:
Jul: +0.26C
Aug: -0.56C
Sep: +0.24C
http://vortex.nsstc.uah.edu/data/msu/t2lt/uahncdc.lt
I don’t have the RSS figures to hand but they won’t be very different in any event.
Quite tangential, but maybe of interest to one of the climate scientists:
We have another ‘natural experiment’ going on, rather like the 3-day shutdown of air travel in the USA after 9/11, and like the Beijing Olympics clean-air experiment.
http://www.nakedcapitalism.com/2008/10/baltic-dry-index-falls-another-107-down.html
Total ocean shipping activity has dropped dramatically. That’s almost all fueled by high-sulfur fuel oil; shipping is known to make linear clouds along shipping tracks.
The change in ship travel ought to show up in satellite imaging, I’d think — and in anything else?
Re: the volcanic forcings. The reason I raised the issue before was that simply eyeballing the graph at the top of p5 of the factsheet reveals the following: the main reason that the tropospheric trend is steeper appears to be the lack of peaks in the 1980s with higher tropospheric anomalies than surface anomalies.
There are in fact several points that I would make here:
1. La Nina events in 1985, 1989 and 2008 (beyond timescale of graph) show a pattern of tropospheric negative anomalies significantly exceeding surface negative anomalies. (Not so much 1999/2000 though.)
2. The El Nino event in 1998 (and weaker El Nino events from the mid-1990s) imply that the reverse pattern ought to be true for El Nino events.
3. However, this did not occur for the El Nino events of 1983, 1987 and 1992 – the major apparent cause of the relative steepness of the tropospheric trend.
4. El Nino 1983 and 1992 were associated with major volcanic cooling events – a plausible explanation for the subdued tropospheric temperatures relative to surface.
5. El Nino 1987 was not associated with such an event.
However, I note from
http://vortex.nsstc.uah.edu/data/msu/t2lt/uahncdc.lt
that the tropics reached an 6-month anomaly of +0.48C for the second half of 1987, a figure which does not appear to have been surpassed since in any year except 1998.
Also I’ve just located the RSS data at
ftp://ftp.ssmi.com/msu/monthly_time_series/rss_monthly_msu_amsu_channel_tlt_anomalies_land_and_ocean_v03_1.txt
and these show +0.47C for the second half of 1987, here just edged out by the two peaks in the 2000s (+0.50C and +0.53C) as well as the 1998 peak. (The graph in the factsheet shows the peaks in the 2000s at ~0.15C higher than the peak in 1987)
6. At the same time, I notice that the standard RSS figures for the tropics (as in link above) are for latitudes 20S to 20N, whereas the key figure in the factsheet (and the paper?) is for 30S to 30N – which extends beyond the tropics as commonly understood:
http://en.wikipedia.org/wiki/Tropical
7. La Nina 2008 will almost certainly have made the tropical tropospheric trend significantly shallower (relative to the surface trend)
I don’t draw any particular conclusion from this still cursory examination, except that it seems to me there’s significant leeway to come up with differing trends depending on one’s choices of parameters for the data.
chris schoneveld writes:
We know what caused it — the Maunder Minimum. Increasing solar constant brought us out of it. But the solar constant hasn’t increased significantly in 50 years, so it can’t be driving the sharp upturn in global warming of the last 30 years. We’re not still “coming out of the Little Ice Age.” We came out of it in 1850.
#100: Chris Says
Hadley’s Crutem3 SH land figures for Jul and Aug (Sep isn’t out yet) were +0.66C and +0.14C respectively, or +0.40C for the two months, compared with +0.72C for GISS. I suggest they are also skewed, but not so badly.
http://www.cru.uea.ac.uk/cru/data/temperature/crutem3sh.txt
Note that according to Crutem3, SH land temperatures in August were the coldest since 1992 and 1994 (post-Pinatubo cooling) and below the 1980s average of +0.15C. This would fit with anecdotal evidence from various countries in the SH, notably Australia (coldest August since 1944 in Sydney, for example).
UAH had Aug SH anomaly at -0.19C
http://climate.uah.edu/august2008.htm
You are comparing apples to oranges to pears. All the different data sets have different base periods for the anomalies. The levels of numbers should be different, therefore. Go back and adjust all numbers according to the anomaly base year. Then there would be at least some sense in the comparison.
By the way, GISS data does not show +.72C. For the globe it is .49C
http://data.giss.nasa.gov/gistemp/graphs/
and I am sure it is less for SH.
Also, the satellite data include lower stratosphere and are, by definition, not comparable to surface temps.
Chris, 2008 monthly S.H. temperature averages by NOAA NCDC are unlike the S.H. averages by NASA but are like the averages you show at your links. Your comment seems to explain that well.
However, I still don’t think anyone should become receptive to arguments for lower-than-consensus climate sensitivities.
