In search of the perfect thermometer.


Lon Glazner, a fellow blogger and local electronics engineer made some comments about my post on the NASA/CSU study on California temperatures. Well that got me started…so below are Lon’s comments and my reply along with a fun technical challenge. For those of you that read this blog, but disagree with my views, I invite you to read this carefully.


You make a number of good points. Particularly in the fact that the writers may have applied changes in urban temperature measurements over large regions for graphical impact.

As someone who has designed and built electronic temperature sensors I have certain concerns about the data itself.

Unless temperature sensors are regularly calibrated I think it is unreasonable to expect accuracy of greater than a couple of degrees.

Even some that are calibrated may not have good accuracy. The LM34 which is a commonly used semiconductor for measuring temperature is +/-2 degrees F. This is pretty typical of analog or digital semconductor sensors. The temperature error for this part is also non-linear, and so it’s not a simple offset that you have to account for during data collection. Furthermore, there are lots of additional errors that can creep into a temperature measuring device beyond the sensor itself.

One could argue that numerical analysis done on data points would tease out errors. But if a scientist doesn’t know the exact accuracy of a temperature sensor then they couldn’t account for errors in their system.

Some of the temperature sensing stations may be very accurate and regularly calibrated. But maybe they’re not?

I have a hard time trusting that the data is accurate to the level of identifying 1 or 2 degree changes over decades. This is especially true since the techniques of making these measurements have changes over that time frame.


Lon, thank you for the comments. FINALLY somebody who understands the kind of biases that creep into temperature measurements!

I’m innately familiar with National Semi’s LM34 and it’s accuracy problems. One of my early jobs at my university as a research assistant was to create remote electronic weather stations. I soon learned how inaccurate many electronic devices can be in temperature measurement.

The problem with the National Weather Service temperature data sets (and world data sets too) is that they are full of biases and errors that I’m not sure have been accurately accounted for. People such as Jim Price, from CSUC who is on the IPCC say they have been, yet nobody has shown me any hard evidence of such. I’d be a lot less skeptical if I could see how the IPCC accounted for temperature measurement biases. But they won’t share.

Some people that I try to explain this to accuse me of splitting hairs. But these bias problems in temperature measurement are quite real.

What works against my arguments about the difficulty in getting accurate temperature records is the everyday simplicity of temperature and its common measurement. We live by temperature, we have it reported constantly, we all have thermometers at home, we measure our childrens fevers with thermometers, we barbeque with thermometers.

Measuring temperature is easy right? You just stick the thermometer in whatever gas, liquid, or solid you want to measure the temperature of and voila’ there it is. People tend to think of thermometers as perfect devices. Some very expensive calibrated thermometers, are close to perfect, especially when taking measurements in a closed system, like a fermenatation vat at Sierra Nevada.

But in an open system in our atmosphere, there are many many more biases that can affect the measurement within a few inches or feet of the thermometer. Here’s just a few:

– Reflected sunlight from nearby building or objects

– Re-radiated infrared from nearby cement or asphalt surfaces or the ground itself (which is why airports make terrible places for temperature measurement)

– The structure that the thermometer is mounted to, can conduct heat to the thermometer

Now add to that:

– Accuracy of the thermometer itself

– Linearity of the thermometer over its measurement range

– Long term repeatability of the thermometer’s accuracy

– Long term repeatability of the thermometer’s linearity

And then we have urban effects such as:

– Localized vegetatation removal or addition over time

– Localized building changes over time

– Localized asphalt or concrete surfaces addition or removal

And finally within the global temperature records data set we find instances of:

– Changing the location of the weather station and/or its thermometer

– Changing the thermometer itself at some point – i.e. repair/replace

– Changing the thermometer type, from mercury, to electronic (thats been done at thousands of weather stations worldwide)

– Variations in temperature measurement devices from country to country, even though the World Meteorological Organization has specifications, they are not always followed.

– Changes in thermometer shelter, different types of paint over time, all which have different absorptive and reflective properties.

– Changes in the observer recording the temperature, some may round up, others round down numbers. BTW for about 75 years, all temperature records were manually recorded.

Ok with all these biases and possible errors that you have to account for to make long term temperature measurement reflect the true temperature of the location, can you be absolutely sure of the data integrity? Especially when you are looking for trends that may be 1 degree or less over 50-100 years? I can tell you that I’ve looked at these climatological data sets, and NONE of them come with a calibration record for the thermometer, or even a description of the make/model used at that location. There are notations in the records that say things like “station relocated to accomodate construction” or “thermometer replaced” which can give clues to the data integrity possibly changing but the climate researcher is left to make a judgement call on the viability of the data without anything to gauge the sensor or its local environment.

Or lets try a thought experiment Lon, you’ve been commissioned by the IPCC to make a new thermometer for use around the world at climate measurement stations. As an electrical engineer, could you design an air temperature thermometer that is:

– Linear to within 0.1% over a temperature range of -20F to 120F

– Accurate to within 0.1 degree F over that same range

– Repeatable in linearity and accuracy defined above for a period of 20 years. Or even 10 years.

– Identical withing the specs above, so that if one fails, it can be immediately swapped with another one from parts stock with no worry about introducing bias

Ok there’s your challenge. Could you do it?

0 0 votes
Article Rating
Newest Most Voted
Inline Feedbacks
View all comments
April 1, 2007 8:29 am

People should know that temperature sensors are a lot like clocks. Back in the old days they were mechanical, and today they are electrical. Different types of clocks have different types of errors and accuracies. Most of the errors and accuracies are not documented well(if at all). But nobody needs a clock that is highly accurate year after year, and they typically aren’t. The same is true of temperature sensors. A few degree here or there didn’t used to matter.
Temperature sensors were never designed to provide accuracy of measurement to the levels being reported. Without regular calibration, and standardized or known errors these measurements cannot be highly accurate.
Using the clock analogy, if you and your friend’s watches are off by 1 minute how do you know which one is correct? You might call “Time”-POPCORN, which is essentially calibrating your clock? That might make both watches accurate today, but which one is more accurate in 10 years? Now what if you wanted to make sure your watch had been accurate 50 years ago? How would you calibrate your watch to be accurate to half a second if you knew calling “Time” was accurate, but only had a resolution of 10 seconds?
Citing a 50 year old temperature measurement as accurate to the degree is pretty far-fetched. That is espeacially true if the calibration and accuracy tests for each sensor are not provided. However, there may be other indicators of a warming trend like glaciers retreating that support certain measurements.
Almost anything can be “done”. But the specifications you gave would hardly be affordable IF they can be accomplished.
The accuracy could be determined by calibrating the sensor against a known standard. Usually the “standard” is a physical property based measurement like the triple point of water. Some of the retail “standards” can be seen below.
The linearity would be determined through characterization of the sensor over the temperature range through an extensive calibration using multiple standards. The sensor itself would not be linear in its error, but a look-up table would be used in the sensor microcontroller to adjust for errors.
Ensuring accuracy over time would not be possible wthout regular and expensive calibrations.
To be realistic I wouldn’t take a contract to design the temperature sensor you described.

Art Roberts
April 2, 2007 5:32 am

Very intriquing comments. I have a home weather station and the one-wire sensor that I use for outside temp is accurate to +/- 2 degrees centigrade – but it reads out the temperature in tenths. So 50.6 degrees looks really accurate, doesn’t it? I always thought it was just funny……

Walter E Wallis
April 2, 2007 7:52 am

I see now why your response to my question about how to determine an average temperature.
In the 50’s I was a test engineer reporting the min and max for Trona, California. While I was reasonably good at what I did, I sure wouldn’t justify changing the way the world is run based on those readings.

April 2, 2007 9:01 am

To answer, the clock question. You get one clock that you can measure against others. You try and get all of the unknowns of the one clock figured out, and then can always measure gain/loss from the rest that are based on that clock. Of course this only works if you can make a valid comparision, which you can’t for the temperature measurements.
Well if you want to get a full time solution then just use SEL clocks.

Marek Suchenek
June 18, 2007 11:02 am

Thank you for the photos. They just illustrate the fact that the global warming lobby didn’t bother to define what the “average earth temperature” is, never mind any specific indication how to accurately measure it.
It’s mind boggling that so many otherwise intelligent and educated people bought that “average earth temperature” fiction.

June 18, 2007 11:08 am

What’s interesting is that I live in Western North Carolina and we are experienincing global cooling for the last 50 years. I wonder if it has anything to do with the fact that the National Climactic lab is nearby and they do a better job at keeping their temperature sensors up to date. One is near me so I plan on checking it out

June 18, 2007 11:34 am

Looks like Drudge has sent you some traffic. Good job!

