Well, possibly a more interesting title than ‘Is the near surface temperature record robust?’ All of the above, plus more besides, can affect the readings at temperature stations. Anthony Watts has been collecting photographs of the USHCN climate stations. He is now up to ‘How not to measure temperature, part 22.’ He has two websites Watts up with that? and surfacestations.org
Obviously, in order for surface temperature measurements to accurately reflect temperature trends, all non-climatic influences must be removed. Initially, Anthony examined the fact that Stevenson screens that house the temperature sensor used to be painted with whitewash, but have been painted with latex paint since 1979. He has purchased 3 Stevenson screens, one has bare wood, one is painted with whitewash, and the other with latex. The initial results were posted here. The latest results will be available in about a week’s time.
Anthony’s work has also caused a ‘blog war’ between Hockey Stick/Big Man-Made Warming defenders Real Climate and Roger Pielke Sr’s Climate Science blog.
Moving on to the peer reviewed science cited by IPCC WG1 in order to support the robustness of near-surface air temperature trends, Roger Pielke Sr claims on his blog, “The IPCC WG1 Chapter 3 Report clearly cherrypicked information on the robustness of the land near-surface air temperature to bolster their advocacy of a particular perspective on the role of humans within the climate system. As a result, policymakers and the public have been given a false (or at best an incomplete) assessment of the multi-decadal global average near-surface air temperature trends.“ Pielke Sr has listed the papers cited by the IPCC, and those that weren’t. Additional evidence has been posted here.
You might agree that life is never dull in climate science. The good news is that Australia’s Reference Climate Station Network seems to be a model for how temperature should be measured.
I now have a list of UK stations used by Phil Jones et al. Maybe I’ll try and obtain photographs of each site. There again, maybe not!
Regards,
Paul Biggs
Jennifer Marohasy BSc PhD is a critical thinker with expertise in the scientific method.
My first post, unaided. I haven’t worked out how to upload photographs, otherwise I would have probably shown the one with the car and air-con unit next to the temperature sensor.
Jennifer, others,
I’m in need of a bit of explanation here. What are “Stephenson screens” and what is their function?
Also, I note “air temperature” is plotted separately. Isn’t the entire point to plot air temperature? Maybe these are two different sorts of devices. If so, someone should at least settle on one which (a) logs air temperature and (b) is not paint-dependent.
Paul
You might be interested in a post I did on the station at the Royal Observatory at Edinburgh, which appears to be close to buildings and carparking.
Hmm. Links not accepted here, I guess. There’s a search function on my blog, so I imagine you can find it if you’re interested.
Paul, quite a few of the Aust sites are at airports; the e-pan rate at Richmond RAAF base was always a little higher than the one at Hawkesbury Ag College, just down the road.
More exposure?
So Schiller – you’ve been arguing with us about climate matters for how long and you don’t even know what a Stevenson screen is – Gads !
http://en.wikipedia.org/wiki/Stevenson_screen
If you click on ‘3 Stevenson screens’ the link shows a picture of them. There’s always google!
Luke,
Thank you for your reply to my request for information.
I see that you are about as informative as usual, which is to say, not informative.
Perhaps others here will take up your slack.
Bishop, others add links without any trouble … just provide the text of the url but without formating.
“As a result, policymakers and the public have been given a false (or at best an incomplete) assessment of the multi-decadal global average near-surface air temperature trends” – this would have to be a huge try-on wouldn’t it?
What does the the well regarded “Paul Biggs certified” Australian anomaly pattern look like. Boreholes? SST’s. Artic and glacier melt rates.
Has anyone found what size error we’re talking about as a result of a few bung stations. Tamino at Open Mind’s issue with large samples?
So if the data series are really no good why do contrarians bang on incessantly about cooling since 1998 and 1940-1970 – and hang on every wiggle in the graphs – surely you don’t know then? Can’t have it both ways.
All this is an ongoing desperate try-on to destabilise and increase the level of uncertainty. Usual denialist nonsense. Typical climateaudit style shit – why don’t they get of their arses and do something constructive.
Gee this means all Sid Reynold’s cherry picked extremes aren’t worth a bumper either. His thermometers are probably in someone’s freezer – has anyone checked? Maybe they’re in a hollow getting too much cold air drainage.
Paul: All this focus on ST measurement station perfection is an extreme. I bet this whole lot have not done a series of practical measurements out side since school lab days.
Luke, with all that scintillating defensive footwork, have you ever considered coaching?
Perhaps these defensive talents would be better employed with the Australian soccer team.
They could use any help you could give them, in that area, at present.