Also, I would like to see your explanation, if you have one, for the for why 1931 was extremely hot and dry in the U.S.
http://picasaweb.google.com/npatnew/ClimateDataMidwestAK#
The Van Hoof et al. paper is interesting. Not very familiar with proxies but I’ve been led to believe I should trust ice cores more, but if the results hold up, would the amplitude of CO2 from Vostok over glacial-interglacials also be effected (if so, linearly?) or would this mostly be a higher frequency problem? I’d like to see a paleoclimate person here do something on that.
The LIA is partly due to solar forcing, but volcanic effects had a lot to do with it as well.
#105 Lauri
“You are comparing apples to oranges to pears. All the different data sets have different base periods for the anomalies. The levels of numbers should be different, therefore. Go back and adjust all numbers according to the anomaly base year. Then there would be at least some sense in the comparison.”
Why don’t you do it yourself? GISS has Jul-Aug for 2006, 2007 and 2008 at +0.51, +0.58 and +0.72.
http://data.giss.nasa.gov/gistemp/tabledata/SH.Ts.txt
Crutem3 has them at +0.56, +0.12 and +0.41.
http://www.cru.uea.ac.uk/cru/data/temperature/crutem3sh.txt
Thus it’s obvious that GISS anomalies were going in a different direction to those of Crutem3 from 2006.
If I had one set of shelves in one room with an apple on one of the shelves, another set of shelves in another room with an orange on one of the shelves, then moved the apple up one shelf but the orange down one shelf, would you dispute that they’d moved in opposite directions if you didn’t know which shelf each one had started on?
“By the way, GISS data does not show +.72C. For the globe it is .49C
http://data.giss.nasa.gov/gistemp/graphs/
and I am sure it is less for SH.”
Pat was referring to SH land temperatures
“Also, the satellite data include lower stratosphere and are, by definition, not comparable to surface temps.”
What you say is misleading. See:
http://www.ssmi.com/msu/msu_data_description.html
“…The MSUs are cross-track scanners with measurements of microwave radiance in four channels ranging from 50.3 to 57.95 GHz on the lower shoulder of the Oxygen absorption band. These four channels measure the atmospheric temperature in four thick layers spanning the surface through the stratosphere…
…The brightness temperature for each channel corresponds to an average temperature of the atmosphere averaged over that channel’s weighting function. In the case of channel TMT, most of the signal is from a thick layer in the middle troposphere at altitudes from 4 to 7 km, with smaller contributions from both the surface and the stratosphere. Channel TLT uses a weighted average between the near-limb and nadir views to extrapolate the data to lower altitude, thus removing almost all of the stratospheric influence. For each channel, the brightness temperature can be thought of as the averaged temperature over a thick atmospheric layer….”
I will concede though that I omitted to highlight that the satellite data does not cover central Antarctica, so here I guess I was unintentionally misleading too. (I would emphasise of course that the satellite data has better coverage around the edges of the continent/Antarctic Circle than GISS, and it is these areas which are the best indicators of temperature since it is here that much of the atmospheric mixing takes place.)
Going back to your analogy, when it comes to the satellite data the pears weren’t moving up a shelf either so it’s up to you to explain the divergence.
There is a new paper out that we can expect to see circulated and quoted by the skeptics:
CO2 EMISSIONS BY ECONOMIC ACTIVITIES ARE NOT REALLY RESPONSIBLE FOR THE GLOBAL WARMING: ANOTHER VIEW
International Journal of Transdisciplinary Research Vol. 3, No. 1, 2008
http://www.ijtr.org/Vol3No1/Tsuchida_IJTRPaper_Formatted.pdf
The journal is apparently peer reviewed, but its focus is an interdisciplinary approach to economics, which makes it an odd choice for a climate paper. The author is a retired physics professor.
I read a good bit of it, and it doesn’t impress me as high quality science, but I’m no climatologist and am not really able to rebut it effectively.
So, I’m posting this here in the hope that someone(s) with more knowledge will take up the challenge.
[Response: Nonsense I’m afraid. The attribution of the recent rise of CO2 in the atmosphere to industrial activities and deforestation is incontrovertible through dozens of lines of evidence. Anyone who insists otherwise (that it comes from the ocean – despite the isotopic evidence, budget and direct measurements of increasing ocean carbon) is living in cloud-cuckoo land. Retired professors of physics notwithstanding. – gavin]
Oh, man — he says
“8.5 The nuclear power industry: a tactician behind the CO2 based global warming
In my view, the CO2 based global warming theory was contrived to revive a nuclear power generation industry that suffered from high cost infrastructure and from a bad public image after the disastrous Chernobyl accident in 1986. This contrivance was
effective and appears now to be achieving its aim.”
D’oh.
International Journal of Transdisciplinary Research
And to think I was just getting used to the term : metadisciplinary.
Tsuchida’s paper is an absolute mess. All it does is regurgitate the “temperature leads CO2” argument and then try to pass off the whole CO2 increase as being due to decreased solubility of CO2 in H20 at higher temperature.
