Bishop Hill Large-scale temperature trends
Jun 5, 2012 This is a guest post by Rob Wilson. It addresses some concerns I raised when I spoke in St Andrews the other week. I had discussed Climategate and the lack of trust this engendered and then went on to briefly cover other issues that made me uncomfortable. In particular, I mentioned the tendency of corrections to the temperature records to produce cooling in the nineteenth century and warming in the twentieth, and the recent lack of warming.
In light of Myles’ and other comments w.r.t. instrumental data, I thought this might be a good time to quickly try and address some of Andrew’s observations that he made in his talk at St Andrews in April.
I had hoped to write a guest post along with colleagues from the Met Office showing temperature trends along with AR4 projections, but I already see the summer running away and what spare time I have, I would rather concentrate on a series of later guest posts focussing on dendroclimatology. So below, I concentrate only on temperature trends in the HADCRU and HADSST data-sets. Thanks to Ed Hawkins and John Kennedy for providing feedback.
There were two issues that Andrew raised:
- That updates of large scale temperature data-sets appear to depress 19th century and raise 21st century temperature values.
- That over the last decade or so, there had been a flattening off in temperature trends.
At the time, I could not comment on either as I had not looked at the new data-sets in detail.
So – the following link will take you to a series of figures that compares CRUTEM3 and 4 (land temperatures) and HADSST2 and 3 (SST) for northern and southern extra-tropics (ET) and tropical (TROP) latitudinal bands. References at end.
http://www.st-andrews.ac.uk/~rjsw/ftp/TempTrends.htm
For those of you who want to check and replicate my plots, the data can be easily accessed from the Met office website:
http://www.metoffice.gov.uk/hadobs/hadcrut3/
http://www.metoffice.gov.uk/hadobs/hadcrut4/
http://www.metoffice.gov.uk/hadobs/hadsst2/
http://www.metoffice.gov.uk/hadobs/hadsst3/
or perhaps an even more user friendly site is:
I have purposely not added trend lines, or smoothing functions and have just plotted the temperature anomalies (w.r.t. 1961-1990).
I am not going to describe trends in exhaustive detail, but really want to address Andrew’s two main concerns.
Older vs newer temperature data-sets
There has been little change in the NH ET input data-sets. The major changes I am aware of are some early instrumental corrections of temperature data-sets in the Greater Alpine Region. However, this is only a small number of records in the extensive NH data-set so does not impact the large scale mean series. For those interested, the Alpine data-set correction is detailed here:
Böhm R, Jones PD, Hiebl J, Frank D, Brunetti M, Maugeri M (2010) The early instrumental warm-bias: a solution for long central European temperature series 1760 – 2007. Climatic Change 101, 41-67.
For TROP and southern hemisphere ET land temperatures, the major changes are in the 19th century which reflects the addition of newly digitized station records – probably mainly from Australia. Early instrumental temperatures are always going to be less certain and there is less data.
Changes in the late 20th century appear to be minimal.
w.r.t. SST, again little difference between HADSST2 and 3 in the ET NH.
The period of greatest difference in the TROP SST data is around the post 1940’s period which are related to biases in HADSST2 w.r.t. an “uncorrected change from engine room intake measurements
(US ships) to uninsulated bucket measurements (UK ships) at the end of the Second World War.” These have been adjusted in HADSST3. For those interested, the relevant paper is:
Thompson et al. (2008) A large discontinuity in the mid-twentieth century in observed global-mean surface temperature. Nature 453, 646-649
http://www.roberts.cmc.edu/159/2009%20Combined%20pdfs/Feb%2010%202009.pdf
For the southern hemisphere ET SSTs, we again see similar corrections in the 1940s as in the tropical SSTs, but interestingly, the HADSST3 are actually marginally “colder” in the recent period than HADSST2. I only highlight this to show that correction can go both ways.
As a final note, correction for homogeneity biases in temperature record is very important and if you want more information on the basic theory, a really good review paper is:
Peterson,T.C. et al, (1998). “Homogeneity adjustments of in situ atmospheric climate data: a review.” International Journal of Climatology, 18 1493-1517
http://www.st-andrews.ac.uk/~rjsw/PalaeoPDFs/Peterson-etal-1998.pdf
The recent flattening of temperature trends
As for the recent flattening. Well this appears to vary markedly. For NH ET winter temperatures, there is clearly an “eye-ball” flattening in winter temperatures, but likewise, a continued increase in summer temperatures. Tropical land temperatures appear to show continued warming for all seasons, but tropical SST records could be argued to have flattened. SH ET land temperatures is a little mixed – perhaps a flattening in summer, but still increasing in spring and autumn.
