While Joe Romm squawks about record highs being “obscenely hot” over at Think-Climate Progress, there’s a quiet bit of questioning going on within NOAA about the veracity of the surface temperature measurements, particularly related to ASOS stations at airports, which have made up a significant number of recent record high temperatures in the USA in June.
Below is an extraordinary interchange between Dr. Roger Pielke Sr., and Greg Carbin, of the NOAA Storm Prediction Center.
Figure and Data Analysis Provided by Greg Carbin of NOAA
I received the e-mail below from Conrad Ziegler that alerted me to a really important analysis of siting issues with respect to surface temperature data. The relevant part of Conrad’s e-mail, and the e-mails on this figure by Greg Carbin and Richard Grumm are reproduced below (with their permission).
Date: Thu, 30 Jun 2011 08:44:32 -0500
From: Conrad Ziegler <xxxxxx>
To: pielkesr@xxxxxxCc: gregory.carbin@xxxxx, richard.grumm@xxxxxxx Conrad Ziegler <conrad.ziegler@xxxxxx>
Subject: Fwd: [Map Discussion] All-time maximum temperatures broken or tied in
June
Hello Roger!
I hope this message finds you well!
I thought you’d be interested in the correspondence below between Rich Grumm and Greg Carbin about US record high temperatures since 1950. I’ve attached Greg’s interesting figure that accompanies his message below (Greg, I hope you don’t mind my sharing this). I know you have long advocated the need for adequate siting of surface operational meteorological sensors worldwide, and have written on the subject of effects of poor siting on temperature time series. Rich’s message raises the bar so to speak…the instrumentation itself needs to be properly designed and sensors aspirated correctly. One would have hoped this was at minimum a standard thorughout the US, though reading Rich’s message I’m no longer so sure.
Best Regards,
Conrad
Following is Greg’s initial e-mail on his analysis
From: gregory.carbin@xxxxxDate: Wednesday, June 29, 2011 10:37 pm
Subject: All-time maximum temperatures broken or tied in June
To: HWT Map Discussion Listserv <discussion@xxxxxx>, SUNY
Map Listserv <map@xxxxxx>
Maps,
I was wondering how the 41 all-time max temperature records broken or tied in the U.S. during June 2011 compared to other Junes of the past. So, I wrote a script this evening to grab the all-time annual maximum temperatures that were broken or tied during the month of June, for all Junes, back to 1950. The resulting chart is attached. June 2011 has had the most all-time maxes broken or tied (16 broken/32 tied through June 29) since June 1998 when 89 all-time max temperatures where either broken (31) or tied (58). However, the month of June 1994 really bakes the cake! That’s when a total of 229 all-time max temperatures where either broken (112) or tied (117). June 1990 was also another hot one with a lot of records either tied or broken. See the chart for more details.
-Greg
Greg Carbin
Warning Coordination Meteorologist
NOAA-NWS Storm Prediction Center
National Weather Center
Norman, Oklahoma 73072-7268
Richard Grumm’s of NOAA replied
From: Richard.Grumm@xxxxx Date: June 30, 2011 7:49:12 AM CDT
To: gregory.carbin@xxxx Cc: SUNY Map Listserv <map@xxxxxxx>, HWT Map Discussion
Listserv <discussion@xxxxxx>
Subject: Re: [Map Discussion] All-time maximum temperatures broken or tied in June
Greg,
Nicely done. Kind of warms the soul.
I ponder if the sensors used in ASOS massively deployed circa 1991 played some role in this, that is [a] downward trend though most of this would be COOP site data. Did we go through some program to improve those sites in the 1990s?
As for ASOS sites, if used, we had one really badly placed sensor at KMDT when we opened in State College. The ASOS was aspirated and on grass, not a roof. It ran cooler in parallel to our older unit (was it an unapirated HO83?). We did not officially began using the ASOS until AFTER your 1994 spike. Hitting 100F at KMDT has been extremely difficult since the new sensor. Being on grass and on the ground helps.
Additionally, we had 2 very warm ASOS units and I know an office not too far west of me that did. It turns out how the fans are installed is critical. One does not want to pull hot air up into the system. We had a sensor that was really warm last year and in the July heat wave (KSEG) was often an eastern US hot spot. New fans lead to local cooling.
