I’ve recently updated the www.surfacestations.org website with the latest surveys and numbers. We have 534 stations surveyed. Here is where we stand now with USHCN station surveys:
click for a larger image
Climate Reference Network Rating Guide – adopted from NCDC Climate Reference Network Handbook, 2002, specifications for siting (section 2.2.1) of NOAA’s new Climate Reference Network:
Class 1 – Flat and horizontal ground surrounded by a clear surface with a slope below 1/3 (<19deg). Grass/low vegetation ground cover <10 centimeters high. Sensors located at least 100 meters from artificial heating or reflecting surfaces, such as buildings, concrete surfaces, and parking lots. Far from large bodies of water, except if it is representative of the area, and then located at least 100 meters away. No shading when the sun elevation >3 degrees.
Class 2 – Same as Class 1 with the following differences. Surrounding Vegetation <25 centimeters. No artificial heating sources within 30m. No shading for a sun elevation >5deg.
Class 3 (error ~1C) – Same as Class 2, except no artificial heating sources within 10 meters.
Class 4 (error >~= 2C) – Artificial heating sources <10 meters.
Class 5 (error >~= 5C) – Temperature sensor located next to/above an artificial heating source, such a building, roof top, parking lot, or concrete surface.”
During the next week, I plan to add a number of stations during my road trip, and Russ Steele is getting many also on his 3 month cross country road trip through the USA by mobile home.
If you are planning trips this summer, why not check out which stations have been surveyed here and see if any at the bottom of the list that have not been surveyed will be near your travels? We still have over 600 stations to go, and your help is needed!
Anthony,
Just a thought – If you can’t get enough USCHN sites with CRN1 or CRN2 classifications to set up a credible temp grid across the U.S., what if you use some GISS sites to fill in the gaps and find CRN1 and CRN2 sites and have your volunteers evaluate the sites if it’s easier for them than to hunt down than the USCHN sites? Does that corrupt the whole surfacestations effort?
REPLY: Interesting idea, I’ll give it thought
Tom: Yes, I know about the effects of water. But CRN1 also stipulates “unless it is representative of the area”. That would have to be determined by site. The SHAP adjustment would be expected to adjust for the offset if a station is moved. (But all too often, SHAP fails to do this.)
Finally, When I compared the class 5 sites to the Class 1 & 2 sites I found differences in trend. Thats what you look for.
Yes, LaDochy et al (12/2007) points this out.
But you dont see anything close to 5C differences in absolute temps. Why? because the bias signal is not constant
Only T-Max and T-Min are accounted for. Other times are not taken into account at all. So only Max and Min times are relevant.
Lets just take the effect of concrete. When a site is over concrete the daily TMAX may actually get depressed because the concrete is a great heat sink.
But when the sun goes down, then the TMIN spikes.
La Dochy claims the trend (sic) in urbanized areas concrete does indeed exaggerate T-Max trend (sic), but far more so at T-Min.
The twofold problem is to factor out the offset so it will not become “part of the trend”, and the heat sink trend change as well.
There are large differences in albedo between different paving materials, which change substantially as the materials weather over time. Fresh asphalt has a very low albedo, Fresh concrete made from portland cement has a very high albedo.
Local effects on temperatures are of the order of 5C between different paving materials. That’s an order of magnitude larger than the claimed warming.
http://www.pavement.com/Downloads/RT/RT3.05.pdf
I’d say that the presence of a nearby manmade surface, should automatically exclude a station from consideration in determining temperature trends.
Bah! looks like somebody already surveyed your Fayetteville Arkansas station, i was pondering doing that. Ive passed it probably a dozen times going to work and school. I swear you can see it from the interstate. What kept stopping me was the whole its behind a massive barbed wire fence that surrounds the agriculture test facility they have going on there. They probably wouldnt like me poking around there. And hunting somebody down to ask is such a pain…haha.
The BBC caving to environmentalists? Come on, they have a fine reputation of balanced reporting. Like CBS.
There isnt a qualitative study to back up the bias figures, so that is what we are undertaking. I take the ratings exactly as Dr. LeRoy proposed them, as estimates.