His motivation is clear–he is afraid remedying climate change will require more nuclear power. This is a classic sad case of rejecting the science because you don’t like it’s implications.
#106 Pat: thanks for the reference to NOAA NCDC. I just looked it up and it has SH land temps for Sep 08 as +0.44C or 10th warmest, well below Sep 05 at +0.85C. (Of course, it has very limited Antarctic coverage, but then again I don’t know how many more stations GISS takes into account.)
http://www.ncdc.noaa.gov/oa/climate/research/2008/sep/global.html#temp
If I’m picking only one surface dataset, it tends to be from Hadley, but maybe that’s my British bias :)
(I would have quite a lot to say about the CET dataset though.)
“However, I still don’t think anyone should become receptive to arguments for lower-than-consensus climate sensitivities.
Also, I would like to see your explanation, if you have one, for why 1931 was extremely hot and dry in the U.S.”
Not sure why you’re asking me the latter question. The obvious answer (from someone who is indeed receptive to arguments for lower-than-consensus climate sensitivities) is that it was on a par with recent hot years because temperatures at US latitudes of the globe really weren’t as much cooler in the 1930s/1940s (compared to the present) than GISS/Hadley’s best estimates (from often sketchy global coverage) suggest.
http://data.giss.nasa.gov/gistemp/graphs/Fig.D.lrg.gif
However, there are various other plausible explanations, for example:
– changes in US temperatures since the 1930s/1940s show regional variation within the overall warming trend at those latitudes;
– actually I’m struggling to think of any others, apart from inaccuracies in the US temperature record but these have tended to point the other way.
Obviously 1930s/1940s had coincident warm phase PDO and AMO.
BTW I do think there is a need for further examination of the causes of the record Sep 08 temperatures on the Antarctic Peninsula/nearby islands, as they are quite striking. On the face of it, and given that the records were limited in time and space to that month in that region, I would say that unusually volatile jet stream shifts are a likely intermediate cause. The next mildest Sep was in 1984 I think it was, not sure if that gives us much clue as to what caused any jet streams shifts/increased volatility. Obviously, as ever, global warming will be a factor, question is how much and causal relationships.
Benjamin,
I read (very, very quickly) about half of the piece you linked to. So far the author is not discounting the radiative effects of CO2 on temperature, but whether the CO2 we emitted is actually what is causing the rise in concentration. He doesn’t understand the relationship between emmissions and concentrations, but aside from that we know very well that the CO2 rise is due to anthropogenic activities, and we also know the magnitude and rate of the CO2 rise is far outside the bounds of natural variability.
He brings up quite a bit of the “CO2 lags temperature in the Vostok ice core” stuff which has been thorouhgly refuted (at least in the context that this is contradictory to AGW). Rebuttals are not hard to find (if you go to ‘start here’ on the top of the RC page, and scroll to the section that says “Informed, but seeking serious discussion of common contrarian talking points” each one of those links discusses this claim, and similar ones).
It’s pieces like this that David Archer was referring to when he said Denialism is mean for laymen, not for scientists. Just speaking s a somewhat informed laymen, I’ll assure you this guy won’t be getting any nobels.
Barton Paul Levenson (# 104),
Who has determined the official end of “the coming out of the Little Ice Age”? If it was, for whatever reason, indeed around 1850 what mechanism caused the warming between 1900 and 1940?
At least you appear to be one of the rare AGW’ers admitting the importance of solar forcing and who at the same time implicitly refutes Mann’s hockey stick by acknowledging the proxies that established the existence of the Little Ice Age, and for that matter, the Medieval Warm Period.
Chris Says:
17 October 2008 at 8:54 AM
#105 Lauri
“You are comparing apples to oranges to pears. All the different data sets have different base periods for the anomalies. The levels of numbers should be different, therefore. Go back and adjust all numbers according to the anomaly base year. Then there would be at least some sense in the comparison.”
Why don’t you do it yourself?
I am not doing it myself, because common sense says, that if someone proposes something, (s)he is responsible for stating the argument without error. It is not the listener’s task to correct the argument.
Pat was referring to SH land temperatures
So you propose it’s sensible to compare land surface temperatures with satellite observations that span all around the globe. Doesn’t make any sense to me.
However, I take back the comment on stratosphere. I had an erroneous picture in mind.
Chris, I asked about 1931 because I haven’t been able to understand why that year was so warm at and dry at so many climate stations especially in the Midwest. El Nino may explain the warm temperatures which occurred later, in the mid 1930s-1940s, but there was no El Nino in 1931.
Also, I’d like to see someone try to explain the extreme precipitation deficit which occurred during the Dust Bowl years in the Midwest and West. The Mississippi River nearly dried up then and it hasn’t been close to that low since. Dams and land use changes don’t explain the lack of significant precip.
Chris, who decided we HAD an ice age to come out of?
The same people decided when it ended.
