The importance of measuring temperature away from human influence

Many of you have followed my “how not to measure temperature” series showing many examples of the folly of thermometer placement in the USHCN network. But what about the rest of the world?

One of the most important things we look for is finding weather stations that are as far away as possible from human influence, so that they can resolve the “climate signal” unhindered and without need for retroactive data adjustments.

Weather stations on a remote mountaintop would seem to be a good candidate. Fred Harwood writes to me with one such example of a remote mountain station: Pointe Helbronner, Mont Blanc in the Alps, near the France/Italy border at 3462 meters high, about 11,358 feet.

Fred trekked to the top, to get this photo for us:

The temperature sensor is properly mounted inside the round louvered white radiation screen on the left side of the mast. The remote weather station also has a live webcam, as you can see in the photo (inverted glass dome) You can see a live webcam of this view here:

The latest image is a few days ago, so there may be a transmission problem.

The remoteness of such a station surely is impressive. Getting an accurate temperature measurement devoid of human influence would almost certainly be guaranteed, and we would not have to worry about nearby objects, people or buildings at such a remote location.

Well, maybe not….

And you can see more of the station and how it is situated (on the roof of a building) here in this YouTube video:

I suppose I’m not surprised. I’ll point out that this likely is not an official climate station. I don’t see it listed in GISS or NCDC. But the point remains, why go to all the trouble of a research quality temperature shield if you put the temperature sensor on a rooftop within reach of tourists? I’m sure that wood has an albedo of consequence and dozens of tourists also create a warm air pocket.

0 0 vote
Article Rating
Newest Most Voted
Inline Feedbacks
View all comments
July 15, 2008 11:52 am

That is great. I really wasn’t expecting that.
Exactly, what is the point? No kind of consequential research can be done from there.

Jeff Alberts
July 15, 2008 12:02 pm

Looks like one person was so overcome with grief at the placement, they just collapsed! They’ve been irrigated!

July 15, 2008 12:10 pm

[…] Tags: temperature Related Posts […]

Tony Edwards
July 15, 2008 12:14 pm

Anthony, off thread, but I don’t have an E-mail address for you.
Have you seen these two papers?
and this one from Professor Alexander
If you are aware of them, sorry for bothering you.
The second one seems to be especially interesting on several fronts.
Hope your house and family are OK and look forward to assisting you in your project.

July 15, 2008 12:27 pm

How about just 12 thermometers? 1 for each continent and ocean. At what point do we say, ok we have enough measuring stations?
If I had a “fever” all you would need is one thermometer to take my temperature. Is the planet not 1 body? Are the oceans not connected?
I hate seeing soooo much money wasted on this HYPE.
There is a war coming in the middle east, Israel and Iran. Compile that with the developing nations demand for fuel and you will see prices sky rocket. You think they’re high now?
I suspect that some alarmists are happy with the price of fuel? If so, that is just sick! We need to start drilling from our own resources AND continue expanding alternative energy (not corn based ethanol).

July 15, 2008 12:29 pm
July 15, 2008 12:29 pm

Bet it has an UHI adjustment in GISS…
REPLY: It does not appear in the GISS database that I could find. – Anthony

Mike C
July 15, 2008 12:39 pm

You really need to watch the video and see that it is a large structure with chimneys protruding from the roof. So much for accurate air temperature readings on calm nights. You might also notice the glaciers on the nearby mountains. On the other hand, I need to get there somehow, I wanna pull out the snowboard and hang ten on that run! =:O~

July 15, 2008 1:35 pm

Latest news; fewer then 100 fires burning n the State.
AP Lindlaw said onshore flow and rainstorms weren’t helping firefighters last Sunday. It was widely parroted by the reality challenged media.
Anthony, should have let me cuss that guy. Yahoo/AP/MSM axis of evil did an airbrush of their previous horsecarp article.
Truth is always a defence in libel cases.

The engineer
July 15, 2008 1:35 pm

The french will be really upset that you manage to put Mt. Blanc in Italy. Its in France and is the second highest point in europe (about 4.5 km).
Incredible amounts of snow at easter – I was there.