Statistically, due to internally forced multi-decadal variability expressed in all of these records and the fact that we are “at the end of the time-series”, I think it is really very difficult to “quantify” a flattening or even a continued increase. Yes, we can fit linear trend lines to the latter end of the time-series, but with the known naturally forced decadal variability expressed in these data-series, I personally think that such exercises are really not very helpful. This issue will simply become clearer over the next 10-20 years............but should we wait until we have statistical certainty?
Final thought
So – my take home message. Let’s not generalise too much. The newer HADCRU4 and HADSST3 data-series are incrementally improved data-sets using new data and utilising corrections related to robust theory and methods. Many on this blog will disagree with this statement, but all I can urge is please read the papers below. Much effort is focussed on the uncertainties and biases in these records. I do not see a systematic change (between old and new) to cooling (warming) of early (late) large scale instrumental series – rather I see improved data-sets (HADCRU4 and HADSST3) with well documented uncertainties.
As for temperature trends, in the same way that it does not really matter if the medieval period was warmer or cooler than today, it does not really matter if a particular seasonal time series shows an increase or flattening in temperatures. What MATTERS is that we need to understand the drivers of these changes. Natural or anthropogenic (or a mix of both). CO2 cannot explain all trends since the 1850s, but likewise internal dynamics (PDO, ENSO, NAO etc) or changes in the sun or large-scale volcanic events cannot alone explain the variability in climate.
References:
Jones, P. D., D. H. Lister, T. J. Osborn, C. Harpham, M. Salmon, and C. P. Morice (2012), Hemispheric and large-scale land surface air temperature variations: An extensive revision and an update to 2010,
J. Geophys. Res., 117, D05127, doi:10.1029/2011JD017139.
http://www.metoffice.gov.uk/hadobs/crutem4/CRUTEM4_accepted.pdf
Kennedy J.J., Rayner, N.A., Smith, R.O., Saunby, M. and Parker, D.E. (2011b). Reassessing biases and other uncertainties in sea-surface temperature observations since 1850 part 1: measurement and sampling errors. J. Geophys. Res., 116, D14103, doi:10.1029/2010JD015218
http://www.metoffice.gov.uk/hadobs/hadsst3/part_1_figinline.pdf
Kennedy J.J., Rayner, N.A., Smith, R.O., Saunby, M. and Parker, D.E. (2011c). Reassessing biases and other uncertainties in sea-surface temperature observations since 1850 part 2: biases and homogenisation. J. Geophys. Res., 116, D14104, doi:10.1029/2010JD015220
http://www.metoffice.gov.uk/hadobs/hadsst3/part_2_figinline.pdf
Morice, C. P., J. J. Kennedy, N. A. Rayner, and P. D. Jones (2012), Quantifying uncertainties in global and regional temperature change using an ensemble of observational estimates: The HadCRUT4 dataset, J. Geophys. Res., doi:10.1029/2011JD017187, in press.
http://www.metoffice.gov.uk/hadobs/hadcrut4/HadCRUT4_accepted.pdf
Reader Comments (190)
Bish!
He's back!
Except ...
While agreeing that known errors in instruments need to be corrected (as the cave troll pointed out yesterday) it would be nice to be able to trust the scientists involved to do that job honestly. Then we could all say with some authority behind us that subsequent adjustments had no validity. As you say, the data are the data. Live with it.
But wasn't there a case last year reported (if I recall) by Anthony Watts where some blatantly incorrect — as in virtually impossible — high readings were pointed out but ignored and left in the data.
There appears to be a constant undercurrent of low-level "misbehaviour" like this with each incident perhaps being insignificant in itself but in accumulation undermining the credibility of the data and therefore the conclusions drawn from it.
It seems that no-one in the climate science community is prepared to address this even by the simple expedient of demanding that adjustments to data require to be checked and signed off independently if only to remove the temptation to "ease to fit"!
Rob, can I add my voice to the others on this thread who are exhorting you to publish the insulting emails you've received. They should be outed and admonished by those of us who try to keep on this side of politeness.
You could learn from us how a community is supposed to act when it's members are acting in a way that brings the community into disrepute. Then maybe you could visit the Climategate emails.
Ah, the golden age when scientists were just trying to do their best with their data
- see here for 'A glossary for research reports' from 1957...
http://www.astro.rug.nl/~terlouw/glossary.html
Example: Typical results are shown ........(= the best results are shown).