Chris Schoneveld (#115)
The forcing from 1900 to mid-century was mostly natural…mostly solar, a bit of lack-of-volcanic, maybe some black carbon in the arctic. The majority of the warming over the 20th century however has been in the last few decades, and the anthropogenic RF relative to 1750 outweighs natural forcings by at least an order of magnitude.
You do not understand Mann’s work: he never denies the existence of a MWP or LIA, just that the late 20th century is anomalous in the context of the last 400 years, probably the last 1,000 and maybe longer.
“Palaeoclimatologists developing region-specific climate reconstructions of past centuries conventionally label their coldest interval as “LIA” and their warmest interval as the “MWP”. Others follow the convention and when a significant climate event is found in the “LIA” or “MWP” time frames, associate their events to the period. Some “MWP” events are thus wet events or cold events rather than strictly warm events, particularly in central Antarctica where climate patterns opposite to the North Atlantic area have been noticed.”
from
http://en.wikipedia.org/wiki/Medieval_Warm_Period
In particular, the MWP and LIA for particular regions are not necessarily precisely synchronous with those in Europe; indeed precise dating for some does not exist. For example, glacial terminal moraines demonstrate local LIA in many parts of the world, both hemispheres. But I, at least, don’t know very precise datings for these maxima. For another, liminological studies in Patagonia demonstrate a (warm) MWP at approximately the same time as in Europe, within a century or so.
Lauri, sorry if I was short with you before. Effect of difference in base years between GISS and Crutem3 appears to be ~0.1C. I just thought the implications of the evidence would be obvious without further detail.
Evidence from satellite data does have some weight when examining possible discrepancies in surface data, I just didn’t feel like spending extra time spelling out the detail, nuances and limitations on this occasion as I thought the person I was replying to would know what I meant by referring to it.
Satellite data is broken down by NH/SH, land/ocean etc see
http://vortex.nsstc.uah.edu/data/msu/t2lt/uahncdc.lt
There is also data available from RSS e.g.
ftp://ftp.ssmi.com/msu/monthly_time_series/rss_monthly_msu_amsu_channel_tlt_anomalies_land_and_ocean_v03_1.txt
Pat, I’m sorry I don’t have the answers to your questions. Though I would comment re: 1931 that regional warm events are not necessarily correlated very well with El Nino conditions e.g. 2006 was the warmest year on record in the UK (HadCET), and 2003 saw a record hot summer in western Europe. Also 1976 remains the hottest ever summer in the UK at 0.4C above 2003.
None of these were major El Nino years.
(Sorry this probably doesn’t help much, as you will doubtless be aware of all this in any event. Maybe someone with more knowledge of US weather than myself can comment?)
Re #108
“Also, the satellite data include lower stratosphere and are, by definition, not comparable to surface temps.”
What you say is misleading. See:
http://www.ssmi.com/msu/msu_data_description.html
“…The MSUs are cross-track scanners with measurements of microwave radiance in four channels ranging from 50.3 to 57.95 GHz on the lower shoulder of the Oxygen absorption band. These four channels measure the atmospheric temperature in four thick layers spanning the surface through the stratosphere…
…The brightness temperature for each channel corresponds to an average temperature of the atmosphere averaged over that channel’s weighting function. In the case of channel TMT, most of the signal is from a thick layer in the middle troposphere at altitudes from 4 to 7 km, with smaller contributions from both the surface and the stratosphere. Channel TLT uses a weighted average between the near-limb and nadir views to extrapolate the data to lower altitude, thus removing almost all of the stratospheric influence. For each channel, the brightness temperature can be thought of as the averaged temperature over a thick atmospheric layer….”
This description is from RSS, as I recall UAH uses a somewhat different approach for TLT which is thought to include more stratospheric signal
I will concede though that I omitted to highlight that the satellite data does not cover central Antarctica, so here I guess I was unintentionally misleading too. (I would emphasise of course that the satellite data has better coverage around the edges of the continent/Antarctic Circle than GISS, and it is these areas which are the best indicators of temperature since it is here that much of the atmospheric mixing takes place.)
The Satellite data doesn’t cover any of the antarctic continent (see:
http://www.remss.com/data/msu/graphics/plots/MSU_AMSU_Channel_TLT_Trend_Map_v03_1.png )
so you’re misleading again.
See also:
http://initforthegold.blogspot.com/2008/10/climate-sensitivity-is-not-central.html
http://initforthegold.blogspot.com/2008/10/followup-re-huge-sensitivity-arguments.html
Phil, so where did e.g.
http://cat.inist.fr/?aModele=afficheN&cpsidt=18948717
get their data from?
[Response: MSU-LT is highly suspect near the poles because of the non-nadir issues with sea ice and perhaps the near-surface temperature structure. MSU 2 and 4 channels are much less affected by surface effects and that is what is used by Johnson and Fu. – gavin]
#121 Chris Says:
17 October 2008 at 5:42 PM
Lauri, sorry if I was short with you before. Effect of difference in base years between GISS and Crutem3 appears to be ~0.1C. I just thought the implications of the evidence would be obvious without further detail.