Forest stations in southwestern Australia used (1960s) to measure evaporation from a pan outside as some indicator of fire hazard. One station showed very anomalous evaporation, until it was noted that the officer-in-charge had a thirsty dog. There were also startling tales of sprinklers and Stevenson screens (we even had them then, Schiller). That was just before the GAGCS (Great Anthropogenic Global Cooling Scare). By the way, didn’t I see Prof. Schneider looking wise somewhere recently?
Davey – negative evaporation has also been observed after research station Xmas parties. 🙂
Lukes about about cooling being measured globally during the 40’s to 70’s? All I hear is the usual crickets chirping. What happened to the UHI during that period of massive industrialisation?
So Phluke would have us conduct this discussion without any data at all? Give us a break you plodder. The fact that we use the data does not preclude us from identifying shortfalls in the data or from improving that data, you casuist fop.
And the fact that white wash is a paint that is not widely used these days, and needs to be re-applied more frequently than latex based paints means that there is likely to have been a wholesale shift in the coatings of all Stevenson screens over recent decades. This is NOT an issue of a few wayward records but, rather, a shift in the majority of records.
And the initial indications are that this shift, along with UHI and related land use changes, is capable of explaining most, if not all, of the reported increase in global temperatures.
These are obviously early days, and too soon for drawing firm conclusions. But it is clearly an issue that needs further investigation.
It is a topic on which the late John Daly had quite a bit to say. http://www.john-daly.com/
The “initial results” are not of Stevenson screens, but of planks of wood sitting in the sun.
Davey, you will note that anomolous readings were detected in an instrument and rectified. Isn’t that an example of the system working?
Ian Mott – “And the fact that white wash is a paint that is not widely used these days, and needs to be re-applied more frequently than latex based paints means that there is likely to have been a wholesale shift in the coatings of all Stevenson screens over recent decades.”
Why? How do you know that the whitewash was not applied at the correct intervals? How is photographing them going to show this? Are all the cameras used for the ‘survey’ color calibrated centrally to ensure that all photographs are the have the same color balance? It is not at all likely that just because whitewash paints were used that this would bias the record – you are grasping at straws now.
SJT,
No, I think the pan system (an old ploughshare actually) was abandoned in favour of pine dowels which were weighed each morning – this was regarded as more scientific. Until it was found that pine dowels vary significantly in water uptake. I suspect Luke is right about negative evaporation (or unexplained cloudless prepissitation) after parties. Really, can we rely on any instruments? Back to the Earl Grey chaps…
It is interesting to note that all of the various bias factors affecting climate stations result in increased temperatures and their influence also tends to increase over time. One wonders how much of the 0.6°C warming of the past century is due to this cause.
Good afternoon Ian. I trust you had a pleasant weekend killing saplings. How are you today.
Noun: casuist. Someone whose reasoning is subtle and often specious
Noun: fop. A man who is much concerned with his dress and appearance
hmmmm – OK I can dig it.
The meteorological data are no good because they’re trending upwards. Every met station is next to an air conditioner eh?
So yes that’s really intelligent – so all the other accelerated warming trends in other data sets are pigments of our imagination too (giggle).
Ian come back when you can string a credible story together. John Daly indeed – you mean bulldust central – ROTFL.
In terms of white wash – there’s stacks in warehouses – it’s been bought up by the denialist movement.
BTW – all this is an aside – we’re still waiting for your maths that can show that the an orbital change alone can make the interglacial temperatures without a CO2 forcing.
Hey; some one ran off with the plot here. I reckon most if not all these stations have gross imperfections when it comes to blog criteria.
Weather stations in general were only set up for monitoring daily changes as part of a local forecasting system. Few were baseline and I guess hardly any were ever a national standard in the strict sense. Also as numerous as they are today we only get what we pay for.
In the early days it was all hand recorded round this country. Folks minding the gear after it was commissioned probably smoked while they had a coffee on the job too.
I once did a lot of measurement for big biz during our transition to automation. What we had fifty years back in industry cost heaps but what the ST stations cost then in total was probably a great deal less by comparison.
This latest post Daly sortie courtesy P the M etc can become the latest cult but in all my time doing system start ups and peripheral device calibrations I never used a camera, not even once to show an error! I would be strapped to find say a dozen shots of stuff we built from scratch too.
The best analogy I can think of for our readers in understanding this zealous search for faults in the global sensing process is some village group setting up cameras around our motorways.
One photo link shows a sensor near the AC unit for the electronics. I reckon they used a decent tele lens to get it and with the angle depicted we don’t know what the actual separation distance is. We certainly don’t know what effect that local AC had on ambient conditions around the station from a dozen or more photos.
One wonders how the weather stations around the globe seemed to go up and down at roughly the same time over the last century. Did they all have airconditioners installed next to them and the roads paved next to them at the same time, including the records taken at sea?