:-)
Ross McKitrick, Christopher Essex, and Bjarne Anderesen have written a paper “Does Global Temperature Exist?” which argues on the basis of mathematics, statistics and physics there can be no global temperature as it is possible to define any number of global temperatures.
http://www.uoguelph.ca/~rmckitri/research/globaltemp/GlobTemp.JNET.pdf
Clive Best has writing at WUWT has argued a average temperatures make no sense
as there may be dry air over some weather stations and damp air containing ice particles over others
http://wattsupwiththat.com/2012/02/07/how-reliable-are-global-temperature-anomalies/
I have posted how a global temperature makes no sense on as the latent heat of ice is so much greater than specific heat of air
http://jeremyshiers.com/blog/global-temperature-is-it-rising-or-falling-does-it-even-exist/
If global average temperatues make no sense, then a collection of historic global average temperatures make no sense either.
In his book Chill Peter Taylor narrows the CO2 global warming down to one sentence
IPCC claim that CO2 was responsible for 0.8 watts/square meter warming from 1990 to 2000, yet over the same period increases in solar output and decreases in cloud cover resulted in an increase of at least 6 watts/square meter.
Cloud cover has recently increased, the suns output has decreased and temperatures are down.
All of which indicates that warming due to greenhouse gases, if it occurs at all, is not a major driver of climate and probably does not even have a measurable effect.
Oh and just to nit pick the temperature datasets are not of temperature but of temperature anomaly
and a very brief search will find evidence of these temperature anomaly datasets being "adjusted" to increase warming eg
http://stevengoddard.wordpress.com/2012/05/11/hansen-covering-his-tracks-in-iceland/
A bit marginal, but does anybody has a link to the complete methodology for computing the global temperature trend at CRUTEM and GISS?
I haven't managed to find which lapse rate they use between the grid mean elevation and the actual elevation of stations within that grid.
Thanks
It does not really matter if the medieval period was warmer or cooler than today.
If it was warmer than today CO2 cannot be the main driver of todays slight warming, if it was cooler artifacts would not still be being dug out of the permafrost in greenland.
http://www.archaeology.org/online/features/greenland/
Robuk: here's another study showing the MWP also existed in Chile: http://notrickszone.com/2012/06/06/medieval-warm-period-shows-up-in-south-america-too-far-far-away-from-north-atlantic/
Y en Patagonia:
http://www.clim-past.net/8/403/2012/cp-8-403-2012.html
Jun 5, 2012 at 9:41 PM | drcrinum
Said it much better than I.
Troll posts and follow up removed. Please DNFTT.
When it takes careful aim and shoots itself in both feet at once, it's hard to resist!
Likewise I didn't get up this morning and say "it's not very anomalous today, dear; shall we stay indoors and do a jigsaw puzzle?"
These "anomalies" — or as I prefer to call them "variations" — must be based on something and that something can only be observed or calculated temperatures. And if the idea of a global average temperature is meaningless and the base is supposedly the averaged global average for an arbitrary 30-year period then the whole thing is patently either a house of cards or a bucket of horse droppings depending on how your mind works.
It's not simply that the emperor has no clothes; the emperor himself is a figment of the imagination.
Jun 6, 2012 at 11:12 AM geronimo
Agreed.
Accordingly, I have submitted an FOI for all emails received by Dr Wilson as a result of his posting on Large-scale temperature trends that would be embarrassing to BH commenters or which Dr Wilson considered insulting.
Instead of concentrating on the records used, I think the the records not used are more telling.
If anyone can point me to where the Geraldton town & Kalgoorlie P.O. records are used in GHCN or HADCRUT x (just to highlight a couple) I will be most grateful.
IMHO: "I Overreacted"
Mike Jackson: you would think wouldn't you that a data set with the world's finest minds poring over it for half a century or more would by now be clean? I'd wager they would have adjusted the temperatures up if they'd have been found to be anomonously low.
Rob Wilson
Thank you for your response. I see Barry Woods has made the point that the statement from the UK science community was clearly a response to the Climategate emails. On one level I can see that it might be regarded as commendable that you wanted to be seen to be standing ‘shoulder to shoulder’ (to borrow a Blairism) with your scientific colleagues in the UEA bunker. On the other hand, the leaked emails show those same colleagues engaging in non-scientific practices. So the way I see it is that you and dozens of others who should have known better were prepared to condone malfeasance - and the point, if you do not get it, is that this sort of behaviour goes to credibility. You stand by your act. Your credibility, at least in my eyes, remains in minus territory notwithstanding your willingness to post here. Sorry, but that is how it is in the real world.