June 18, 2007 11:59 am

Being an avid vegetable gardener (and retired research engineer), I woke one morning to frost in the front yard (faces east). But there were no alarms on the weather station indicating 32F. I immediately suspected temp sensor accuracy, and ran into the south yard still in pajamas with a Fluke temp measuring device. I found that yes, the temp sensor on the station agreed with the Fluke hand held sensor. It was 34F degrees on the south side of the house, but 31.5F on the east side. The difference: a two lane road — some 30 feet away from the sensor — is on the south side.
And, unfortunately, the effects on calibration vary with temperature, humidity, traffic, and so forth. So, if you can live with maybe 3 degrees accuracy, no problem.
I had all along suspected that the temps were “urban warming” or related to modern conveniences.
In Atlanta, I’ve actually seen a rainstorm vanish while crossing I-85 at the intersection of I-285, only to reconstruct on the other side. It went from red to green to red again on the doppler radar. Sixteen lanes of asphalt really does modify weather!

william kinder
June 18, 2007 12:29 pm

as an engineer, and neither a warmer, a cooler, or a don’t worrier, all of you have mush between the ears. the only way to get a true measurement of global average temp is to blanket the planet with sensors, land and sea.
the type of sensor does not matter, mercury, T/C, RTD, melting ice cubes and a stopwatch. what does matter is consistency of results and long term measurements to show a trend, if any.
quit griping about stated accuracy of sensors. it is statistical anyway. what they really mean when they say +/- 2 deg is that there is a 95% probability that the REAL temp is within 2 deg of the displayed number(if they use 2 sigma standards. 99.7% for 3 sigma). this particular error cannot be corrected for and is where the long term trend comes into play, which tend to cancel this type of error.
the errors resulting from placement and calibration are a different type of error which tends to skew the data in one direction or another. this is where many data points are important, or a completely consistent single data point per period.
if it is a calibration problem, all data points between calibrations need to be adjusted.
if it is positioning then many points per day are necessary, or just a single point taken at a fixed time. this tends to equalize the error and make it consistent from period to period.

Wallace Milam
June 18, 2007 1:02 pm

A little ombudsmanship never hurt anyone. I don’t know if global warming is occurring or not–neither does anyone else, I suspect. It certainly is the flavor of the month, though. I’m a Tennessean who has known Al Gore back before he invented the Internet, went to the same university in fact. As we say in the South, “Bless his heart; he means well.” Gore kinda stands as a perfect illustration of Ambrose Bierce’s definition of the word “positive,” “to be mistaken at the top of one’s voice.”
Your examination of the data-gathering stations reminds me of my late father’s favorite saying (He had a 4th grade education.): “A fool with a tool is still a fool.”

June 18, 2007 1:30 pm

Your research of the Historical Climatology Network sites in your area, and the deplorable conditions in which the sites are located, has piqued my interest. How can I go about finding the locations of the sensors in my area? I’m in the St. Louis metropolitan area.
***REPLY We’ll have the website back up in about 24-48 hours, just check back then. It was overwhelmed by traffic today.

Paul F
June 18, 2007 2:01 pm

You’re making a heck of a lot of sense with your thinking here, good job! I’m definately in the ‘skeptic’ camp, and I see all too clearly the political motivations of the global warming crowd. I truly hope mother nature will cooperate and give us a nice cooling trend so we can watch the ‘warming’ crowd scramble for explainations as to why capitalism is freezing the planet! Thanks ~ Paul

June 18, 2007 2:04 pm

Ditto Jill’s comments. I would love to go and inspect the weather stations here in SE Minnesota. I’ve been scouring the web today to find the locations, but alas, no luck.

Anthony Watts
June 18, 2007 2:07 pm

There is a locator available at but the site was overwhelmed and is offline at the moment.
When its back up there will be a step by step procedure to locate and survey the station in an orderly way.

June 18, 2007 2:51 pm

I knew it! I have been observing this trend for years. I work in agriculture, and I run crews in groves in Florida. Last summer was the relative coolest summer that I remembersince 1983 when I started working. Now mind you, it is always hot, But I was amazed last year when our, in field minimum temp thermometers that we use in freeze analysis rarely registered in the 90s. These are very accurate thermometers because we calibrate them every year. You could just feel that it was not as hot as usual. Sadly the statistitions showed the temps almost every day to be hotter than it was in the field. I just hate the weather channel. Also last week in the southern Applacians, the actual temp in the city I was in was generally 5 to 10 degrees cooler than the weather channel predicted and reported. For 5 days they misreported the temp. We are being duped. goto for some interesting “global warming” myth busting

June 18, 2007 5:35 pm

Watts is biased. He’s only in this to make money for himself.
***REPLY Hmmm interesting take. Please show me where I’m accepting money or donations on the website. I’m financing this all myself, so please by all means if you think I’ve made money, please show me where.

June 19, 2007 6:00 am

As a retired electrical engineer that designed experimental systems for an environmental physics group at a university, I can attest to the wildly inaccurate readings obtained by electronic sensors. Not only temperature, but other parameters, especially humidity, are generally inaccurate. Even systems designed for balloon sonds that only have to operate for a couple of hours cannot hold their calibration for that period of time under changing conditions.

June 19, 2007 12:44 pm

There is a digital solution that is acurate +/- 0.5 degree C.
The DS18S20
communicates over a 1-Wire bus that by definition requires only one data line (and ground) for communication with a central microprocessor. It has an operating temperature range of –55°C to +125°C and is accurate to 0.5°C over the range of –10°C to +85°C.

June 19, 2007 6:36 pm

Ive often been curious as to the accuracy of the thermometers myself. As I work in a trade where I often see the latest in high tech digital thermometers costing hundreds of dollars that are 98% accurate +/- 2-3 degrees.
Now I always figured G8 type countries like the U.S., Canada, France, etc as well as major world Universities in the 21 Century would be using the most accurate instruments. But Ive often wondered what the thermometers in 3rd world countries were like, not to mention the Temps taken from around the world since the late 19th century. I wonder how accurate those thermometers were.
Since Golbal Warming is based on how the Entire Earths avg temp has risen 2 or 3 drgrees over the past 120 years or so.
Perhaps much of that is based on more accurate thermometers, which we know arent perfect in the U.S. in the 21st Century!

June 19, 2007 8:20 pm

Those may be the wrong ways to measure temperature, but here in Oshawa, Ontario, our weather, from the news stations, is gathered from Buttonville Airport, which is 25miles away, through 5 different cities! (All part of the Greater Toronto Area), its crazy, because we have our own airport, which is quite large (capable of handling jumbo jets in case of emergency).

Robert Stathers
June 19, 2007 11:12 pm

Back in the days when air temperatures were measured with maximum and minimum recording thermometers, the daily average temperature was calculated from the max and min readings. When strip chart recorders came along, the daily average was usually calculated by simply finding the daily max and min from the chart.
For example, if the daily max was 20C and the min was 10 C, the avg = (max-min)/2 = 15 C. Since the advent of dataloggers in the 1980’s, the daily average temperature is calculated by averaging hundreds of sampled temperatures taken throughout the day.
This has also lead to large errors, because if you think about integrating all those samples over an 18 hour summer day, you’d likely get a higher average than simply taking the average of the daily max and min.
Something that appears to be so simple as a daily average temperature record that has been measured for more than 100 years is full of many biases, not just heat island related or sensor related.

June 20, 2007 5:05 am

Great work Anthony- be prepared for an assualt of posers like “jesse”
In the hope of being useful, I found this mapping page for station sites:
if you click on “search by map” under “data and products” it is fairly useful
*** REPLY Thanks I’ll add it to the link on

June 20, 2007 6:24 am

I like the depth of investigation with regard to each individual meteo station. And certainly agree with the conclusions on accuracy, etc.
However, all of those details mentioned at each and every station would affect only the absolute values they are reading while the trends would not be affected. That is to say, given so many of them have contributed to the NASA study, and given these are data for over 100 years, the upward trend (of the means) must be a reality. It is of no real concern that the actual values of these means could be biased (up or down, dependent on the actual situations) locally.
This upward trend is also corroborated by the diminishing glaciers all over the planet, and a lot more; temporary cooler regions as reported here are isolated phenomena, and are normally to be expected, given the Earth’s size as an eco-system.
Whether this warming up is man-made or not can not be judged upon seriously at this time, as there is not enough truly scientific evidence. And it shouldn’t, as it is a distant secondary issue.
The global warming could be coming as well e.g. from enhanced solar activity, a fraction higher than some hundred years back. It could be that this fraction higher is still within the accuracy we measure it, it could be that it has blipped up a hundred years ago but the effect has been accumulating to only be visible strating as of recently, it could be else. And then again: frankly, the change of the atmospheric adsorption characteristics can not and should not at all be discounted.
The real problem is, it will probably take us until the Amazon jungle is annihilated before we could agree we needed it. Maybe not for lowering the solar radiation (or enhancing its absorption) but to escape in its shadows… I can tell you, in Europe we are already looking for darker shade of pale way more than 40 years ago.

Ralph A. Zingaro
June 20, 2007 8:27 am

How do I get a copy of this e mail? The printed copy off of the computer is too small to read. The arguments make so much sense and are documented. My mailing addrress is: Department of Chemistry, Texas A&M Unibversity College Station, TX 77843
Thank you
*** REPLY: try cut and paste into Mircosoft Word

June 20, 2007 10:17 am

Thanks for your work. The Chicken Littles out there like to listen to journalist too much. After all, they must know what they are talking about.
Guess what most undeclared second year students at UGA pick for a major.
Also, one of my favorite questions to ask the Chicken Littles is this: Columbia, SC used to be on the coast long before man ever walked on the Earth. There are fossils of sea creatures to prove it. Knowing this, how can you consider the current coast line the correct one?

June 21, 2007 2:47 pm

I agree that measuring temperature is a tricky thing, and placing sensors near parking lots, a/c units, and other natural or man-made structures could lead to potential errors. Now we just need to tell the glaciers that the measurements were wrong so they can stop melting!

Rick Fanning
June 28, 2007 6:00 am

Just wanted to comment on the receding glacier issue. Two points: First, not all glaciers are receding and a good number of them are advancing. Second, glaciers have been receding overall world-wide since the end of the last ice age. That’s how it ended – the glaciers receded. When the majority of the glaciers start to advance is when we need to be concerned.

Joe Elebracht
June 30, 2007 11:19 am

Three things, First, your link for weather station standards does not seem to be the right link. It is on a related topic, but says nothing about the 100′ rule, for example. My apologies if I missed something there.
Second, When doing station audits, I think there would be some value in getting a picture or assessment of the roof of the Stephenson shelter, where feasible (one can just hold a camera up there if it is above eye level). This is probably the hardest part to keep painted.
Three, I think that it would be a fair question to ask the supervisor of the station if it is ever in the shade. This might not matter for max/min temperatures but could be relevant for other averaging.
***REPLY FROM ANTHONY*** There is a link to a PDF document with specs on the NOAA page I reference about siting

July 21, 2007 8:47 am

You are doing a wonderful job!
Thank you for injecting a much-needed note of reality into green hysteria. Their desperation to manufacture a problem they can use to increase taxes and regulations is more blatantly obvious than ever. Thanks, again.

Tim Bartlett
August 5, 2007 1:14 pm

I live in Palm Desert CA and my temp readings match AccuWeather’s until temps reach above 100 then they post too high readings which increase even more as temp rises suggesting that thier observation sites may be impacted by re-radiated infrared. The sites are located in the middle of acres of asphalt runways at Palm Springs and Thermal Airports. Each City in the area is given a temperature based on historical data relative to the airports’ readings. Gee…do you think the posted temps may not accurately depict the actual temps, (unless you happen to live in the middle of the runways at the airports)?

August 24, 2007 12:01 am

I always thought it was the opposite.
Our weather stations may report a high in town of 92F but our thermometers out here in the country, near no asphalt or the like, read 102 to 105. It is horrifyingly hot to feel.
SO I figured the weather stations were just reporting the temps below what they really are to keep public panic down, as usual. Now you say it’s the opposite, it isn’t getting warmer on average?
I understand people will use global warming as an agenda to make money and gain power. This ticks me off, it’s taking the focus off of the real environmental issues at hand. All I hear now is a lot of hot air from both sides.
People like you will say it’s not getting warmer on average because your money is already tied up in the things that have screwed things up to begin with. Nice try. Both sides aggravate me.
I have to listen a geologist elder in my family yap on about the temps not really getting any warmer and how everything is fine, yada yada yada. Guess where his money is tied up? Coal, gas, mining, oil. Surprise surprise. *sigh*

Verified by MonsterInsights