The CRN “estimates” are based on the results of a French study. The Rev has posted the link, but, foolishly, I have not yet captured it.
The CRN handbook was approved in late 2002, so it would not be a particularly recent study.
An up to date look would definitely seem in order on light of the surface stations flap.
Note two post-1980 factors:
–The MMTS conversion. That created a majority of the CRN4 violations. (Many may well have been CRN3 prior to that, but it is usually impossible to say.)
–The air conditioning revolution (a favorite feature in surface station pics).
And of course there was a huge amount of exurban creep during this time that swallowed up who knows how many of the stations.
A portion of those offsets will probably have found their way into the record.
Also consider that heat sinks exaggerate trends, either up or down (LaDochy 12/2007 notes how the latest warm trend was exaggerated in California), and there was a genuine warming trend from 1980 – 1998.
LaDochy also says UHI exaggerates T-Max, but nit as much as T-Min.
Those are the “estimated” (NOAA’s phraseology) MINIMUM error. (All “plusses” are excluded from the calculation.)
again, this is total nonsense.
the source of the calssification can be found here:
http://tinyurl.com/4zcky9
in Leroy’s article “Classification d’un site”
if Leroy was speaking about a minimum error, a site above a parking space (obviously cat V) would show a +5° error even when the surface was covered with snow. this doesnt make any sense.
even regulars here do NOT understand the classification scheme. this happens, because the presentation is misleading.
hm.
REPLY: You are confusing single temperature events with averages for a site. When the snow melts in the parking space, it is back to + bias for the rest of the year. Bottom line, a CRN site by its description and biases listed will be warmer in the long term average, even though events like snow, ice, or an attack on the MMTS by a Slushee machine may change that temporarily. Prove it won’t be on the long term average for station as described by NOAA’s standard as CRN5.
This system is good enough for NOAA to use for the new Climate Reference Network, so complain to them. In fact why not write a letter to Dr. Thomas Karl at NCDC, who signed of on it, and tell him just how flawed you think the system they use is…ooops that would require you to put your real name to paper. Can’t have that.
Purely anecdotal and for the contemporary record.
Raw data for SEATAC to April 17, 2008, it has been cold out here.
From NOAA, at SEA-TAC, the following are monthly departure averages from the current record as of 4-18-08 -/+ 0.1 degrees F. I didn’t use a calculator.
These notes are contemporaneous.
November 07: -0.5
December 07: -0.67
January 08: -2.00
February 08: +1.00
March 08: -2.50
April 08 to date (17): -2.15
It snowed again today in Tacoma. It has never done this in the most recent 16 years I have lived in the NW. In fact, the latest snow in those 16 years was in February. To see this month claimed to be one of the hottest is beyond the pale. This is not just a local anomoly.
I am convinced raw temperature data may forever become non recoverable, even on the internet. Thus I am doing my little part here.
Of course, I may be wrong.
You are confusing single temperature events with averages for a site. When the snow melts in the parking space, it is back to + bias for the rest of the year.
i don t confuse that. i was simply pointing out that the number given is NOT a minimum error. if it was a minimum number, it would happen under all conditions, even when snow is covering the concrete…
REPLY: It is open ended, the error may be greater than or equal to, and we’ve seen examples of CRN5 stations (like Baltimore) than show this.
This system is good enough for NOAA to use for the new Climate Reference Network, so complain to them.
the system of station clasiification is fine. i don t have any complaints about it.
but i do disagree with the way you represent it on this page!