On the temperature record others have made the point that it is not ‘scientific’. As I understand it, the earth receives a wagonload of electromagnetic radiation from the big burny thing in the sky and when the radiation hits stuff it turns into ‘heat’ which goes off and hides in the sea and swirls around in the atmosphere and then finds its way back into space and we humans have thermometers dotted about the place (no suggestion that the thermometers are placed in any sort of representative way?) which were formerly used to tell us what the temperature was (to within the sort of tolerances that would indicate whether another sweater might be required) and those records of the temperature of the air are ‘averaged’ notwithstanding that temperature is an intensive quality and when I went to school we weren’t supposed to average intensives and then what about enthalpy I hear you saying because well there is more of that ‘heat’ stuff we were talking about in humid air than dry air and it might follow that temperatures in the tropics are apples and temperatures at the poles are oranges and not to mention that the mean is calculated by taking the daily high and the daily low and ending up in the middle and how does that tell you if it has been mostly warm or mostly cold look at May 2012 in the UK bloody freezing for the first 3 weeks hot as a pork pie for the last 10 days and overall average I mean to say the whole thing is a bag of mashings. If you see what I mean?
"Accordingly, I have submitted an FOI for all emails received by Dr Wilson as a result of his posting on Large-scale temperature trends that would be embarrassing to BH commenters or which Dr Wilson considered insulting."
You have got to be kidding!!
I think this is enough from me on this thread.
Dr.Wilon, a lot of fact based criticism has been levied on your post on the science part and t=on the insulting e-mails part. Instead of facing up to them and giving straight answers, you are picking on mail to take offence and are saying " that's it " and attempting to use that as an excuse to run away, not answering the valid questions. We've seen too much of such behaviour from the " climate science " community and so pardon my cynicism with such an attitude. It doesn't wash.
To be blunt are you honest and man enough to answer the questions and criticisms related to your post and also show up the hateful e-mails you alleged which came from this blog's denizens?
Well, yes, geronimo. But you are assuming "the world's finest minds".The best I could ever manage was human nature+ benefit of the doubt.
I was never a fan of temperature adjustments but when I realised that they were finding reasons for them 50 years after the event benefit of the doubt flew out of the window.
The same applies to arctic interpolations or extrapolations or whatever.
If you don't have data, you don't have data. Stop pretending. And above all stop making up data that just happenes to reinforce your personal world view. Are you listening, Dr Hansen? Right answers for the wrong reasons are bad enough; wrong answers for the "right" reasons are on a different scale altogether.
What's the matter Robby. taking your ball and going home?
It seems likely that Rob Wilson's latest post will be the subject for "discussion" for a while but for anyone who's still interested in the science stuff...
Some people seem to have some problems understanding the idea of an average temperature and/or a global average temperature. The problems seem to be at various levels.
1: No such thing as average temperature as temp is an intensive variable
Two baths, each with 1 m^3 of water; uniform temps, one at 10, other at 30 degrees. Average temp is 20 degrees. Average can obviously be weighted for different volumes. Averaging intensive variables is well-defined (but they can't be meaningfully added)
2: Average temp isn't "physically meaningful"
Pour water from the above mentioned two baths into a single previously empty bath (of volume >2m^3 to counter any pedants). Temp of well-mixed water is 20 degrees. So calculated average temp is temp resulting from mixing separate volumes into a single volume and hence "meaningful".
3: There are lots of global average temps
True but not a problem. There are an infinite number of ways in which any set of measurements can be weighted to produce an average. That doesn't prevent calculation of averages. Choose a physically meaningful method, apply it consistently and then you can look at changes over time. For AGW, the aim is to assess changes in the heat content of the Earth's atmosphere. Temp measures mean kinetic energy so the appropriate average is the (weighted) mean of individual measurements. One can of course argue about the accuracy of this average but it's meaningful, well-defined and its variations over time can be assessed.
4: Weather station coverage is too variable for an average to be meaningful
Interpolate, approximate, assess variations and calculate error on average.
5: Temperatures vary depending on time of day, height, nearby air-conditioning units, ground cover...
These are all difficulties for ground station measurements. Not insurmountable but tedious and all adding to errors.
So average global temp can be (and has been) calculated from ground station measurements. However, currently the best method of measuring global temps (and hence averages) is via satellite. Such measurements have been made for about 20 years. UAH's website gives globally averaged temps in degrees K to 3 decimal places.
The claims of AGW have plenty of problems but the notion of global average atmospheric temperature (and anomaly) isn't (I think) one of them.
I think I addressed anything I am capable of addressing w.r.t. instrumental data in my comment last night and originally only wanted to address Andrew's concerns. I have done that.
Bye for now, but I am a glutton for punishment and will probably be back to pontificate about tree-rings at some point.
I am sure that will get some of you going.
For now, feel free to read any of my papers:
http://www.st-andrews.ac.uk/~rjsw/cv/publications.html
signing off
Simon Anthony
1: No such thing as average temperature as temp is an intensive variable
2: Average temp isn't "physically meaningful"
3: There are lots of global average temps
Simon you have given an example with 2 baths of known size
In real life we don't know the size of the baths, what substance is in them
A better analogy would be to try and estimate the heat content of an unkown number of baths
each filled with an unknown amount of unknown stuff, with unknown specific or latent heats
and to make this estimate from a thermometer measuring the temperature of the air some way
from these baths.
4: Weather station coverage is too variable for an average to be meaningful
5: Temperatures vary depending on time of day, height, nearby air-conditioning units, ground cover...
There comes a point (for example when you've interpolated your way across australia) the errors in your calculations are so large as to make your results meaningless.
The meaninglessness is increased when the datasets are subject to periodic review.
The temperature anomaly datasets are being used to justify the claim the heat content of the earth has increased. I belief the datasets are so flaky they can not be used to justify this claim.
"Accordingly, I have submitted an FOI for all emails received by Dr Wilson as a result of his posting on Large-scale temperature trends that would be embarrassing to BH commenters or which Dr Wilson considered insulting. Jun 6, 2012 at 2:16 PM | Martin A"
Oh dear, how deeply undiplomatic of you to have announced that here and now. It might fit your current frame of mind but it does not help the thread when Dr Wilson leaves it.
Jun 6, 2012 at 5:10 PM | Jeremy Shiers
Your analogy is false.
It's simplest to see with satellite measurements (which for some reason you don't mention): temps are (almost) directly measured; volumes are known; composition of atmosphere is well-enough known.
So rather than discuss the details of surface station temp measurements (and, as I said, one can argue about the size of errors and I share at least some of your reservations), what problems (if any) do you have with satellite measurements?
And if you don't have significant objections to the latter, then since they agree fairly well with surface station measurements, the latter can't be very wrong.
Why do you need the average temp of two baths when there is no water at that temp, and you can't get in two baths at once? And of course, for the same average both baths might be temperate or both lethal. It's meaningless because you can't do anything with it. The average anomaly likewise.
The thing to do with temps is take every source, every station, every data point and plot against time. Compare highs and lows, compare daily average, compare year for year or month for month. All those plots and their metadata contain buckets of information which are gone at the global anomaly level. That is the information which is of interest. The man behind the curtain wants you to look at the wonderful confection he has made of all those ingredients. Don't fall for it. The truth is in station data, unadulterated. And in the way the stations he chooses vary over the years. No averaging, no smooshing, no interpolation.
Jun 6, 2012 at 5:33 PM Simon Anthony
I don't believe the analogy is false as I don't believe the composition of the atmosphere is that well known simply because the amount of water in it changes
Moreover the state of the water in the atmosphere changes (ice, water, gas).
Temperatures may be measured directly but they are converted to temperature anomalies.
And we've been talking about temperature anomalies (and adjusted ones at that) where the entire change is compressed in to about 0.5C so pretty much anything will agree with anything.
Temperature is being used to estimate the heat content of the earth which doesnt seem practicable.
I have no particular knowledge of satellite temperature measurements. Could you provide a reference and some sort of justifcation for your claim that these measurements agree with surface measurements. It seems unlikely that there will be a satellite measuring the temperature of Lerwick weather station, for example.
Was there any specific reason why you thought I ought to have mentioned satellite temperature measurements.
As far as I know sea surface temperature measurements havent been mentioned either.
I gave examples earlier of temperature datasets having been adjusted
http://stevengoddard.wordpress.com/2012/05/11/hansen-covering-his-tracks-in-iceland/
do the satellites agree with pre or post adjustment land temps?
Jun 6, 2012 at 5:19 PM Cassio
Dr Wilson has a record of walking away when he does not like the comments on BH - see the previous thread he was involved in.
What did he expect?
He has said that BH posters have sent him insulting emails. It's not a trivial accusation.
It fits a pattern, as someone pointed out here recently, that fits climate scientists who debate with CAGW sceptics.:
- Assume a position of superiority, either moral or intellectual.
- Display the assumption of a right to control the direction tone of the debate, with a readiness to flounce off if the discussion does not go in the desired direction.
There seems to be a "sceptics send nasty emails" pattern too. It's reminiscent of the Australian "death threats" to climate scientists - which turned out to be nothing of the sort after FOI requests.
It also fits the "Poor Phil" meme - who was said by UEA to have received death threats - of which nothing more whatever has been heard, despite such threats normally being taken extremely seriously by the police.
Jun 6, 2012 at 5:52 PM | Rhoda
None of your objections seems to apply to satellite measurements of global average temp (which do all the necessary averaging in time and space).
Do you have any objections to such satellite measurements?
Rob talks about what he wishes too, and then signs off in a manners that I'm afraid comes across as patronising
I had hope he would have anaswered my earlier question..
Ie had he actually looked at the emails, it seated if the issues, foi , hide the decline, before he signed that Met Office statement
I would if Tamsin, Richard would be sign it so easily(as they did) ,if similar circumstances arose again..
Do I really need to submit a paper to prove the null hypothesis? Which will be the best organ for a paper written by an oxfordshire housewife? Nature? Woman's Own? Heat? Will I be peer-reviewed by housewives from other counties?
Or, what a pathetic argument these communicators put up. I can't address your question on XYZ until you reall ALL my papers and then put in one of your ownm if you can get it past the gatekeepers. Life does not begin and end at the university gates.
Simon, satellite temperatures are far from “(almost) directly measured”! On the contrary they are probably the Earth system metric that is most messed around with in order to arrive at discrete parameter values. Satellite MSU temperatures are reconstructed from atmospheric “brightness” in the microwave frequencies radiated by O2 atoms through the column of atmosphere below the satellites. The temperatures are arrived at by weighting different MSU channels to attempt to remove contaminations from the stratosphere (cold) and to correct for that proportion of the microwave “field” resulting from surface emission. Since surface emissivity varies markedly between ice and water, both the fact of seasonal sea ice expansion and retreat, and the melt pools of water with very different emissivities from ice, means that satellite atmospheric temperatures miss out parts of the polar regions. So the satellite measures aren’t “global averages” anymore.
And if one is interested in temperature trends the effects of orbital decay and diurnal drift need to be accounted for. In fact all of these effects meant that the UAH data that you refer to were pretty useless right through to 2005 when other groups addressed many of the problems with the UAH data analysis, and it’s not obvious that their current temperatures/anomalies are yet accurate.
In fact it’s useful to get to grips with the difference between precision and accuracy in measuring parameters like temperature. UAH may quote their temperatures with high precision but that doesn’t mean that they are accurate. The situation is rather similar to the NOAA average temperature metric we discussed on the other thread recently that has a spurious precision; the precision is in the temperature anomaly and not the baseline notional “average temperature” which is poorly specified in all of the data analyses (NOAA, Hadcrut, NASA Giss, BEST) for many of the reasons that Jeremy Shiers mentioned in his post just above.
Clearly if UAH and RSS (not to mention other groups) come to different values for their tropospheric warming trend, and yet quote temperature anomalies to 3 significant figures, each of these values cannot be accurate however precise the metric might be!
Simon Anthony
But it's not the satellite measurement which is being used as the base, is it?
And even then we don't know precisely what the satellites are measuring: ground? six feet above ground? city-centre? Hyde Park? beautiful downtown Burbank? (showing my age, there) mid-Atlantic? top of Everest?
Possibly all of the above, but as Rhoda says (and I seem to remember saying the other day) the only way you get a base is to take daily readings (max/min) every day for 30 years, take an average, and you have a base for that station.and only that station.
Ten yards away is different.
I'm not suggesting that calculating some form of average is impossible just that the concept of a global average has no significance unless, as somebody once put it, you happen to live in Global Average. The error bars are just too big.
And a 0.1 increase is equally meaningless because the earth is not going to react as a whole and whether whatever global warming there may be is going to be beneficial/neutral/adverse/catastrophic is going to depend on which parts of the planet heat/cool by what amount and when (summer/winter/night/day/max/min).
And they're not all going to do it in the same way at the same time.
Simon, I don't understand all the machinations of the satellite temperature figures. They are far from thermometers and the two series are not directly compatible with the two surface-based series. I do accept then as an indication of a trend in..something related to average temperatures. If we have to do this by satellite, let it be a radiation budget and albedo measurement. Thirty years of THAT anomaly might be more relevant. But not conclusive, as thirty years is nowhere near enough. Maybe a thousand years is not enough either.
We are mistaken in believing this can be resolved with proxies, models and a flaky imaginary number or global average temperature anomaly. The way the team concentrate on these is what Chiefio calls a 'dig here'. Why not better measurements? Don't they have the approved message?
Jun 6, 2012 at 5:10 PM | Jeremy Shiers
Having agreed with you re the problems associated with the usefulness of a notional parameter like "gobal average temperature", I don't believe much of the problems you describe apply to the temperature anomaly. The key point is that although absolute temperatures aren't well correlated spatially (for some of the reasons you describe), temperature change is spatially quite well correlated.
In fact the temperature anomaly eliminates much of the effect of interpolation, and the remaining problems then relate to the lack of coverage of really large areas of the globe who's temperature trend (anomaly) differs from the global average (this might apply to the Australian outback, and certainly applies to the Arctic).
So although, for example, we can't easily conceive of an average surface temperature of a mountainous region with one temperature station in this valley (av temp 17 oC say), and another on top of such and such a summit (av temp 5 oC) and another in a high pasture somewhere else (av temp 11 oC), the spatial correlation of temperature change means that if one station experiences a temperature rise of 0.5 oC over 30 years, then the others will likely also rise by about that amount.
So where an "average temperature" of the area bounded by the 3 stations with some sort of interpolation and area weighting, is a potentially dodgy metric, the temperature anomaly (near 0.5 oC) is much better specified. In effect the temperature anomaly ties the temperature metric right back to the individual temperature station data somewhat in the manner that Rhoda prefers, eliminating much of the potential problems arising from somewhat subjective interpolation procedures.
Of course if you're not happy with the absolute temperature station data then that's another problem altogether! But the use of temperature anomalies largely eliminates the problems listed as #s 1,2,3, and 4 in your post...
chris, I don't really get the magic of anomalies, but never mind. I have another tiny question regarding interpolation. Does anybody ever check the value predicted by interpolation against that of a known staion? That is to say, where there is a live station, compare what interpolation betwen neighbours says compared to the true figure.
Jun 6, 2012 at 6:37 PM | chris
Your objections to the "indirectness" of satellite temp measurements apply to remote sensing of temperature or of any other quantity. It's a rather Victorian view to suggest that only thermometers at weather stations are "proper" measures of temperatures.
As for the exclusion of poles - so what? Call it global temp (excl poles) or sub-polar temps. It doesn't matter. With a combination of different orbits the poles would be included too. You're getting too hung up on details.
Let's look at how you'd actually estimate global average temp in the most simple-minded way.
If there was just one thermometer placed somewhere at random on Earth's surface, and its temp sampled hourly and averaged over a year (could be longer - it's the long term trend that's interesting), that could be taken for global average temp. That average would differ from that of any other randomly placed thermometer on the Earth's surface by a maximum of +/-100 degrees, or 30%-50% of the absolute value. So that's a first estimate. Not great but not wrong by, say, 1000 degrees.
Now most randomly placed pairs would agree much better than that. I haven't done the calculation but at a rough guess I'd say the standard deviation of the difference in randomly placed pairs is ~10% of the measured temp. So a typical randomly placed thermometer, anywhere on Earth's surface, doesn't do badly in estimating the temperature "everywhere". Ten thousand thermometers, genuinely randomly placed, might improve the accuracy of that estimate by a factor of 100. Unfortunately, the surface station network, although it has about that many thermometers, doesn't sample in this way but is skewed in various ways. So perhaps the relevant number of surface stations is only, say, 1000. Nonetheless, that would improve the single thermometer estimate by a factor of ~30.
So, to a very rough estimate, I'd expect the surface station network to give global average annual temp
to ~0.3%, or ~1 degree, which, coincidentally or otherwise, is of the same order as the accuracy of the global temp given by Richard Betts in an earlier thread.
Now you say that global average temp is "poorly specified". I'm not sure what you mean by this. I've given a specification which I'd argue is likely to be accurate to within ~1%. Is that "poorly specified"?
As for your points about the corrections to satellite readings, well, this thread began with a post about adjustments to the ground station record. Agreed that both need adjusting (and may need further adjustments); what do you conclude?
On your precision/accuracy point, fortunately I'm well aware of the distinction and I'm glad that you've recognised it. UAH quotes sub-polar average temps to 3 dp. What do you think their accuracy is?
Chris
your point about the anomaly is well made and I will accept that calculating the anomaly is likely to show a general trend. What I find difficult to accept as being scientific is a claim that eg the earth has warmed by 0.8c since 1850. I have seen the world. It is big. And then there's the bits I haven't seen. I do not believe that it is possible to take the temperature of a body the size and complexity of the earth over a period of 150 years with all the social upheaval that has beset our home planet with decimal accuracy. The error bars are greater than the alleged increase by some measure IMHO. And the alarmists and politicians spout this crap as if it was handed down from Him, you know, the bigyun.
Rob Wilson,
This change does not affect the comparison CRUTEM3-4. Böhm et al 2010 concerns only the years before 1850. There have been a change for CruTem but version 3 was already corrected. The current release in 2000 showed significantly lower global warming trends for the Alpine area (0.5 ° C per century cooler than CruTem3). For details, see Böhm et al 2001.
Jun 6, 2012 at 4:48 PM | Simon Anthony
For AGW, the aim is to assess changes in the heat content of the Earth's atmosphere.
Which temperature doesn't do. Global average enthalpy, now there's something to conjure with.
Jun 6, 2012 at 6:15 PM | Rhoda
'Heat' magazine sounds like an appropriate journal for a global warming paper. I never heard of it; does it have good reputation?
So Rob, where are these "insulting emails"?
Enquiring minds want to know.
If you have them post them on this blog, with names/IDs.
I'm sure that the Bish would take a dim view of those that engaged in such behaviour.
On the other hand, you have simply "upped sticks" stating
"I think this is enough from me on this thread"
Makes me think that said emails are rather less than "insulting".
Jun 6, 2012 at 7:33 PM | Rhoda
Yup, an excellent question. I don't know the answer, but I expect that checking the robustness of an interpolation using the method you suggest is something that will have been done. No doubt if we could be bothered to trawl through the papers describing methodologies for surface temperature measurement we could find out!
Heh, it was the emails alright; the ones from East Anglia. He's left here, but he's still standing by his corrupt letter.
============
Jun 6, 2012 at 8:49 PM Don Keiller
Don, That possibility had crossed my mind. See my posting at Jun 6, 2012 at 2:16 PM . It will be interesting to see what comes to light.
Rob used one argument that many "warmists" have used which is totally bogus in my opinion.
We all know that even the temp figures from the people we dont trust tell us that the world has not warmed for about 14 years. However the warmists tell us that 14 years is not statistically significant and that we should wait until 30 years have passed before a trend can be established.
That argument holds true if what you are doing is trying to calculate statistical trends but we are not.
We are discussing the influence of CO2 on temperature, I am not aware of any simple fisics wot demands a 30 year wait!
Levels of CO2 in the atmosphere are rising year on year. The IPCC claims that CO2 is the be all and end all of global warming/climate change. The world is not warming therefore CO2 is NOT the main driver.
CO2 works by "absorbing" or holding radiation that would otherwise be reflected back into space, there is no time lag in the so called simple physics.
Another somewhat O/T but check
http://wattsupwiththat.com/2012/06/06/noaas-national-climatic-data-center-caught-cooling-the-past-modern-processed-records-dont-match-paper-records/
Jun 6, 2012 at 7:37 PM | Simon Anthony
It all comes down to what you're trying to assess Simon. If you wish to determine the surface warming since the start 20th century (say) which has a value around 0.8 oC, and your metric for average surface temperature might be accurate only within 1 oC, then the latter metric isn't very useful for assessing temperature change.
In that context the global average temperature is "poorly specified". In fact both NASA Giss and Hadcrut/Met don't attempt to specifiy error bounds (though Hadcrut think the average global temperature is between 13.5 and 14.5 oC, if I remember correctly). I emailed NOAA in the context of the discussion on the earlier thread to determine what they meant by their value for a global average temperature, and they don't consider this an accurate metric either. They use what they call an "off the shelf" value from a paper as a "baseline" number from which anomalies are determined (obviously to calculate a temperature anomaly you have to have some sort of a baseline, but this can be rather arbitrary).
Your "accurate to within 1%". That's also somewhat arbitrary, since we can change the apparent accuracy by our choice of temperature scales. It's accurate to within 0.3% if we use the range of Kelvins down to absolute zero, and it's accurate to only within ~10% if we consider the apparent range of surface temperatures covering the entire Phanerozoic. Context is key.
On "adjustments". Adjustments to surface temperatures from met station and sea surface temperatures seem to be rather small (e.g. comparing th old and updated Hadcrut series) and generally result in refining of the metric as new information becomes available. The satellite adjustments are large and profound. After all, as late as 2005 the UAH team were still interpreting their MSU TLT brightness temperature as indicating a cooling trend in the tropics and a warming globally of around 0.1 K per decade cooler than the RSS analysis. Analyses by other groups gives different trends. Clearly the large range in temperature anomalies in satellite measures of tropospheric temperatures is much greater than the range of temperature anomalies in surface temperature measures amongst NOAA, NASA Giss, Hadcrut-Met, BEST etc.
What is "UAH accuracy"? I'm not sure we know. For absolute temperature it's difficult, since the MSU samples the entire microwave field below the satellites and has to use various weightings to assess temperature contributions from different levels in the troposphere. Again the problem with “average temperatures” is highlighted by the fact that UAH determine a “sea surface temperature” that varies in the range between 20.5 and 21.5 oC (throughout the year) [*], whereas the NOAA site [**] has a baseline “sea surface mean temperatures” that varies in the range 15.5 oC to 16.5 oC.
Obviously they’re measuring different things, but it’s not obvious what (and I’m not going to bother NOAA again with a query about what their metric actually refers to!). But this highlights the problem with measures of some notional global average temperature. If anyone knows the answer to that question that would be much appreciated!
[*] http://discover.itsc.uah.edu/amsutemps/execute.csh?amsutemps+001
[**] http://www.ncdc.noaa.gov/cmb-faq/anomalies.php