What I find amazing, is that the temperature sensors and designs of the buildings are so remarkably accurate compared to the rest of the global temperature record. They follow the same basic curve up to the ’40s, dip down then rise again in the ’70s, have the big jump in 1998. All I can ask is, has he actually compared their output to stations that he thinks he can trust?
Like the Australian reference set or Macquarie Island.
Anyway bring it on – let’s toss the stations that are shown to be compromised and rerun the analysis. Yes we do not want poorly sited stations, we do want stations that are well maintained, and yes full public access to the data would be good.
There are well established standards for these issues.
The number we are not getting is how many stations in the global set are compromised – 1%, 5%, 10% , 50% ??
Bit of a comprehension problem today, Ender. The point is that few people even know what white wash is these days, let alone still use it. It is not a question of whether the paint was reapplied on schedule but, rather, whether the same paint was reapplied. Not likely.
And not to be outdone, SJT gives us the analytical skills of the village idiot. If an AC unit, or paved runway is increasing temperatures by a fixed amount each day, and part of the day, within a normal diurnal variation of 10C or more, then there is no reason why a whole bunch of them will not produce similar numbers.
The total recorded warming since 1939 has only been 0.5C while daily variations are usually more than 10C, more than enough to mask such influences.
But according to Phluke, any attempt at validating the integrity of “convenient” data is nothing but a right wing plot.
Casuist, theologian etc who studies and resolves cases of conscience (or belief) etc, esp. with false but clever argument.
Fop, dandy, vain man.
Bishop Hill – thanks, I’ll have a look.
Temperature trends are a time series, so we need to know when the likes of air-con units were installed.
The point is that, even aside from poor stations, there is likely to be a warm bias in the near-surface ‘global average’ temperature measurements. That is not to say that the world hasn’t warmed, but it might not have warmed as much as is claimed.
I don’t see how anyone can defend poorly sited stations being used as part of a network assessing temperature trends. There is also the issue of land use changes around stations.
There are those, such as Pielke Sr, that believe ocean heat content is a better metric with which to assess temperature trends. There appears to have been little or no ocean warming during the past 5 years.
Ender : (aka Climate nerd)
Anthony Watts IS you. You ARE Anthony Watts You go together like peas and carrots. He has a solar panel on his roof. You talk about putting a solar panel on your roof! He is a climate nerd. You are a climate nerd. BE Anthony Watts. EMBRACE Anthony Watts. You know you want to.
Ian
thanks for that comment Ian. As I said, they track the global record, and I would like to see how they compare to that.
Well better to be a climate nerd than a climate turd I guess. (Mottsian precedent humour – that’s a joke Joyce).
Anyway be that as it may, I just did a scientific experiment – I yelled out “You’re all a bunch of casuist fops” – 6 didn’t flinch, 2 looked up briefly, 1 gave me the finger (but she always does), and 1 said he hadn’t seen the mop. I then yelled “You’re all a bunch of casuist dandies” – 2 nodded, 3 didn’t flinch, 3 gave me the finger, and 2 flexed off.
Anyway – la de dah – on the central limit theorem or something like that – Paul you should know having done some stats.
From:
http://tamino.wordpress.com/2007/07/05/the-power-of-large-numbers/#more-286
A commenter (”Citizen”) on another post opines that because thermometers are imprecise, and not continually calibrated, it’s not possible to discern temperature changes precisely enough to support the temperature increase claimed by climate researchers. In fact, he (she?) states:
Do you think that continually averaging imprecise information makes that information more precise? Don’t you think that’s a stretch? It’s not rational you know.
It’s certainly not intuitive. But in fact it is true.
{ENDS} continues at the site.. .. and worth a read.
I know it’s a tad intellectual for this blog but there you are nonetheless. So up the bum to Ian’s thermometer precision ruse. Probably why when I asked David Jones about this many moons ago he just rolled back his eyes and said “central limit theorem. For heavens sake”.
Anyway I just yelled our “Casuist vain man” – only 5 were left – 3 didn’t flinch, 1 gave me the finger, and 1 said “will you shut up and stop yelling”.
A mathematician has a different view of ‘average global temperature:’
http://www.numberwatch.co.uk/Temperatures.htm
Paul: How do we possibly cope with the usual imperfections after buying a prepacked Kg bag of carrots or brown onions and a 2Kg bag of potatoes at the supermarket?
Sure imperfections in measurements and data are likewise annoying but we should each learn to deal with them in time. Living off the farm instead of the lab helps in both cases.
Paul,
John Brignell is correct. ‘Average global temperature’ is dodgy indeed.
Indeed, ‘average global temperature’ is a function of the number of sites and their location. Ideally, stations would be equally spaced accross the entire surface and away form any non-climatic influences. We are, after all, dealing with fractions of 1C, so accuracy is important.