Leroy makes it perfectly clear that the numbers are an estimate, by adding a question mark behind the temperature number. “Classe 3 (erreur 1 °C ?)” (pdf page 6)
REPLY: Well perhaps a ~ (approximate) would be appropriate if that’s your whole argument. I’ll add the ~, not that it will make any difference to how the project is carried out.
http://tinyurl.com/44tbnp
the NOAA paper states, “The errors for the different classes are estimated values.”
both papers are written for proffesionals. non of them would draw the false conclusions from this notation of error estimates, that your readers do. (minimum error…)
so you should provide MORE information on them, but you actually give less than your sources.
final point: of course a class V station will give higher temperatures than one of class I.
but the effect will be far from 5°C. (Steven Mosher has it about right above)
REPLY: Well our experience looking at stations shows that there are errors that large or larger, so we’re good to go. Again go survey some and see for yourself.
actually this rather nice paper by LaDochy gives 0.5° (factor 10 smaller) as a reasonable estimate of the real difference.
http://ams.confex.com/ams/pdfpapers/119064.pdf
Evan and Bill,
Thanks for taking the time to address some of my statements. I know they are entry level but it helps the learning process for us less than technical folks.
Jerry Magnan: Yes, that sound like a good idea.
Is there a down loadable list of stations? Also is there survey instructions?
sod, to be clear, my assumption is that the Leroy study indicates the average error.
Thus:
Classe 1
Classe 2
Classe 3 (erreur 1 °C)
Classe 4 (erreur 2 °C ou plus)
Classe 5 (erreur 5 °C ou plus)
(Sounds so much better in French.)
To be clear, when I said I was using the minimum estimates, what I meant I was excluding the “ou plus” from the calculations.
From what we know of heat sinks, their effect is most noticeable at T-Max and T-Min (esp. the latter). If the US did an hourly average, rather than a straight average of T-Max and T-Min, there would be a decreased potential for error.
actually this rather nice paper by LaDochy gives 0.5° (factor 10 smaller) as a reasonable estimate of the real difference.
Unfortunately the two sites being compared are both smack in the middle of downtown LA (if I read this correctly). Yes, one site has better microsite creds, but they are both masked by UHI.
Firthermore, even in his December 2007 study, LaDochy does not address microsite issues, but only UHI effect.
Needless to add, if rural stations are compromised by microsite violations, any UHI estimate that is based on them is bound to be severely lowballed.
The infamous Baltimore site the Rev posted earlier shows a much greater urban warming offset (+8°C IRRC).
After reading the last several posts (showing at the time I write this), I am going to go out on a limb and ask: Are we now not seeing the forest because of the trees? Wasn’t the intent of the Surfacestation surveys to show that Hansen and the others were using suspect data to push their AGW views? Hasn’t Anthony shown this already and with each and every station that is in conflict with their own criteria for data collection just another nail in their coffin? Hasn’t Anthony’s work already made Hansen adjust his “findings”? Does it matter if the bias are estimates or appoximations? Or if the bias is 0.5 or 0.7 or whatever number that can be derived by manipuating the data? The sad fact is that Hansen is using GIGO to sway opinion and influence government officials who then pass legislation based on unclear and unproven principles to the detriment of us all. When the master blowhard of the universe, Sir Albert Gore, pitched his boogeyman to children and then declared the debate to be over, huge red flags should have gone up in everyone’s head. I am glad Anthony saw them.
Thanks for taking the time to address some of my statements. I know they are entry level but it helps the learning process for us less than technical folks.
I am a less technical folk. And I am trying to boil the arguments and factors down to language that anyone (including myself) can understand.
For the AGW theory to have even a prayer of holding up, the site violations adduced by the Rev would have to be meaningless. They are demonstrably not meaningless.
As a result, none of it adds up. The other intrusive issue that points out the “not adding up” is the D’Aleo correlation of temperatures with AMO/PDO cycles: If the 1979-1998 warming trend has been exaggerated by a factor of, say, two, the D’Aleo correlation fits even better than D’Aleo has so far indicated.
The “soft spot” seems to be in the comparison of satellite-to-surface measurement. But I do not yet know where that is, though I am trying to find out.
I am new to this, but I don’t notice any qualifications in the classification scheme for:
1. Maintanance of whitewash or paint on shelters (this would impact temperature of course).
2. Aspiration (aspirated instruments show less error in sunshine). Even a well sited instrument will show a higher temperature if it is not aspirated. This creates problems for historical temperatures since earlier instruments were not aspirated. It seems there is a lot to sort out.
Bill D.