July 15, 2008 1:41 pm

maybe the mountain moved.
If you had a choice to be French, or to be Italian, which would you choose?

retired engineer
July 15, 2008 2:03 pm

French or Italian? They both have great food. Been there, done that.
How would you measure things at a remote mountain site? In winter, snow would cover everything. In summer, someone would swipe everything.
Satellite? How do they do initial calibration? Is/was it automatic, or do they base it on land measurements. Like GISS…
Another example of “We don’t know whats really happening.”
(without Anthony, we might have to believe what Al Gore is telling us.)

Peter Hearnden
July 15, 2008 2:25 pm

Lets see if I get this.
You’re saying that on the top of Europe’s highest mountain (15, 000 ft odd) there are going to be enough days when the wind so light that air can be heated up by tourists (yes, tourists) and the roof and rise up to materially effect the temperature record there?
Are you being serious?
REPLY: There are published World Meteorological Organization (WMO) standards for thermometer placement. This violates them, simple as that. Those standards exist for a reason, and the reason is so that the data is free of any potential bias and therefore free of arguments like this one over whether or not the data is or is not biased and to what degree. Adherence to standards negates the issue. Unfortunately we keep seeing this poor placement again and again. – Anthony

July 15, 2008 2:28 pm

Here are two SST anomaly data sets I pulled from ERSST.v2 on NOAA’s NOMADS system. I guess no one told Smith and Reynolds they needed to make adjustments for urbanization, etc.
Gulf of Mexico SST 1854 to Present
Long Island Sound SST 1854 to Present
Funny how current temperatures aren’t close to the peaks of earlier rises.

Richard deSousa
July 15, 2008 2:56 pm

It appears both France and Italy claim Mt. Blanc (French) or Monte Bianco (Italian)

July 15, 2008 4:14 pm

Mr Mark W asked: “If you had a choice to be French, or to be Italian, which would you choose?”
I’ve been saying for years, and will say until my liver takes its final gin, you can’t trust any country that makes runny cheese.
Donning the dunces cap for a moment, I have a question.
I can understand the general proposition that an accurate measurement of temperature trends requires three things:
(i) the measuring device itself must be reasonably accurate (hence one would prefer to use a properly calibrated thermometer rather than a home made device with graduations marked by fingernail scratches),
(ii) the device must be placed where it provides only relevant information not irrelevant information (hence, don’t put it next to the oven at the local bakery or in the ice cream freezer at the supermarket) and
(iii) the readings taken must themselves be directly comparable (measuring the temperature at 8am in January, 9am in February etc up to 7pm in December would be neither use nor ornament).
These basic points would mean that we could compare like with like. The more refined the calibration of the measuring device and the more readings that are used, the more accurate will be the calculation of averages.
I can also understand that it might be practically impossible to avoid all irrelevant variables because measuring devices have to be put somewhere (unless they are remote devices, as I believe satellite measuring devices to be).
It seems to me that all pre-satellite era temperature records are likely to have been affected to a greater or lesser degree by problems of (i) calibration, (ii) irrelevant influences and (iii) an insufficiency of data points. The first and third problems are perhaps more likely to affect older records and the second more likely to arise in records from industrialised times and industrialised places.
When analysing data to provide a trend, I presume that the statistical concept of “noise” can help to even-out the worst excesses of (i) and (ii).
My dunce’s question is whether it can also even-out (iii).
My initial reaction is that it probably can, provided there are sufficient readings from places unaffected by significant irrelevant influences to show two distinct trends – one in the affected places and another in the unaffected – from which the effect of the irrelevancies can itself be displayed.
Am I spouting sense or is it time for a little hemlock to be added to my evening gin?

just Cait
July 15, 2008 4:26 pm

ROFL The AGWers want so badly to have their claim that humans control the climate to be true that they’ll do anything to make it happen!

Michael Hauber
July 15, 2008 4:32 pm

To measure temperature theree, would you measure above the snow? Or above the rock? What if the snow melts or increases over time?
Or would it make better sense to measure above a more stable platform such as wood? Unless the wood platform rots/loses it paint over time. Fake grass matting perhaps?
How would you site a station in such an area?

Jeff Alberts
July 15, 2008 5:44 pm

I’ve been saying for years, and will say until my liver takes its final gin, you can’t trust any country that makes runny cheese.