No one seems to have mentioned the papers cited and not cited by the IPCC, despite the fact that they were available.
Broken links now fixed.
Paul Biggs – “Indeed, ‘average global temperature’ is a function of the number of sites and their location. Ideally, stations would be equally spaced accross the entire surface and away form any non-climatic influences. We are, after all, dealing with fractions of 1C, so accuracy is important.”
And in an ideal world that would be the case. However scientists, lacking the billions of dollars that it would take to create such a perfect system have to work with what they have. Scientists are used to working with imperfect data as it is very rare or impossible to have perfect data.
This is how it is done – if you have an issue with it I suggest that you petition your local member of Government to invest the money that should be invested in the climate network. However as usual the luggage capacity of the government jet is far more important than accurate climate information.
http://www.ncdc.noaa.gov/oa/climate/ghcn-monthly/index.php
Ian – “The point is that few people even know what white wash is these days, let alone still use it. It is not a question of whether the paint was reapplied on schedule but, rather, whether the same paint was reapplied. Not likely.”
As the color affects the temperature gradient of the enclosure I would have thought that as long as the whitewash was the same color as the latex paint then what the paint is made of is immaterial. You have no way of qualifying the change in paint colors over the years.
What are you trying to say that the entire 0.6 degrees of warming is an artifact of paint color or are you just fishing for doubt?
Paul: with most instrument recordings, trend is more important than accuracy and dare I say it that also applies to your pocket tape measure given many still in use have lost some definition near zero over the years.
There are some novel ways to calibrate our weather thermometers without a camera.
The weather forecast on ABC TV tonight for Hobart, Melbourne and Canberra gives the expected maximum for each place in the range 10-11 C and that is so common for a cold snap ridging up from the deep Southern Ocean at this time of the year. This is one good opportunity to see the difference altitude makes in such a fast moving air stream. There have been many others over a long period of my watching this type of event.
What’s the bet Canberra is as cold as Hobart on most of these occasions and what’s the bet all three records come from our airports? The big question I have though from the above comparisons is about the likely UHI effect if any from the old Essendon Airport on past records.
It would surprise me greatly if BoM has not done these relative simple checks over decades to prove their systems respond to various uncertainties appropriately. The public record such as we have in the media every day must be challenged by many amateurs every which way over time.
BoM had quite a few calls from me through periods of extreme weather searching for extra details beyond the daily record based on my own instrumentation and perception of the events. In the end regardless of gear we are left with only the trend. But their max/min like my max/min becomes the basis of your understanding.
If the IPCC were confident in the robustness of the near surface data, then they wouldn’t need to cherry pick papers. Furthermore the station data for Jones et al 1990, which the IPCC relied on in order to claim that the UHIE was small, is ‘lost’ and therefore cannot be verified.
To better understand my slant on the Australian side of your records dilemma I suggest this quick history covers my earlier comments. At any time since about 1960 I had good reason to follow up on international instrument system improvements as well, particularly with remote sensing techniques.
http://www.bom.gov.au/inside/eiab/Bureauhistory.shtml
Australian science in a few areas actually led the way. This country’s weather specialists covered very large areas indeed in terms of distance and service difficulties.
From memory the first electronic devices everywhere related to the development of platinum resistance bulbs. Data logging in marine environments probably came in during the mid 70’s, fully automated weather stations only appeared from about the early 80’s. A few of my technical mates had first hand experience calibrating and recording on the ice.
http://www.bom.gov.au/weather/ant/casey/history.shtml
A really clever move would be for some one else to find out exactly what happens to gear left out overnight in all these other wild regions before worrying too much about instrument skirts designed to provide shade out in the desert.
Is our Ralph your typical target?
http://www.austehc.unimelb.edu.au/fam/1419.html
Or is it these poor guys? Note well the best of British engineering was here too.
http://www.bom.gov.au/weather/sa/inside/history/
Hey; why do I bother? I’m absolutely sure there is enough data tucked away in our stations and their history to sink the UHI furphy as well.
Paul Biggs – “Furthermore the station data for Jones et al 1990, which the IPCC relied on in order to claim that the UHIE was small, is ‘lost’ and therefore cannot be verified.”
Says who – McIntyre or Warwick Hughes????
Show us the reference!! Here is one for you though.
“A paper by Peterson (2003) is of interest because it has been out for a while and is more comprehensive. It agrees with Parker. The paper, “Assessment of urban versus rural in situ surface temperatures in the contiguous United States: No difference found” published in the Journal of Climate finds that the effects of the urban heat island may have been overstated and that “Contrary to generally accepted wisdom, no statistically significant impact of urbanization could be found in annual temperatures.”. This was done by using satellite-based night-light detection of urban areas, and more thorough homogenisation of the time series (with corrections, for example, for the tendency of surrounding rural stations to be slightly higher, and thus cooler, than urban areas). As the paper says, if its conclusion is accepted, then it is necessary to “unravel the mystery of how a global temperature time series created partly from urban in situ stations could show no contamination from urban warming”. The main conclusion is that micro- and local-scale impacts dominate the meso-scale impact of the urban heat island: many sections of towns may be warmer than rural sites, but meteorological observations are likely to be made in park “cool islands”.”
http://www.realclimate.org/index.php/archives/2004/12/the-surface-temperature-record-and-the-urban-heat-island/
Paul: Now let’s have a good look at this girly science Wattsy – Pielke style. Something any school kid should know though is that by examining the shroud we are missing the core.
Sensing errors generally start with the bulb hey
At mid point in those wooden slat curves there is virtually no deviation between them in response to air temperature change. Damping factors nill? At comfortable ambient temperatures the delay is negligible, oh and that’s just above our global average. It all seems like someone has been there before. Hmmmm!
Let’s guess our forbears had some wit hey. I also reckon they had other problems like deviations between their instruments and various readers.
No, Ender, the whole point of the graph was that it clearly indicated temperature differences between the two types of paint. And this suggests that the temp records may vary due to the type of paint used. The scale of the variance due to different paint types appears to be at least equal to the total change in global mean temps and therefore, must be examined further.
That is still a long way short of claiming that this accounts for all of the recorded global warming, as you have rhetorically inquired.
And thanks for that link to John Brignell, Paul. I seem to recall Luke launching into one of his classic undergraduate sneers the last time I tried to say the same thing, albeit far less lucidly than Brignell. I felt a powerful urge to go out and rent some pimples.
Gavin,
In reference to you suggestion that overnight readings in remote areas be analysed, Johnathon Lowe at http://www.gustofhotair.blogspot.com has been accessing met readings from around Australia, and statistically analysing temperatures at set times; 3.00 am. 6.00 am, 9.00 am. His posts are interesting.
Well Ian undegrad sneers beat your grandad at the pub routine.
Anyway back to Brignell – well have a gecko at the maps on http://www.remss.com/msu/msu_data_description.html#msu_amsu_data
Do you reckon the surface temperatures are telling you a grossly different story?
Incidentally I can tell you still haven’t pondered the central limit theorem.
Thanks Helen,
I found Jonathon’s remarks interesting. I think a lot of the debate centres around the use, and misuse, of statistics. It’s a tricky subject, and needs very clear thinking. I have been at it for decades, but am still often baffled. The danger is that Stats 100, or even Stats 200, do not guarantee valid analysis. In fact, such units, combined with ready access to statistical software, can lead to overconfidence. I remember Schiller posting something on psychological research into the inverse relationship between self-confidence and actual competence.
Luke,
My memories of the Central Limit Theorem are festooned in cobwebs. However, I think it tells us that the distribution of sample means converges on the Normal Distribution, even if the parent population is not Normally Distributed. Further, the mean of the means will converge on the population mean, but the variance will diminish with increasing sample size. However, if the parent population is biased in some way (could be instrument error?) then the mean of the means will converge, very nicely, on the biased mean. Have I got it right?
Helen: thanks but overnight info, max/min etc is not quite what I was after today
It goes like this at about 2.30 pm – Hobart 6, Devonport East 6, Wilsons Prom. 6 Essendon 6, Canberra 6, my crude max/min u tube in the back yard is currently 6 and a bit, dry bulb in the sling 6.5 and probably falling
Am I on a heat island?
Hardly! Its too darned cold to stay out side and wait for the Hg bulbs to settle.
Luke
how close are those satellites to the space shuttle exhaust?
Oddly enough Laverton is 6, Moorabbin 6 and Melb. Airport is 7 however East Gippsland, Latrobe Valley are way up at 11C with Mallacoota at 13 as I post.
More disingenuous crap from Luke, as if one could possibly judge whether a data set is valid by looking at a map of that data set.
You have this extraordinary capacity to merge fact with image or perception as equal values. Did you mother happen to have the flu in your second trimester by any chance?
Oh gee I’m sorry I didn’t supply you with a DVD of the data with full documentation. Look up the detail on the site yourself boofhead. I think Ian’s climate life just had a wakeup call so out comes the abuse on cue. Who’s a widdle bit angwy and done his nana then – kitchy koo.
Helen, Realclimate has this to say on average global temperature.