Tom,
There are no entry level in the climate change discussion. We are already in climate change. It is good to try to learn all that you can. Then apply what you learn to your daily life and to how you vote if you are part of a democracy. Climate change happens all the time in the past, in the now, and in the future.The question as I see it is how it happens, how much it happens, and how to react to it. My personal feelings are that, A we must continue to study the climate to be able to reliably predict our future needs for survival, B continue to study to determine how/why climate change occurs, C How we react to it should be to prepare and adapt to what ever nature brings because I don’t believe we can change the climate enough to make any difference. We must do what creatures have done for eons. Adapt or become much fewer in number or extinct. I opt for adaptation, the other option is grim.
Bill Derryberry
EJ,
Like you I have started to keep a record of high low and precip here at home. I find that by inversion and other variables our temps are usually 5= 8 degrees different from the official temps taken at the Birmingham airport some 90 miles away. the difference is notable in both summer and winter mostly in the winter as we are most always 5 – 8 cooler here.
Just to let you know that you are not the only one that is beginning to not trust the official temps.
I used to live in Kissimmee Fl. and know that the weather was often adapted by the tourism index. (the better the weather the better the tourism). and that country is all pretty flat and often most dampish. Mostly we didn’t get official temps but TV temps so that may have been the difference. Here in Alabama we get both the official temp and the TV station temp. the official temp is usually different more so than the tv station temp. May have to do with the tarmac at Birmingham Airport.
Bill
Derryberry
Over the last three days in South Everett we got perhaps 10 inches total (between snowing and melting snowing and melting). That’s more than we usually get in two winters combined. And rarely if ever this late.
Here is another way to look at it. If the average bias is 1.95C per STATION.
Then what kind of INCREASE in trend would we see over the past 100 years?
Geta loada this.
According to Yilmaz, et al (2008):
City of Erzurum (a “new developing city” in E. Anatolia [Turkey], alt. c. 1900m).
Mean Temperature differences:
Asphalt/Concrete – soil: +6.5°C
Soil – Grass: +5.3°C
Asphalt/Concrete – grass: +11.79°C
Mean Temperature differences 2 meters above:
2m above Asphalt/Concrete to 2m above soil: +5.22°C
2m above Soil – 2m above Grass: +2.32°C
2m above Asphalt/Concrete – 2m above grass: +7.54°C
http://www.ejournal.unam.mx/atm/Vol21-2/ATM002100202.pdf
I think that pretty much bears out the CRN ratings. And then some! Certainly an eye-opener.
Here is another way to look at it. If the average bias is 1.95C per STATION.
Then what kind of INCREASE in trend would we see over the past 100 years?
Oh, I totally agree that is the correct question. That is exactly why I keep banging on the fact that most of these-here site violations have occurred since 1980.
To wit:
–The MMTS switchover (which is why there are so many CRN4 violations).
–The air conditioning revolution.
–The massive post-1980 exurban creep.
As I have said previously, my guess is that the increase from 1979 – 1998 has been exaggerated about twofold. (Or, very roughly, 0.3°C.) This would put current temperatures at about the mid 1930s level and would match up (even) better with Joe D’Aleo’s PDO/AMO correlations.
Here in Alabama we get both the official temp and the TV station temp. the official temp is usually different more so than the tv station temp. May have to do with the tarmac at Birmingham Airport.
Hmmm. Funny you should say that.
I happen to know a temperature-reader from Alabama.
Let me bite my tongue and merely comment that in my opinion it would be preferable if temperatures were measured using a strictly automated system, and that the results were immediately and automatically transmitted to the NOAA.
I’m not sayin’ anything. I’m just sayin’ . . .
evan, you wrote
“As I have said previously, my guess is that the increase from 1979 – 1998 has been exaggerated about twofold. (Or, very roughly, 0.3°C.)”
now read your first post, where you say the average bias is 1.95C.
do you see the problem? A CRN5 may very well have a 5C bias, on some days
in some conditions, at some times. When you look at the bias over time,
you’ll find that the average is small, more on the order of .3C as you point out.
looking at data and cruching all the numbers for class5 I found it to be something like .15. When we get more class5 data, I’ll do those calculations and share them