Cheese Shop Owner: “It’s a bit runny, sir.”
Cheese Shop Customer: “I don’t care how excrementally runny it is, hand it over with all speed.”

Brent Matich
July 15, 2008 7:26 pm

Geez, you don’t think kids would be having fun using this equipment as target practice for snowball throwing. I’m still waiting for summer to arrive, it went down to 6 C a couple of nights ago and struggles to get to 20 C here. Calgary has the sunniest weather in Canada and can’t remember a more cloudy year in my 38 years here.
Brent in Calgary

July 15, 2008 10:01 pm

Peter Hearnden,
I would tend to agree with you about the tourists. That ignores the issue with the smoke stacks and general building waste heat that will not be consistent from day to day or season to season!!!
Of course, a group smoking cigars or pipes might actually cause a blip. ;>)
How about Al Gore as the tour guide. That would kick a few extra C on the reading!!

July 16, 2008 12:25 am

The swiss wetter data
i believe they have good wetterstations, all well maintaint
is in the centre off europe almost all mountains..
(beautifull country btw!)
(warmest year is 1994 orso)

July 16, 2008 12:26 am

Re, FatBigot’s (Wonderful name, Sir) query about the errors in measurements. Basically there are two types of error, random and systematic.
Random errors are centred on the “true” value (which you can never know) and roughly follow the “bell curve” distribution. They are otherwise known as “noise”, and can be dealt with by making your instruments as carefully as possible and taking a large number of independent measurements and calculating the mean value.
The big problem with nearly all practical measurements is the systematic errors. These all tend to go the same way, and thus cannot be averaged out. As an example, if you wanted an accurate measurement of the temperature of a pot of boiling water using a hand-held thermometer (DON’T try this at home, it hurts!) you would have to remove the thermometer from the water to read it, and as soon as it leaves the pot it starts to cool down. Therefore, all your readings will be lower than true by varying amounts. Taking the mean here is obviously pointless.
Unfortunately, all the examples FatBigot gives are systematic. Type 1 could, due to a calibration error, have its scale shifted up by half a degree, Type 2 (which is the main point at issue here) be near something that was persistently warmer or cooler than the “true” temperature, and Type 3 will, as he implies, show a totally fictitious periodicity.
We have to be very, very careful about systematic errors, in particular because the buggers are adept at hiding among the random ones and not being allowed for, like UHI.

The engineer
July 16, 2008 12:34 am

This report seem apt and amusing – Global Warming makes Mount Blanc

The engineer
July 16, 2008 12:36 am

Tres amusant – Global warming makes Mount Blanc grow !

Phil M
July 16, 2008 2:19 am

Way off topic, but you are always on about the influence of volcanos on the Antarctic Sea Ice
I think this image is a good illustation of this:
– it would be interesting to see the positions of the volcanos relative to the ice-free areas…

July 16, 2008 2:51 am

The first image had me cheering – but I did wonder at how long the cable might be to power it considering the sensors are so close to buildings because of this issue.
However subsequent images showed me the error of my perception. Actually, what do they use – a generator or solar ?
I too hope that your fire hazards are resolved, Anthony.

July 16, 2008 4:07 am

Actually, the snowfields are troublesome! When you climb high mountains on a hot day, even cold but rather slow winds, sweeping up or down from large snow or ice areas can be very cold indeed and can force you from a state of transpiration to a state of high demand for protective clothings!

July 16, 2008 4:11 am

Can’t help but to smile. I had a barbeque at this station a few years back. The wine, the entrecôte and the lettuce was excellent. The temperature reading at that particular day may not have been excellent…

An Inquirer
July 16, 2008 5:25 am

Okay, with the possible exception of satellite data, it is really tough to get high quality & consistent readings of historical temperatures. However, how much do care about the specific average temperature, up or down a degree or a small fraction? Are we not more concerned about temperature’s effect on climate? And can we get a handle on that through means other than specific temperature readings? I have often thought that the length of the growing season would be a good indication of trends — at least for places that winter visits. I once did a google search for a comprehensive and understandable analysis of growing season lengths, I came up empty. Any thoughts on this?