How much of the 1980s was badly placed sensors? Did we improve COOP sites in the 1990s? How much of this was the pattern? Some of those 1980s years were hot but how much did sensors contribute to this too? Gotta find the eggs for this Duncan Hines cake mix.
Thanks
Rich
The relevant part of my reply to Conrad’s e-mail is reproduced below.
Date: Thu, 30 Jun 2011 08:20:54 -0600 (MDT)
From: Roger A Pielke Sr <pielkesr@xxxxx>
To: Conrad Ziegler <conrad.ziegler@xxxxx>
Cc: Roger A Pielke Sr. <pielkesr@cires.colorado.edu>, gregory.carbin@xxxx,
richard.grumm@xxxx Subject: Re: Fwd: [Map Discussion] All-time maximum temperatures broken or tied
in June
Hi Conrad
Thank you for sharing the information on the surface temperature measurements! I have a question and a list of some of papers on this issue below.
First, Greg/Rich – do I have your permission to post your e-mails and figure on my weblog?
Second, we have been examining the effect of siting quality on the long term surface temperature record, with our most recent paper
Fall, S., A. Watts, J. Nielsen-Gammon, E. Jones, D. Niyogi, J. Christy, and R.A. Pielke Sr., 2011: Analysis of the impacts of station exposure on the U.S. Historical Climatology Network temperatures and temperature trends. J. Geophys. Res., in press. Copyright (2011) American Geophysical Union.
http://pielkeclimatesci.files.wordpress.com/2011/05/r-3671.pdf
We also have other papers on this subject; e.g. see
Klotzbach, P.J., R.A. Pielke Sr., R.A. Pielke Jr., J.R. Christy, and R.T. McNider, 2009: An alternative explanation for differential temperature trends at the surface and in the lower troposphere. J. Geophys. Res., 114, D21102, doi:10.1029/2009JD011841.
http://pielkeclimatesci.files.wordpress.com/2009/11/r-345.pdf
Klotzbach, P.J., R.A. Pielke Sr., R.A. Pielke Jr., J.R. Christy, and R.T. McNider, 2010: Correction to: “An alternative explanation for differential temperature trends at the surface and in the lower troposphere. J. Geophys. Res., 114, D21102, doi:10.1029/2009JD011841″, J. Geophys. Res., 115, D1, doi:10.1029/2009JD013655.
http://pielkeclimatesci.files.wordpress.com/2010/03/r-345a.pdf
Steeneveld, G.J., A.A.M. Holtslag, R.T. McNider, and R.A Pielke Sr, 2011: Screen level temperature increase due to higher atmospheric carbon dioxide in calm and windy nights revisited. J. Geophys. Res., 116, D02122, doi:10.1029/2010JD014612.
http://pielkeclimatesci.files.wordpress.com/2011/02/r-342.pdf
Pielke Sr., R.A., C. Davey, D. Niyogi, S. Fall, J. Steinweg-Woods, K. Hubbard, X. Lin, M. Cai, Y.-K. Lim, H. Li, J. Nielsen-Gammon, K. Gallo, R. Hale, R. Mahmood, S. Foster, R.T. McNider, and P. Blanken, 2007: Unresolved issues with the assessment of multi-decadal global land surface temperature trends. J. Geophys. Res., 112, D24S08, doi:10.1029/2006JD008229.
http://pielkeclimatesci.files.wordpress.com/2009/10/r-321.pdf
Parker, D. E., P. Jones, T. C. Peterson, and J. Kennedy, 2009: Comment on Unresolved issues with the assessment of multidecadal global land surface temperature trends. by Roger A. Pielke Sr. et al.,J. Geophys. Res., 114, D05104, doi:10.1029/2008JD010450.
http://pielkeclimatesci.files.wordpress.com/2009/10/r-321b.pdf
Pielke Sr., R.A., C. Davey, D. Niyogi, S. Fall, J. Steinweg-Woods, K. Hubbard, X. Lin, M. Cai, Y.-K. Lim, H. Li, J. Nielsen-Gammon, K. Gallo, R. Hale, R. Mahmood, S. Foster, R.T. McNider, and P. Blanken, 2009: Reply to comment by David E. Parker, Phil Jones, Thomas C. Peterson, and John Kennedy on .Unresolved issues with the assessment of multi-decadal global land surface temperature trends. J. Geophys. Res., 114, D05105,
doi:10.1029/2008JD010938.
http://pielkeclimatesci.files.wordpress.com/2009/10/r-321a.pdf
Hanamean, J.R. Jr., R.A. Pielke Sr., C.L. Castro, D.S. Ojima, B.C. Reed, and Z. Gao, 2003: Vegetation impacts on maximum and minimum temperatures in northeast Colorado. Meteorological Applications, 10, 203-215.
http://pielkeclimatesci.files.wordpress.com/2009/10/r-254.pdf
Pielke Sr., R.A., T. Stohlgren, L. Schell, W. Parton, N. Doesken, K. Redmond, J. Moeny, T. McKee, and T.G.F. Kittel, 2002: Problems in evaluating regional and local trends in temperature: An example from eastern Colorado, USA. Int. J. Climatol., 22, 421-434.
http://pielkeclimatesci.files.wordpress.com/2009/10/r-234.pdf
We are currently looking at the instrument change issue as a follow on to our Fall et al 2011 paper. Your input on our paper would be very valuable.
This is really important and influential work and I am glad you have shared it!
Best Regards
Roger
Last evening when Greg sent me the latest figure with the June 2011 data, he also updated the information with the e-mail below.
On Thu, 30 Jun 2011, gregory.carbin@xxxx wrote:
Roger,
The updated chart is attached to this e-mail. It may be trivial to add one more record for the month but it was a new record (not a tie) and it was also the only new all-time record high broken in Kansas during the month. There were all-time max temperatures in Kansas that were tied during the month but only this one, on the last day of the month, was a new record. The other states with stations breaking all-time high temperature records in June were: FL, NM, OK, and TX. Again, there were 17 new all-time maximum temperature records set in June 2001, and 25 all-time maxes tied during June 2011 according to the data available from NCDC as of late this evening.
The top three years and numbers from the chart, if you need them handy, are:
Year, # Tied,# Broken, Total
1994, 117, 112, 229
1990, 95, 62, 157
1998, 58, 31, 89
I will look forward to your post.
-Greg
============================================================
From Anthony: I have documented serious and uncorrected problems with the NOAA ASOS system at airports. One notable train wreck of false high temperatures (which still remain in the climate record today) Occurred in Honolulu , HI at the airport. Just look at the difference.
The data from the two Oahu stations, 3.9 miles apart on the south shore. When plotted side by side, was telling. I marked missing data, the record high events, and when the ASOS was repaired.
The ASOS sensor read several degrees high, setting a string of all-time new temperature records. Seems to be a recurring problem, see my analysis over several posts:
This is your Honolulu Temperature. This is your Honolulu Temperature on ASOS. Any questions?
More on NOAA’s FUBAR Honolulu “record highs” ASOS debacle, PLUS finding a long lost GISS station
NOAA: FUBAR high temp/climate records from faulty sensor to remain in place at Honolulu
How not to measure temperature, part 88 – Honolulu’s Official Temperature ±2
Guest Weblog: Ben Herman Of The University Of Arizona – Maximum Temperature Trends and the HO83
Key West, FL sets new subzero “record low” temperature – Update: now snowing!
Hot Air in Washington DC- More ASOS Failures?
=======
I have more coming on this, look for another post on the ASOS/Airport issue related to high temperatures.
David;
— the siting isn’t too great, either!
😉
Thank you Anthony, your work has maybe allowed those blinded by AGW hysteria and money to see. Time will tell how Diogenes does in the progressive new world.
More on ASOS high temp issues? I don’t think that anything could surpass that Honolulu fiasco, made all the worse by the fact that they’ve kept those “record” high temps, when it was obvious that the equipment was showing temps at a minimum of 3° too high. Bet a record low temp would get a much different treatment.
Wonder what is next. Perhaps it is time for some changes at the notable NASA think tank. I am glad that NOAA is noticing.
Bill Derryberry
DR says: July 7, 2011 at 7:14 pm
I’m beginning to think the good old mercury thermometers are still more reliable than the newfangled electronic albatrosses.
Obviously you are mistaken, or why would they have had to revise the early (LIG) USHCN temperatures down so much for USHCNv2? </sarc>
Following up from my previous comment, my memory failed me a little. It was the paper of Christopher Davey and Roger Pielke, Sr. in 2005 BAMS paper [1] that attracted NOAA’s attention.
Davey-Pielke 2005 is a classic critical expose of poor siting. They visited 57 stations in eastern Colorado, documenting the exposure of the temperature sensors. Their BAMS paper is replete with color photographs worthy of the SurfaceStations project, showing sensors next to buildings, in parking lots, surrounded by brush, near an air-conditioning unit, or near mixed ground of varying albedo.
The worst one, with photographs in their Figure 12, was the Las Aminas station. This one would belongs on the wall along with Anthony’s best classics of bad siting. It’s worth quoting Davey and Pielke Sr. at length. The last sentence is a paragon of understatement (my emphasis added):
“The Las Animas site had, by far, the poorest exposure for the USHCN sites we visited (Fig. 12). The temperature sensor is set up over a gravel surface at the southeast corner of the main building of the Las Animas Power Plant. The sensor was moved in the 1980s from an open field about 50 m northwest of the power plant building to its current location on the southeast side. The 10-m building is only about 2 m west of the sensor. It blocks ventilation for the sensor in all directions north and west. Additionally, an exhaust vent is only 2 m north of the sensor; any air discharges from this vent would very likely affect temperature readings. There are also three short stacks between 3 and 10 m north of the sensor. To the north and east, about 10–20 m away, are several sheds with metal siding, along with several other metal storage features. In summary, this is a very poor site for measuring air temperatures.”
Moved from an open field to sit next to a power plant 2 meters from an exhaust vent. Perfect. Davey and Pielke Sr. finished their paper by observing, “Similar variability in the climate observing sites in the worldwide dataset of land-surface temperature trends would raise questions concerning the use of the historical record to assess regional and even global temperature changes.”
And how did NOAA and the NCDC respond? As good public servants, stewards of important data, and professional committed scientists, they responded constructively, right?
Well: Their response was the very next paper in BAMS [2]. What a coincidence. After a lot of blah about how they are dedicated to providing the very best data, the highest quality instrumentation, and how they take care in choosing site locations, Vose & co. get to the nitty-gritty of what they mean to do about the problems Davey and Pieke Sr. revealed.
It’s worth quoting them extensively, too:
“Although undesirable exposures add uncertainty to trend estimates, Davey and Pielke (2005) did not demonstrate that USHCN siting problems systematically bias long-term temperature trends at the regional level. For instance, they did not quantify the temperature bias produced by the exposure problems, nor did they show that those problems actually resulted in spurious temperature trends at any station. Furthermore, their analysis was a static assessment of site exposures over a relatively small part of the country, an area within which station exposures varied considerably. In other words, their results do not show that a large number of USHCN stations have a comparable exposure problem, much less exposures that bias temperature trends over large areas. (emphasis added)”
That is, Vose & co. including Tom Karl, head of the NCDC, after seeing a report that the USHCN may be in disarray, respond with a peremptory dismissal.
Any concerned scientist would have taken alarm and begun a comprehensive survey of the remaining sites, or even just of a representative selection of sites, to determine the extent of the problem. Nope, not a bit of that. The problem is minimized and Davey and Pielke are faulted for not surveying farther afield and for demonstrating that bad siting produces bad data.
Note that Vose & co. do not argue from knowledge. They argue a speculative polemic. That which was not shown to be bad is implicitly asserted to be good. Here’s a plain language restatement of their points: Bad siting not shown to bias temperatures don’t bias temperatures. Siting problems found to be pervasive in bounded regions do not call into question siting in larger regions. Temperatures from poorly sited sensors, not shown to be biased, are not biased. Even if all the examined stations prove badly sited, this does not mean that any of the remaining unexamined stations are badly sited.
Hand-waving dismissal never expected from a professionally competent scientist. Under normal circumstances, a defensive polemic like that would be widely recognized as signaling incompetence.
[1] C. A. Davey and R. A. Pielke Sr. (2005) “Microclimate Exposures of Surface-Based Weather Stations” Bull. Am. Meteor. Soc. 86(4) 497-504.
[2] R. S. Vose, D. R. Easterling, T. R. Karl and M. Helfert, (2005) “Comments on “Microclimate Exposures of Surface-Based Weather Stations”, Bulletin of the American Meteorological Society, 86(4), 504-506.
Oops the end of paragraph 3 from the bottom should be ‘… and for not demonstrating that bad siting produces bad data.’
So we’d better stick to the satellite record instead?
I live in Hawaii, windward side of Oahu. It’s almost unheard of here for temperatures to top 90 F (32 C) yet that seems the norm at Honolulu Airport. The PTWC temperatures mesh with my day to day expectations while the PHNL thermometer must have been directly behind the hot exhaust of departing aircraft. An 8 F (4.4 C) difference is just not credible.
Mariss, there was a hilarious episode on Channel 2 Honolulu where the weatherman at the 2 meteorological stations at/adjacent to Honolulu Airport said that the rise in temperature was entirely due to asphalt, exhaust, and the lack of vegetation. I believe the tape was posted here. But I saw it on TV.
There is another side though. My temperatures in Kekaha regularly exceed the reported temperatures by NOAA by 5 degrees. I do not believe that is abnormal. In fact I believe these temperatures have been the norm in this desert for thousands of years. I believe it has always been true and has been under-reported because NOAA is guessing and the military simply does not give a damn about AGW.
Really?
Maybe they’ll be able to get rid of that pesky super El Nino.
Remember, they started adding heat to the new electronic thermistors in the 90’s.
Scan video forward to 3:00 and play.
What I find incredible is that if they had admitted their problems decades ago and spent even a small percent of the money then doled out on wind energy on proper instrumentation, then there is no way on earth we would be in this mess.
Instead: because the temperature readings were showing the result the “save planet earth” money grabbers wanted, they had not the slightest interest in correcting the many and obvious problems of the temperature series.
Worse still … like “News of the World” the rot spread, the culture of deceit, of sweeping criminality under the carpet and promoting those who “got the results” over those who “got some results — honestly” spread.
OK, Murdoch also owns Fox news, and fox news is amongst the few who report the other side of global warming but ….
…Perhaps if fox news hadn’t been so vociferously anti-global warming based on knew jerk editor bias, then the other MSM might have not bee so anti-fox: perhaps they would have done their job and reported the news accurately and not just done a left-wing gut instinct, react against anything that Fox news supports.
Who’d a thought that badly-sited instruments could skew temps upward. How convenient for the Warmista!
All credit to Anthony for identifying this issue years ago, but why has it taken so long for at least some of the so-called ‘scientific’ community to wake up and catch up?
It does annoy me that the US is stuck using F when C just makes better sense! Although we here in the UK use C, it’s really funny listening to people talk. When it’s cold they talk C, but as soon as it gets hot they start talking F!
It is quite simple – NOAA is not a quality agency. In any other area of science or engineering finding errors of this magnitude would be cause for immediate remedial action and removal of the bad data from any records. Staff could expect to be reprimanded or dismissed for their failure. But with NOAA we see a random slap-dash, unconcerned and often defensive approach. This is the mark of an agency that is unconcerned with the quality of its work and deliverables.
All NOAA records (not just temperatures) should now have an ‘asterisk’ added as not to be trusted or used until a formal technical assessment and quality audit is passed by each site and its data handling. If NOAA happily accepts invalid data and poor quality where they can easily be found, NOAA work in other areas must be equally suspect.
What kind of government is it that has formal congressional hearings on baseball players that may have taken performance enhancing drugs, but has no reaction to an agency generating invalid metrics on which the government is basing fundamental shifts in economic policy?
Hmmm. if you look at the temperature data for Ross-on-Wye, which the UK Met Office has kindly provided here:-
http://www.metoffice.gov.uk/climate/uk/stationdata/
And you then plot the figures for Tmax in Excel or similar for the years 1930 to 2011 you find something strange. The average temperatures jump up by 2 Celsius after 1985. Well the monitoring station has been in the same suburban location since 1930, and its nowhere near an airport.
Could it be something to do with the re-opening if the weather monitoring station by goggle-eyed TV weatherman Iain McKaskill in 1985?????
I think so.
Full history of this montitoring station is shown here:-
http://www.ross-on-wye.com/index.php?page=ross_250-Weather_Station
Conclusion could be that changing from Mercury thermometers to new-fangled electronical thermometers puts yer temp readings up by 2Celsius? And may be this was done all over the Western world between 1980 and 2000? And all those CRU graphs just show an amalgamation of how electronic thermometers read higher than mercury thermometers together with airports getting busier?
“So we’d better stick to the satellite record instead?”
Great idea. Measure the temperature accurate to 0.1Celsius using a satellite a hundred miles up in the sky with equipment on board that only claims to be accurate to 0.5Celsius. Seems to be a reasonable approach for climate scientists but as an engineer I would dispute how this could ever give you sensible results.
What Team-AGW should be doing is pushing for better more reliable monitoring stations. I wonder why they aren’t?????
It’s worse than we thought!
So placement of sensors HAS an impact in the measurement of global temperatures?
Who would have tought…
This gives “Man-Made Global Warming” a whole new meaning, doesn’t it…?
Nifty chart. Really.
Two things though.
1. Doesn’t the probability of setting a new high at any given site fall off with the length of the temperature record for that site? What I’m thinking, perhaps naively, is that if the temperature record is ten years long and everything is random, the probability of setting a new high (or low) in year eleven would be one in ten. Whereas if the record were 100 years long, the chances of setting a new high in year 101 would be one in a hundred.
2. If I’m right about point 1, doesn’t the data plot have entirely the wrong shape for temperatures being random? If they were random, the chart would sweep down to the x axis from a peak on the left and sort of approach the x axis asymptotically? That would suggest that there is a pattern to the data caused by some combination of temperature increases, changing site biases, and possibly other factors?
They continue to rely on fans for aspiration? How often do these get checked as mechanical devices, such as fans, fail regularly. — John M Reynolds
Thanks Pat, the background is helpful. It is hard to understand the hubris that emanates from NCDC.
We have an electronic (LA CROSS TECHNOLOGY) thermometer set up in the back yard, under a tree, which is just in back of our house (south facing). One thing I’ve noticed is that our daily high temps are usually about 5 – 7° F lower than those recorded in Concord NH, roughly 10 or 12 miles West Southwest of us. I believe the thermometer there is at the local airport (surprise surprise). Additionally, the predicted highs by the local TV station are almost always 5 – 10° F higher than our actual highs.
Ian W,
Answer in short form – we have gone to a “bread and circuses” approach in government. Give the people a good show and forget the details. While I’m not a fan of wind power, you have to admire the technology when those huge blades are spinning…
But maybe, just maybe there is some light shining in the outhouse…
Mike
Assuming the site problem as suggested, what is the probable magnitude of the error, how many stations are you talking about, and what would be the impact on the US average? Or is this just a local problem, though it leads to local “records”, that really doesn’t impact the US annual/monthly averages?
Detail or substance?
Regarding the comment about all the F readings ending in .9, .0 and .1, it is apparently quite a fluke. The C readings, as pointed out, appear reasonably randomly distributed, so if one assumes those are the original readings and calculate the resulting F readings, you do indeed get the numbers shown.
Between 10 and 22 degrees C, 30% of the converted F readings would end in .9, .0 or .1, so there’s a 7/10 probability of getting a F conversion in the .2 to .8 decimal range. The chance of getting 24 readings in a row without a .2 to .8 decimal is .7 to the 24th power or about 1 in 5,000, which is why it stood out. (The odds of getting any 3-decimal range (.3,.4,.5 say) would be 10 times that, or 1 in 500.)
This is a decent example of the result of doing data mining when doing research. Apparent conclusions crop up that weren’t proven. The thread a long ways back about research results reverting to the mean over several followup studies also relates to this example. Only the research results that are anomalous get published, but it turns out that many of them are just getting the occasional random outlier that, as illustrated here, will crop up regularly, if not frequently. Do enough studies and some will certainly yield outliers. Then the outliers get published because they seem to indicate a new reality. And if no followup replication studies are done, that new reality can stick around for a long time.