“[Response In the late 19th and early 20th century, when meteorological networks were expanding and government agencies were trying to standardise observations, a lot of thought was given to how best characterise the daily cycle of temperature. Many papers were published about this and eventually the (Max + Min)/2 convention was agreed upon as a simple but effective way of computing the daily average. More recently, studies have looked at daily minima, and daily maxima and how they have changed over time, and how this may have changed the daily (diurnal) cycle. In many places, this has changed due to an increase in nocturnal cloudiness, for example. Such effects are of course rolled into the global mean. Your hypothesis that the record of global mean temperatures might have been affected by the odd warm hour on a spring day here and there has a very low probability of being correct, given the vast amount of data that goes into the global mean, from stations in all pats of the world (from the fully dark Antarctic winter days to the fully illuminated Arctic summers, desert and equatorial forest sites etc etc). However, I encourage you to investigate this using all the hourly data sets that you can lay your hands on, and report back to realclimate.org -Ray]”
They are way ahead of you, and Gusto, the debate was held over a hundred years ago on what would be a simple but representational figure to use to describe average global temperature. As they say, the numerous observations taken will smooth out the erroeous data, not to mention the box the measurements are taken in, that is designed to smooth out any transient changes in temperature.
Scientists today are aware of the issue, but it’s once again making a mountain out of a molehill. As Luke has said, the errors will tend to cancel each other out.
Ender – regarding Jones et al 1990 – the University of East Anglia were unable to comply with the FOI Act request for the data:
In your email of 17 April 2007, you re-iterated your request from your email of 12 March 2007, to see
“B) identification … of the stations used in the gridded network which was used as a comparandum in this study”
I have been in conversation with Dr. Jones and have been advised that, in fact, we are unable to answer (B) as we do not have a copy of the station data as we had it in 1990. The station database has evolved since that time and CRU was not able to keep versions of it as stations were added, amended and deleted. This was a consequence of a lack of data storage comparable to what we have at our disposal currently.
I have been advised that the best equivalent data available is within the current version of CRUTEM3(v) or CRUTEM2(v). The latter is still available on the CRU web site, though not updated beyond 2005.
These latest versions are likely different from what was used in 1990. Australia and China have both released more data since then – it is likely that much of this was not digitized in 1990. Dr. Jones acknowledges that the grid resolution is now different, but this is again due to greater disk storage available.
The details of our updating of the raw station data is discussed in the following article:
Jones, P.D. and Moberg, A., 2003: Hemispheric and large-scale surface air temperature variations: An extensive revision and an update to 2001. J. Climate 16, 206-223.
This is, in effect, our final attempt to resolve this matter informally. If this response is not to your satisfaction, I will initiate the second stage of our internal complaint process and will advise you of progress and outcome as appropriate. For your information, the complaint process is within our Code of Practice and can be found at:
http://www1.uea.ac.uk/polopoly_fs/1.2750!uea_manual_draft_04b.pdf
Yours sincerely
David Palmer
Information Policy Officer
University of East Anglia
Helen: play with this global system map of oz downunder after clicking on your spot. See also some raw station data for our country airports.
http://www.wunderground.com/global/Region/AU/2xTemperature.html
Hint: some local concrete industries use another version for their outdoor jobs
What happened to the Central Limit Theorem?
Ian Mott – “No, Ender, the whole point of the graph was that it clearly indicated temperature differences between the two types of paint. And this suggests that the temp records may vary due to the type of paint used.”
So how would two different types of surface treatment affect the temperature reading inside if they are the same colour?
great link for temp Gavin, thx I will look properly when i get time
I see a lot of desperation here trying to push the notion that blatantly artificial environments are “natural” places to measure temperature.
Poor Alan deserves compassion for his benighted compulsion.
“Save the data” sounds a lot like “Save the Patagonian toothfish.”
Toby: it’s also a great link for our weather charts and their animation.
The point I wanted to make is we have hundreds of measurements virtually on line all the time so we can see any anomalies recorded at home and anywhere else by the hour. Get all your kitchen thermometers out of the draw and any wall mounted devices outside in the shade an have a go at calibrating them against the “system” while it’s steady but near freezing. Have another go in the summer hey.
http://www.wunderground.com/global/Region/AU/2xTemperature.html
Let’s talk about errors: Post war 2 industrial instruments were normally plus and minus one or two percent of max in their working range. Those with linear scales frequently had zero errors but we don’t throw them out because it’s easy to make a mental adjustment each time we use them. U tubes often have two zero errors. Domestic thermometers could be non linear and that’s a bit more of a problem.
The trick is to remain honest about our individual limitations dealing with all such sensors. Going with the tide is a fairly safe response to uncertainties. My job from the late 50’s onwards even with our best instrument engineering for recorders and controllers based around the above sensors was all about catching up with the “tides”. Big system lags in measurement caused major problems in phase relationships and response.
Weather stations likewise are just the peripherals in our climate biz. Stay in tune.
Luke,
You are now silent on the Central Limit Theorem, although you mentioned it twice before. There is a difference between precision and accuracy. Imagine a shooter who can get ten shots in a group the size of a penny – very precise. But what if that group’s centroid (mean point of impact) were on the edge of the target, or off it altogether? The shooter would be precise, but inaccurate. Perhaps adjusting sights would solve the problem.
The Central Limit Theorem guarantees that means of repeated samples will converge on a Normal Distribution, with a mean precisely equal to that of the parent population. But that mean is not necessarily accurate.
Or am I a victim of Earl Grey toxicity?
Pushy – yes your stats recall is correct. The issue I’m talking about is the accuracy in measuring the instrument – given we’re looking for 0.5C variants in temperature. I’m told that annual average global temperatures have a standard error of estimation of about 0.025C. Obviously observors cannot measure that accurately.
Given your bullseye analogy, illustrated well at http://en.wikipedia.org/wiki/Accuracy_and_precision , as long as you’re biased equally around the true measurement, then a lot of samples should converge on a reasonable estimate of the true mean.
If the gauge is inappropriately sited (biased for air-con units – whatever) if that bias is consistent and we’re measuring an anomaly with these effects factored into the base, how much does it matter.
Dave: I could remain intrigued by you guys playing with your math in recall but there are far more basic questions remaining about Stevenson screens, like what is inside today versus what was there way back when these disputed records started.
Let’s assume everybody knows except me however I could bet nobody knows the individual error of daily readings place to place around Australia.
How were temperature readings actually made?
Forget the US and the UK, how were we doing by our own standards over the period since these screens were introduced?
Ender, Ender, Ender, would a gloss white paint reflect more energy than a flat finish? Of course it would. Would the composition of the paint also have an impact? CSIRO seem to think so. See;
http://www.infolink.com.au/dir/Coatings/news
or just google “reflective paint”.
Alan Hunter,
Thank you for agreeing that the “blatantly stupid photos” show how global temperature is measured.
Now you must agree that measurements of global temperature are, not merely questionable, but “blatantly stupid.”
Says Schiller who didn’t even know what a Stevenson screen was. ROTFL.
Luke: Schiller has yet to offer us info about what goes on inside any of those boxes in his territory. Also very few offer a view on the local scene.
Chaps: it’s the inside that counts not the outside.
But we can guess the fillings of some wooden heads hey
RTD linear sensors
http://content.honeywell.com/sensing/prodinfo/temperature/technical/c15_136.pdf
Thermistor
http://en.wikipedia.org/wiki/Thermistor
non wooden heads: NTC self heating and scaling problems see Stuart Ball
http://www.embedded.com/shared/printableArticle.jhtml?articleID=9900485
Thermistors versus RTD’s
http://www.automatedbuildings.com/news/jan02/art/bapi/bapi.htm
In remote sensing it was common to calibrate loop components to each sensor, ie the transducer and receiver can easily accommodate zero errors for say B type RTD’s however lets reckon with the cost of our latest electronics made in China, we throw the first dud sensor, chip, board and all till we find one that’s spot on the desired output.
Luke,
In science inaccuracy due to bias can matter a great deal. The man doing the paint trials is showing excellent scientific scepticism. We should all welcome his initiative, whatever the outcome. Anyone who rejects that may (or may not) be a scientist in name, but certainly not in spirit.
Are you peeved at having your statistical bluff called on the Central Limit Theorem? Have another cucumber sandwich, dear lad. More tea vicar?
P.S. Yes, Wiki explains it better than I did. Thanks for the reference.
Err I wasn’t bluffing nor am peeved – the claim was that the resolution accuracy of the reading (0.5C?) fundamentally limits the ability for change detection in a population of like managed thermometers. I knew you thought you were having a win. Such subtle nudging by yourself.
Also how much does bias matter if we’re dealing with an anomaly – something a bit warmer warms a bit more??
Yep the paint trials are excellent. Different enclosures also need a check – e.g. those beehives. Paint will have different infra-red absorbing properties although it will look white, but I thought all these systems should be installed and managed to a specified standard. I’m surprised that the yanks have this crap. But again – how many are talking about – silence ! 1%, 5%, 10%, 50% or what. Simulated consequence of errors is?? Review of the literature by proponents here is??
Assertions made before checking !
I reckon the UHI issue is overrated. How many are also in parks and gardens within the city??
Anyway the yanks should be tossing all stations that are dodgy – find’em and kill’em. Remove them from the data series and reanalyse. Bring it on !
Gavin – big issue with satellites and sensing is different platforms (spacecraft), sensor drift over time, and nasties like bi-directional reflectance/sun angle. Lotsa traps for young players.
Game, set and match to Davey. Ah, strawberries and cream in the big tent! I say, is that the Poo over there?
Luke: we are not quite up with satellites yet, still playing with those little houses and your average Jo blow at 2 M above ground. Btw those industrial sites as depicted round the blogs are familular zones for routine high quality measurements regardless of their landscape.
Remind me not to bother with a holiday in the US hey
Surely accuracy in temperature measurement is important, the stations are not measuring an anomaly, they are measuring temperature, the anomaly is calculated later, using the data from this “high quality” network.
The fact that so many AGW proponents seem dismissive of this project astounds me. If the data is dodgy, surely we all agree that it needs to be upgraded, then let the results fall where they may.
I agree with Luke that the dodgy stations should be removed from the record. (I don’t often agree with Luke!). Then let’s see what we are left with.
Maybe something like this! http://www.john-daly.com/stations/ireland.gif
Ian Mott -“Ender, Ender, Ender, would a gloss white paint reflect more energy than a flat finish? Of course it would. Would the composition of the paint also have an impact? CSIRO seem to think so. See;
http://www.infolink.com.au/dir/Coatings/news
or just google “reflective paint”.”
Ian Ian Ian, how is one happy snap going to verify the finish, matt or otherwise, of the various paint surfaces of the non-verified record of measurement stations?
BTW great link – I bookmarked it however nothing on the surface finish affecting the internal temperature of a Stevenson screen enclosure.
The standard Stevenson screen, in use all over the world, allows ventilation as it is designed. The surface finish I am sure will have only a tiny effect. Actually it would be pretty easy to work out. Isolate the temperature records from areas of very low average windspeed and compare them to areas of high average windspeed and see if the temperature has a bias. The screens in areas of low windspeed would be more affected by the finish of the enclosure.
There is a project for you Ian – better than taking happy snaps.
BTW this is the word on Stevenson screens in OZ
“Temperature measurement and the Stevenson screen
To measure the temperature of the air accurately,
it is important that the thermometer is shielded
from direct sunlight but is still exposed to a good
airflow. The standard screen used internationally
to shelter instruments is a double-louvred wooden
box, with the instruments 1.2 to 2.0 metres above
ground level. This screen, known as ‘a Stevenson
screen’, was designed by Thomas Stevenson
(1818–1887), a British civil engineer and father
of Robert Louis Stevenson. The use of a standard
screen allows temperatures to be compared
accurately with those measured in earlier years
and at different places.
The Stevenson screen was first introduced to
Australia in the 1880s and was installed
everywhere, with a few exceptions, by 1910. Prior
to this date, thermometers were located in
various types of shelter, as well as under verandas
and even in unheated rooms indoors. Because of
this lack of standardisation, many pre-1910
temperatures in Australia are not strictly
comparable with those measured after that date,
and therefore must be used with care in analyses
of climate change within Australia.”
From
http://www.ausstats.abs.gov.au/ausstats/subscriber.nsf/0/E8B0AEE832A75948CA256F8F00710605/$File/13010_2005.pdf
Thanks but all this stuff on screens avoids discussion on thermometers. Screens are no more or less than standard trees to provide the essential shade that stops bulb runaway in direct sunlight. In fact a verandah is quite good enough to do the job considering the other errors of the day.
Readers may be surprised but I rejected about 10% of all MIG thermometers found in labs over a long period of time. Although they mostly had simple errors from manufacturing they upset daily production quality checks. As previously stated, 1% of max was the usual target. For air temp it’s only one in a hundred degree C.
Chasing screen errors will not fix up the old records but extending the sweep will. Sorry no one bothered with calibration checks when air temp is uniform over wide areas.
First principle in obtaining standards is gradual development from endless comparisons.
Paul
It’s just there are numerous amateur pundits around the globe who think they are teaching their grandmother to suck eggs. If you don’t believe the measurements, look at the satellite records, for which a link has been provided, or the Arctic, permafrost and glaciers. It’s all there.
Looks to me like if nobody can credibly monitor “the temperature of the Earth,” they’re about as full of booshwah as people who claim to know what Earth’s ideal temperature is.
For now, though, it seems “the consensus” is that the best of all climates is the coldest. Which proves “the consensus” is booshwah.
Schiller: when was the last time you stepped outside with a thermometer and tried to develop your climate monitoring skill?
Schiller
you keep on making up stuff that people don’t think, then criticise it. You’re just wasting your time.
SJT – I refer you to the IPCC’s cherry picked peer reviewed science:
http://climatesci.colorado.edu/2007/06/20/documentation-of-ipcc-wg1-bias-by-roger-a-pielke-sr-and-dallas-staley-part-i/
and
http://climatesci.colorado.edu/2007/06/25/additional-evidence-on-the-bias-in-the-ipcc-wg1-report-on-the-assessment-of-near-surface-air-temperature-trends/