Peter Hearnden
July 16, 2008 5:47 am

Anthony, you say: “There are published World Meteorological Organization (WMO) standards for thermometer placement. This violates them, simple as that. Those standards exist for a reason, and the reason is so that the data is free of any potential bias and therefore free of arguments like this one over whether or not the data is or is not biased and to what degree. Adherence to standards negates the issue. Unfortunately we keep seeing this poor placement again and again. – Anthony
But you then say:
I’ll point out that this likely is not an official climate station. I don’t see it listed in GISS or NCDC.
So, you point out there are (rightly) standards for official stations, and then you complain a non standard station is non standard??? Huh??? Run that by me again will you 😉
REPLY: The standard applies to stations whether they are used for synoptic forecast analysis (such a s a standard USA COOP station)or if it has been chosen to be a climate station. I’ll point out that the USHCN network was hand picked from what was considered the “best” USA COOP stations in the early 1990’s. Unfortunately they only examined the length of record and number of station moves as the criteria for making it USHCN. GISS and HadCRUT also have not examined station instrument placement characteristics.
So having instrument placement right in the first place is key, since you never know what someone will do with the data later. The standards exist to ensure accurate readings, and should be followed.

July 16, 2008 5:54 am

Interesting, thnx for the post.

Pierre Gosselin
July 16, 2008 8:08 am

Quality has a price.
I bet it costs the hell to do maintenace on that station.
Still, better a few quality stations and quality data, then a load of rubbish.

The engineer
July 16, 2008 8:15 am
Tom Richard
July 16, 2008 9:25 am

Hi Anthony,
Did you see this blurb from NOAA? Any truth to it or is it another example of choosing data to get an end result? Thanks for any info.

Gary Gulrud
July 16, 2008 9:31 am

OT from sayanything:
Seems ethanol may not be economically viable at yesterday’s corn prices. Who da thunk.

July 16, 2008 11:16 am

An Inquirer,
ecos are complaining bout this continually. Of course, this varies from area to area almost on a yearly basis. In general, the growing season HAS lengthened in the last 30 years. Of course, more CO2 helps plants grow better in harsher conditions, so, all that lengthened season is not necessarily temp based.
A better idea is altitude (which is also affected by co2) and areas of historic farming. Areas like Greenland, Alps, and others around the world show that, while we have currently warmed, the Medieval Optimum and Roman Optimum were both warmer.
This site references a load of geo and bio type current and paleo studies:

Leon Brozyna
July 16, 2008 11:26 am

This newest image from Icecap is just too good to pass up:
Is this how anyone can keep on saying their forecast is accurate?

July 16, 2008 11:28 am

Jerry (00:26:37) :
Re, FatBigot’s (Wonderful name, Sir) query about the errors in measurements. Basically there are two types of error, random and systematic…
Thank you kindly, Mr Jerry. I knew I was right to don the dunce’s cap. It will remain atop my balding pate for the foreseeable future. But fear not, my ignorance will be no handicap to my bigotry.

July 16, 2008 11:28 am

The more I research global warming, the more I believe its a massive and carefully orchestrated con job.
Thanks for the post.

July 16, 2008 11:35 am

Brent from Calgary – I’ve only had one previous summer in Saskatchewan, but this one defintely feels cooler – only one, maybe two days with a temperature in excess (and not by much) of 30 C so far. But the farmers that have farmed here since the first farms on the prairies told me the same – this is a cool summer (so far).

July 16, 2008 12:10 pm

Re The Engineer’s post
I have read a couple of Stephen Wilde’s articles and this one is singularly clear. I posted a couple of days ago asking for comments from the scientific community on his article suggesting that with a negative PDO and weak TSI we may expect to see the AGW theory put to bed and it would be great to hear what people think about this latest article “Greenhouse Confusion Resolved”. Perhaps a grand title given the complexity of the climate but certainly one of the most lucid articles I have read. Moreover it seemed to fit with much of what I have read on this blog and over at CA.