Chris,
You are correct in that the difference, in practice, between GISS and Crutem3 baselines is small. However, I just computed the difference (in GISS values) of baselines 1951-80 and 1979-07 and it is 0.32C. (I once tried to find out the baseline period for UAH but did not succeed. It seemed to me that the baseline is the whole data set.) This 0.32C is significant and should be used as a correction when comparing GISS to UAH data.
chris schoneveld writes:
The sun increased in brightness slightly, and carbon dioxide was beginning to ramp up.
Nobody disputes that there was a Medieval Warm Period or a Little Ice Age. What’s disputed is whether they were global or not — the former probably was not, the latter probably was. And the MWP was not warmer than the present, as some have tried to say.
More than a dozen independent studies have come up with “hockey sticks” since Mann et al. 1998. The latter was just the first to use that kind of method to establish a historical temperature curve.
Chris Colose #119,
Chris, to demonstrate that Mann acknowledges the LIA and MWP, please indicate on the hockey stick graph well defined intervals that supposedly correspond with the MWP and LIA. To allege that I don’t understand Mann’s work is a bit stiff, thank you very much.
And as to the rate of warming in the late 20th century, it is not much different from the rate of warming in the early 20th century.
BTW, there are too many bloody Chris’s responding to this thread, it’s getting confusing.
Mark #118, thank you for your enlightening response.
Chris, thank you for commenting on my questions. I realize now it’s probably a distraction for me to dwell on the 1930s when we know (as Chris Colose said in #119) that … “the late 20th century is anomalous in the context of the last 400 years, probably the last 1,000 and maybe longer.”
Barton Paul Levenson (126) — Here is a review of a book by anthropologist Brian Fagan which suggests a fairly global extent for MWP:
http://www.nytimes.com/2008/03/21/books/21book.html?ref=science
Re: #127: For what Mann thinks about the Little Ice Age, I would read: http://holocene.meteo.psu.edu/shared/articles/littleiceage.pdf
Also, read Chapter 6 of the IPCC 4th Assessment Report for the overview of what (at least a year or two ago) the literature had to say about the MWP, the LIA, and temperatures in the last 1000 years in general.
I don’t think anyone denies that the sun matters for climate, but the question is whether the variability of the sun in recent history has had the impact that we project from greenhouse gases over the next 100 – and there, I think, a majority of your “AGW’ers” would think the evidence suggests that changes in human forcing will likely be several times (at least) larger than any solar variability we’ve seen in a thousand years or more.
Interesting — look at the top hits in this search in Scholar.
They’re from McIntyre, Michaels, and World Climate Report.
http://scholar.google.com/scholar?q=Mann+%22hockey+stick%22+MWP+LIA
One of the many penalties of ageing is that people insist on explaining, slowly and carefully, ideas with which you have been completely familiar for many decades.
Tamino (29) (Open-Mind, Central England Temperatures) predicted an up-tick (not a least-squares trend) in surface temperatures of 0.5 degrees centigrade this decade, starting from an annual average of about 10.5 degrees. With just 15 months to go, Tamino, sooner rather than later you will have to accept that you picked the wrong decade.
I have commented many times in these posts of the near impossibility of measuring the changes in global temperatures which AGW theory predicts. The variability of all the data sets is obvious. The warming is everywhere slight. To make its presence clear any particular forcing needs to compete with natural cycles and trends, events, as well as measurement errors, both systematic and random.
Random errors (noise in these posts) will accentuate the variability of the data and increase the demands on the trend needed to establish significance. Systematic errors will distort the trends, and are not easy to detect. With satellite data, systematic errors like diurnal drift, altitude change, instrument deterioration etc are particularly trying. Surface (or more accurately near-surface) temperatures are vulnerable to urban development, which I observe regularly driving along the (flat) M4 motorway into London.
Those whose business it is attempt to correct for some of these, and we are presented with a variety of data, always plotted as deviations from a mean which accentuates the false zero effect, exaggerating the changes in relation to the absolute temperature.
Tamino asks why I quote the UAH satellite data and not, say, the GISS “surface” data. The answer is that UAH can cross-check their results against an independent source – the long running radio-sonde data. Have a look at the Hadley Centre Frequently Used HADAT graphics. Hadley (hardly a denialist organisation) have weighted their radio-sonde data to match the satellite altitudes. They plot surface, HadAT2, UAH and RSS data together, as does Tamino in his first “this”, 29.
The plots are similar – little or no warming from 1978 to before the El Nino peak in 1998, a sharp increase from 1999 to 2002, and varying degrees of fall-back from about 2005/2006 led, surprisingly, by the radio-sonde data.
As well as the trends themselves we can test the fundamental AGW assumption that temperature increases must be uniform across the lapse rate temperature distribution – the surface, lower and upper troposphere temperatures must increase together. Over 30 years from 1978 (the trough following the 1945 peak) the trend increases per decade are:
Surface: 0.16 degrees
UAH : 0.12 degrees
RSS : 0.18 degrees
Radio-Sonde : 0.16 degrees
The UAH mid-troposphere trend is 0.05 degrees per decade.
All these trends depend on the step change from 1999 to 2002, and the persistence of these (marginally higher) temperatures over the next 3 or 4 years to 2005/2006. If the current 1978 temperatures continue, the upward trends, such as they are, will fall away rapidly.
No one can know what will happen over the next decade, but this data does not support the IPCC assertion that we can be 90% certain that increasing CO2 concentrations have been responsible for a substantial part of the 20th century warming, or that we an expect 3 degrees C of warming over the next century.
[Response: I find that the penalty of running this blog is the amount of aging I end up doing patiently explaining to people the same facts over and over and having them not listen each and every time. Your confidence in the logic and physical relevance of your thinking is unfortunately highly misplaced. – gavin]
19: Figen Mekik Says:
“To avoid all that, I think peer review should be double blind so the work is judged on its own merits”
But, AFTER the paper has been accepted, the name of the reviewers AND [even more importantly] the text of the review should be published along with the paper [in the electronic version, at least]
133: [Response: I find that the penalty of running this blog is the amount of aging I end up doing patiently explaining to people the same facts over and over and having them not listen each and every time. […] – gavin]
Gavin, how true !
133 Gavin, pontification is an export issue. Importation of data is NOT desired or allowed. Fred decided long ago what is ABSOLUTE FACT, and so current data is surpurfulous.
You waste your time trying to educate the uneducatable Fred, but there are thousands of others whose brains aren’t stuck in the muck of past beliefs. Fred’s spouts are far removed from reality. The graphs of temperature ALL show that current temps are far higher than in the 70s, but Fred still spouts goop. “current 1978 temperatures?!” Egad, Fred, you’re embarrassing yourself. Fred, please, please, PLEASE go look at some recent data. ANY data. Don’t do it on a clean carpet, though.
I’m just here to remind everybody to watch the upcoming PBS Frontline Special “Heat — A Global Investigation” which will be shown both on the air and online at 9 p.m. EST, October 21st.
http://www.pbs.org/wgbh/pages/frontline/heat/
Needed — a link the Contribs can drop into such Inline Responses that automagically searches out all prior inline responses to that user.
(That would mostly benefit the new reader coming along who will only see the “over and over” but not find where it had been explained to the same person over and over. Heck, even Fred might benefit. And it’d save retyping, perhaps.)
(It’s a truism from Usenet days that one way to get good information on the Net is to post what you think and await correction. Without good search tools some folks won’t recall or retrieve the corrections, but rather go on posting the same thing heedless of help.)
OT
Some of you may remember Alexi, who has been here several times, says enough to show that he has a grasp of the science, then delves into opinions not supported by the science.
Well, his favorite sport these days is to hang around Yahoo CCD and pick on people with less education. We had the following exchange recently and if some science expert would like to make sport of replying to Alexi’s assertions, I’d be grateful, since I’m not likely to ever know as much science as he does, and he loves to use that as a bludgeon, as if I must be wrong since I don’t even understand his answer.
I wrote: “CO2 has been shown in laboratory conditions to increase radiative forcing at the surface.”
Alexi replied: “It does not look as you have grasped the concept at all, since the above statement contains at least THREE factual errors.
“First, no controlled laboratory experiments have been conducted at the scale of 70,000 meters long, with varying air density and a special temperature profile.
“Second, the AGW theory alleges that the “forcing” occurs at the top of atmosphere, not at the surface. The actual amount of “forcing at the surface” is a subject of heated debates because many non-trivial factors are acting at the surface.
“Third, the “forcing” cannot be possibly ever observed in laboratory conditions because the definition of forcing contains three physically unrealizable conditions: (a) CO2 mixes up, but temperature profile stays the same, (b) stratospheric cooling equilibrates with tropopause conditions before troposphere does, and (c) troposphere-surface interface is kept frozen. None of these conditions can possibly exist, and therefore the whole concept of “forcing”, expecially the claim that is is accurately calculated to be 3.7 W/m2, is highly dubious.
“It is really amazing how it is possible to make three brutal mistakes in a statement of 10 words long.”
Thanks in advance for any contributions.
– Walt
#122 Phil “The Satellite data doesn’t cover any of the antarctic continent”
Apart from unquestionably the majority of the outer coastal zones (i.e. from 70S)
http://maps.grida.no/go/graphic/antarctica-topographic-map
unlike GISS which only extrapolates from a few scattered stations (and a main concentration on the Peninsula). Not criticising GISS by the way, that’s just the reality of Antarctic surface coverage.
Also (re: Gavin’s response) I note that both TLT and TMT use MSU 2 channel (see below). Perhaps the UAH guys use the same method as Johnson and Fu when extending their anomaly maps further over Antarctica?
TLT = Temperature Lower Troposphere
MSU 2 and AMSU 5
TMT = Temperature Middle Troposphere
MSU 2 and AMSU 5
TTS = Temperature Troposphere / Stratosphere
MSU 3 and AMSU 7
TLS = Temperature Lower Stratosphere
MSU 4 and AMSU 9
from http://www.ssmi.com/msu/msu_data_description.html#channels
Another point, I thought it was more high-altitude continental ice that was the problem than sea ice. RSS set their north pole band up to 82.5N, and this includes large areas covered by sea ice. Perhaps they simply want a single latitude limit, therefore they stop at 70S to avoid including a significant amount of the high-altitude ice sheet, even though there are large areas south of 70S which generate reliable enough data (as with between 70N and 82.5N).
Sorry that this is all rather speculative, and I realise that this should probably be for the UAH guys to make clear, rather for Gavin to have to comment on.
In any event, the only reason I referred to the satellite data in my earlier post was as a (limited) check against the GISS SH land data, in particular its record temperature anomaly for Sep. Since then my attention has been drawn to the NOAA NCDC data, which only show SH land as 12th warmest
http://www.ncdc.noaa.gov/oa/climate/research/2008/sep/global.html#temp
I wouldn’t therefore say that GISS is “wrong”, just that I had a valid point in raising questions about the definitiveness of the record emphasised by an earlier poster.
And sorry if none of this has anything to do with the tropical troposphere, can’t even remember how this diversion started now!
[Response: MSU-LT uses off-nadir radiances (unlike MSU 2 or MSU 4) and they are the problematic ones. – gavin]
There is no “near impossibility of measuring the changes in global temperatures”. Furthermore, it’s easy to see that regional warming is taking place at many climate stations in the Upper Midwest and Alaska, which is being driven by global warming.
http://picasaweb.google.com/npatnew/ClimateDataMidwestAK#
Walt, what is the link?
The graph above may demonstrate the “you can’t prove our models are wrong” thesis, but they do little to convince a skeptic like myself the models are right. Look at the (A) section: all the sonde datasets are left of the grey confidence zone of the models for a big chunk of the lower troposphere. Look at the (B) section: all the sonde data means plus UAH MSU means are below the model means. And I think it would be worse if the data were run out to 2008.
Correct me if I’m wrong but this paper does little more than regurgitate the Douglass results, but with a wider “confidence zone” for the models. If your aim is to to prove that current GHG modeling of the atmosphere is sound, that’s not near good enough.
Regards, BRK
[Response: Well, since that wasn’t the aim, we are doing ok. What is needed to show that models have skill are comparisons to data that is a) well characterised, and b) with a large signal to noise ratio. Both elements conspicuously absent in the tropical tropospheric trends, but abundant in the 20th Century global mean surface anomaly, or in the stratospheric cooling, or the Arctic amplification or the response to Pinatubo, or the increase in water vapour, or the enhancement of the Antarctic polar vortex etc. Please see the IPCC report for more detailed assessment of these issues. – gavin]
Re: #142
Ray,
http://tech.groups.yahoo.com/group/climatechangedebate/message/22190
This is Yahoo CCD. You have to be a member to jump in, but membership is automatic.
Gavin’s response to #143
“What is needed to show that models have skill are comparisons to data that is a) well characterised, and b) with a large signal to noise ratio.”
Maybe I misunderstand, but wouldn’t that imply showing that models have skill is depending on a clear signal existing in the first place? My understanding of most of the (lets call it) skeptical positions from people like Roy Spencer is that they essentially claim exactly that: the absence of a large signal compared to noise (or natural variability) and the entire debate is essentially about the question, whether noise is a measurement/statistical problem or the very nature of climate itself? Or is the noise in this case something that can be clearly separated from natural variability?
[Response: This is a real issue, but for attribution purposes the signal vs. noise has to be determined using the models. Within the models the late 20th C trend is way outside of the ‘noise’, while the 1979-1999 tropical tropospheric trend is not. Part of the uncertainty in the attribution is of course how realistic the ‘noise’ in the models is – and that can be assessed by looking at hindcasts, paleo-climate etc. Roy Spencer notwithstanding, there is no evidence that the models hugely underestimate the background variability. – gavin]
Walt #139
1- why must it be on the same scale? It wouks on 1m scale. why does it change at larger scale? At what scale do you say it changes? Why?
2- Incorrect. Prove it says that.
3- CO2 mixes so on the scale of km it is homogonous. b. says nothing. c is ridiculous. frozen means solid. there’s no solid surface in the sky.
note: Won’t change him
Walt —
Alexi’s analysis is completely wrong. To begin with, you don’t need to have a 70,000-foot tube to measure CO2 illuminance depletion. All you need is enough different lengths to verify that the Beer-Lambert-Bouguer Law holds. We can also measure the natural solar and back-radiation from the sky.
The statement that “CO2 mixes up, but temperature stays the same” is contextless and scientifically illiterate. What temperature, under what circumstances? The phrase “stratospheric cooling equilibrates with tropopause conditions before troposphere does” is also meaningless. Alexi is just stringing sciency-sounding phrases and words together in an attempt to intimidate readers who don’t know those terms. How can you compare “stratospheric cooling” with “troposphere?” How does cooling in the stratosphere “equilibrate” with conditions at the tropopause? The phrase is just meaningless. So is “troposphere-surface interface is kept frozen.” All three “prerequisites” exist solely in Alexi’s mind.
Walt #139, the guy is browbeating you.
He doesn’t have any deep understanding of the science, just read enough of it to spell the difficult words right.
I spotted one error offhand: he says
“Second, the AGW theory alleges that the “forcing” occurs at the top of atmosphere, not at the surface. …”
i.e., the same mistake you made, only a few kilometres further up :-)
(For clarity, the “forcing” is a concept that applies to the atmosphere as a whole, not to any particular level. What happens at the “top of atmosphere” — the level where outgoing radiation leaves for space, not itself a very easy concept — is the restoration of equilibrium, the increase in temperature that, through Helmholtz-Boltzmann at the Earth’s brightness temperature 255K, restores the balance between incoming and outgoing energies. This then propagates downward along the temperature gradient, which may also change a little as temperature goes up. Contrary to what Alexi claims, there are no great uncertainties there. Not as long as we talk about the CO2 forcing sec, before the feedbacks.)
[Response: It’s probably also worth noting that the standard radiative forcing is a diagnostic of what is going on designed so that different forcing mechanisms can be usefully compared. The physics of any particular forcing happens wherever it’s appropriate – ie. every grid box has a change in radiation if there is a change in GHG or aerosol composition; ozone changes affect the stratosphere and troposphere directly; land-use changes affect the surface reflectivity and water cycling etc. – gavin]
Gavin:
A few comments on your comments in your statement above:
Arctic amplification – What about the Arctic warming of the 30’s? I don’t know the details but I doubt that the models do a good job of replicating this. So if the models “don’t get” non GHG Arctic warming, why should I believe they “get” GHG-driven polar amplification? You could argue the data on 30’s arctic warming is too sparse both spatially and temporally to be a realistic test of model performance, but then you start to weaken your point about “well characterized” global surface temperature. As a final comment on polar amplification, I don’t think you could argue the south pole temperature trends offer much support for model performance.
As for stratospheric cooling, the trend since 1996 looks flat. Sure it’s only 12 years but as pointed out above, it’s got a low noise to signal ratio. The question is regarding stratospheric temperatures: What have you done for me lately? (model-performance wise)
Which brings me to my last point. If you’re going to convince skeptics like myself the models are sound, you’re going to have to get them past the 20th century. You had a posting some months back on an ocean heat content paper which ended in 2002 (I think), leaving the last 5 years blank. And now this paper ends the analysis at 1999, leaving the last 8 years blank. And yet the best data we have for many climate parameters has been gathered over the last 10 years or so.
[Response: Your comments indicate to me that there is not much point in trying to convince you of anything. All of those features I mentioned were predicted decades before they were observed, and yet you seem to think that just one or two more years will make a difference to your thinking. Sorry, I’m not convinced. What I read instead is a grasping of straws in order to provide a fig leaf of credibility when not paying attention to what’s robust. However, you should check on the details occasionally. You appear to think that the MSU-4 (lower stratosphere) record is what I was talking about – it’s not. That record is mainly driven by ozone depletion (which is slowing due to restrictions on CFC emissions). Cooling on the upper stratosphere and mesosphere is the signal of GHG-related cooling and that is ongoing (look up Stratospheric Sounding Units (SSU)). I think hardly anyone would agree with you that Antarctic trends are ‘well characterised’, but the general fact that the southern hemisphere is expected to warm less rapidly than the north is itself predicted by the models.
As for dates for model runs, these depend on the availability of emission data sets (which take time to collate) and the cutoff dates for submitting data to the central databases. The runs for IPCC AR4 were completed in 2004, with emission datasets that went mainly only up to 2000 and so many of the simulations only went to the end of 1999. However, note that we were assessing the work done in Douglass et al, and I don’t recall you complaining in those earlier discussions that they should have used a longer period. Curiously, in their figure they used 1979-1999 for the models and 1979-2004 for the observations. It turns out not to make much difference, but we preferred to compare like with like. The latest model runs will start next year and they will be up-to-date to at least 2005. – gavin]
re 125
Baseline of satellite records (incl. UAH) usually is 1979-1998 (20 years). They might change in the future when a 30 year baseline (1979-2008) is available.