John in Oregon
July 16, 2008 12:12 pm

The weather station isnt the only problem with this installation.
Notice all those funny looking things mounted all around the weather station? Those are radio and TV transmitting antennas. So now we have two more issues.
First. How accurate are the weather station instruments in a high Radio Frequency environment? Have they even been tested for an RF environment?
Second. If this was the United States and I as a site manager allowed the general public this close to transmitting antennas I would get a severe reprimand from the FCC. Reprimand as in, fine and loss of license.

July 16, 2008 12:46 pm

Anthony, this is OT, but you might find this article at CO2Sceptics by Stephen Wilde to be interesting GREENHOUSE CONFUSION RESOLVED

The engineer
July 16, 2008 12:58 pm

To Paul H.
My thoughts exactly.

Gary Gulrud
July 16, 2008 1:26 pm

Leon Brozyna:
How can they keep their job? And they would not have put any real confidence in the date of minimum.
The maximum count is what all the predictions are about. So the abyss of their embarassment waits until, say early 2013 when that peak is reached.
The Russians and Clilverd, perhaps an continental Indian or two, not all of whom are recognized as making ‘scientifically verifiable’ predictions, are the only potential winners left standing.

July 16, 2008 2:27 pm

Somehow, all this fight regarding AGW vs long term climatic cycles reminds of another war within the scientific community – that one was about math –
in the late 18xx years was a fight between the supporters of the quaternion algebra vs the complex algebra – the supporters of the complex algebra won.
I allways had the impression that the quaternion expression of Maxwell’s law
is better to understand than the complex expression .
After nearly 150 years of the wrong way gone, the quaternian algebra is preferred again (since about 1990)
From Wikipedia: ( when it’s about math, wikipedia is OK so far)
Quaternions are often used in computer graphics (and associated geometric analysis) to represent rotations (see quaternions and spatial rotation) and orientations of objects in three-dimensional space. Certain fractals can plot in quaternion coordinates. They are smaller than other representations such as matrices, and operations on them such as composition can be computed more efficiently. Quaternions also see use in control theory, signal processing, attitude control, physics, bioinformatics (see: Root mean square deviation (bioinformatics)), and orbital mechanics, mainly for representing rotations/orientations in three dimensions. For example, it is common for spacecraft attitude-control systems to be commanded in terms of quaternions, which are also used to telemeter their current attitude. The rationale is that combining many quaternion transformations is more numerically stable than combining many matrix transformations. There is also less overhead in using quaternions compared to using rotation matrices, because a quaternion has only four components instead of nine, so the multiplication algorithms to combine successive rotations are faster, and the result is much easier to renormalize afterwards.

July 16, 2008 5:13 pm

Hi, I’m just a dog guy, but is it even possible to find some small corner of our little planet, that hasn’t been influenced by humans ?
Obviously, the more remote a place is, the less influence we will have on the findings. But isn’t weather global. Meaning, although major metropolitan areas can be thousands of miles away, they still must have some minute influence on that region.
I’m serious, can somebody explain this to me. What effect if any, do we have on these remote areas.
Thank you

July 16, 2008 9:10 pm

John in Oregon (12:12:39) :
“Notice all those funny looking things mounted all around the weather station? Those are radio and TV transmitting antennas.”
Are you sure about that? The high power antennae I’ve seen are more like those at . What I see in the photos are microwave links (the round dish reflectors), directional UHF antennas (the bar with the vertical elements) which are likely for communication with the building, and maybe a lowish power omnidirectional antenna like those used for police and fire radios in the USA.
The transmitting antennae likely aren’t putting out much power, the receiving ones aren’t putting out much of anything.
There’s a fire tower a few miles from my home with a bigger cluster of antennae that are used by public radio to link to remote transmitters, transmit seismic data from a nearby seismometer, relay telephone traffic, etc. There’s enough to significantly interfere with the view from just below the lookout cabin on top.

retired engineer
July 17, 2008 8:43 am

In working with wireless instrumentation, even a a few milliwatts, we saw disruption in some analog circuits, usually as an increase in DC offset. The gurus at ADI explained that this was due to geometry, fabrication techniques, and black magic. The last part made it hard to determine what would happen under all circumstances. You don’t need a lot of RF power to mess up a sensitive measurement. Nor can you easily predict the outcome.
Answer: Don’t put radios near low level sensors.

%d bloggers like this: