By Steve Goddard and Anthony Watts
NCDC has done an first rate job raising Arizona summer temperatures, as seen in the graph below.
How did they accomplish this? – by magic! My favorite Arizona station is Ajo, near the Mexican border. Until 1984, temperatures were dropping – as seen in the USHCN (RAW) plot below.
Apparently someone at USHCN didn’t like that trend, so they made up homogenized an extra 25 years of data with a sharp upwards trend. This spreadsheet shows the USHCN data. Note that there aren’t any years after 1985 which have a full year’s data, and no years after 1985 with a full summer’s data.
For example, note this B91 form from the Ajo observer for July 1987, missing 10 out of 31 days of data:
Here is the adjusted monthly mean maximum data plotted from NCDC:
The image below shows Ajo adjusted maximum mean (black) on top of raw maximum mean (red.) Note that they are identical until 1970, when the magical adjustments kicked in. Click on it for a clearer image.
The station is not well sited. Note the MMTS temperature sensor is inside the white stucco patio wall enclosure at right:
http://gallery.surfacestations.org/main.php?g2_itemId=33437
Here’s another view:
http://gallery.surfacestations.org/main.php?g2_itemId=33463
Photos by surfacestations.org volunteer Bob Thompson
While the near A/C heat exchanger units are comical, wind sheltering and building proximity are also likely contributors. According to NCDC MMS metadatabase, in 2002 the station was switched from a Stevenson Screen to the MMTS sensor in the location shown above. Since NCDC does not make the site sketches that exist for all stations public, we can’t see the plan map showing where the Stevenson Screen was. However, the site survey from Bob Thompson tells us:
Site description and known history: The station was previously located on a nearby hilltop, but is now close behind a Phelps Dodge administration building adjacent to an open mine. I did not find any record of the relocation, but there is nothing any longer atop the hill.
There was a notation in the NCDC MMS Metadata remarks though, saying that the station had been moved 845 feet to the northeast.
The dates don’t match the date of the equipment change in 2002 though, and since the MMTS sensor requires a cable, it is likely that it was moved when the equipment change was noted in 2002.
Most likely the metadata citing the date of the move is wrong, and/or it took NCDC time to catch up with the change made by NWS personnel.
This Google Earth view, dated July 13th, 2006 shows the location of the temperature sensor at Ajo at the Phelps Dodge plant. Basically in the middle of an industrial zone:
In this more recent aerial photo from Bing Maps, it appears the facility has been closed down, and the buildings removed. They even abandoned 3 locomotives previously used to shuttle ore cars:
Note while the buildings are missing, the asphalt parking lot to the SE of the office is new.
Here’s a view with the GE ruler, showing where the Stevenson Screen likely was:
Here’s a closeup view of where the MMTS and rain gauges are:
Interactive view at Bing Maps is here
The point of all this is that this station has the following problems:
- Poor siting – building proximity
- Station move
- Sensor change from Mercury/Stevenson Screen to MMTS
- A nearby dynamic industrial environment with rapidly changing infrastructure and albedo as shown by aerial maps
- Missing/incomplete observer data over a long period, likely due to observer not recording data on weekends, holidays, vacation days, sick days.
- Incorrect/conflicting metadata at NCDC
- post facto adjustments to infill missing/incomplete observer data
That’s a lot of uncertainty added to the base measurement. Many stations have similar problems. The measurement environment is hardly static, yet we are looking for small variations in the climate in the midst of all this noise and uncertainty.
Other Arizona USHCN raw station data is below, showing about equal numbers of stations with declining and increasing maximum mean temperatures over the last 80 years. In Arizona, it’s all about the daytime heat, not the nighttime low.
Raising Arizona was probably Nicholas Cage’s best movie. In the end, they decided to be honest and give Nathan Arizona’s baby back. Can we expect the same?
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
Sorry about the quality of that previous comment. I rewrote and edited for clarity and brevity and then neglected to delete that which was rewritten which appears below the first three paragraphs. I hate when that happens!
REPLY: Deleted, can’t figure out what your before/after was, feel free to resubmit. -A
Aside from the MAX vs AVG issue, which has already been noted, the NCDC graph is for summer (JJA), while the A&S rebuttal shows annual averages. Is the NCDC distortion just because of their choice of season, or because of their “homogenization”?
pat says:
July 10, 2010 at 9:57 am
This is just fraud, pure and simple.
REPLY: I tend to think of it this way:
Never assume malice where simple incompetence will do.
Government excels in incompetence.
-A
…………………………………………………………………………………………………………….
Case in point: an oil leak coming up on 80 days now.
Steven, I noticed in Mike’s link regarding NCDC’s quality:
http://www.ncdc.noaa.gov/oa/climate/research/ushcn/ushcn.html#QUAL
The first paragraph has two things I wonder about:
1. A quality control procedure is performed that uses trimmed means and standard deviations in comparison with surrounding stations to identify suspects (> 3.5 standard deviations away from the mean) and outliers (> 5.0 standard deviations). Until recently these suspects and outliers were hand-verified with the original records. However, with the development at the NCDC of more sophisticated QC procedures this has been found to be unnecessary.
Referring to the ‘suspect’ and ‘outlier’ data points, do you have an idea what order of magnitude the std. dev. is for the data? I assume they are talking about that for the data in a given grid cell. Secondly, did the methodology change in going from hand-verification to ‘the development at the NCDC of more sophisticated QC procedures’?
It seems to me that is an unusually large number of std. dev. they are using as limits. As few as 1.5-2 would give a fair chance the point was representative of the population, but I still would not bet on it.
I didn’t see it posted here, my apologies if I missed this elsewhere or am in error, I think these are the adjustments made through 1999.
http://i42.tinypic.com/2luqma8.jpg
REPLY: That is USHCN1, now they are on to USHCN2, and they have not published a similar graph. Perhaps you could inquire? – A
AZ Ajo temp data were taken are shown on plots at:
http://gallery.surfacestations.org/main.php?g2_itemId=79205
and
http://gallery.surfacestations.org/main.php?g2_itemId=79207
Looks like they have not been recording data on the weekends .
Joseph D’Aleo covers adjustments, sitting, UHI, etc. in 20 minutes:
“Joseph D’Aleo, PART 1, runs the gauntlet of data problems!”
“Joseph D’Aleo, PART 2, runs the gauntlet of data problems!”
Hansen, Wigley, Jones, they’re all here building a hockey stick
Anthony, after everything uncovered in the CRU tape letters and other documents and speaches given by the ACGW alarmists, do you really feel it’s pure incompetence vs. malfeasance?
Could they really be that incompetent to the point of screwing things up that badly rather than intentional mishandling with ulterior motives?
I know you can have a subtle dry whit sometimes Anthony but how can you believe that it’s just incompetence, or are you just trying to be diplomatic and full of grace?
I mean after seeing the difference in the charts between Marysville and Chico, the charts in Mexico used for that ludicrous paper on disappearing lizard populations in Mexico, and all the pictures of the messed up sensor depots are you seriously saying it’s just mere incompetence?
Granted after looking at Hansen at Nasa and Gore in politics it is easy to aim many things to the side of incompetence but I think there is a little bit more to this than mere incompetence.
Anthony you are a truly gracious man, because you give them far more credit than they deserve.
The odd thing is, here you are being gracious to them, and on other websites they’re constantly saying what a hatemonger and dictator you are.
Brad aka 1personofdifference says:
July 10, 2010 at 12:41 pm
“[…]The odd thing is, here you are being gracious to them, and on other websites they’re constantly saying what a hatemonger and dictator you are.”
That’s not odd. They know they’re failing. They’re desperate, that’s all.
the really obvious sad thing is that the alarmists will use the adjusted charts and the average person will believe the lie. thanks for your continued investigations !!!!!!!!!
latitude says:
July 10, 2010 at 11:27 am
“Mike says:
July 10, 2010 at 7:15 am
There are reasons why data is adjusted rather than used raw.”
obviously
“But explain why raw data generally is cooler, and adjusted data is usually adjusted up, by the exact same people that describe UHI.
How can UHI be described as remaining warmer at night, and temp data adjusted higher to compensate for that?”
Read the link I gave. UHI is only one of six factors used in the adjustment methods.
http://www.ncdc.noaa.gov/oa/climate/research/ushcn/ushcn.html#QUAL
“Application of the Station History Adjustment Procedure (yellow line) resulted in an average increase in US temperatures, especially from 1950 to 1980. During this time, many sites were relocated from city locations to airports and from roof tops to grassy areas. This often resulted in cooler readings than were observed at the previous sites. When adjustments were applied to correct for these artificial changes, average US temperature anomalies were cooler in the first half of the 20th century and effectively warmed throughout the later half.”
See also:
http://www.realclimate.org/index.php/archives/2004/12/the-surface-temperature-record-and-the-urban-heat-island/
REPLY: Mike, you are behind the curve. You are arguing using USHCN Version 1 references, which are no longer used. NCDC data that we used here is USHCN Version 2 with a whole different set of adjustments. -Anthony
Anthony, can you please point me to the data you’re using for your figures?
At this site: http://cdiac.ornl.gov/r3d/ushcn/ushcn_r3.html you can download the USHCN data for Ajo, maximum, mean and minimum, and the data do not match the data you show in your figures. I wonder why that is. This data tells a different story than the data you show.
REPLY: I did all the metadata and imagery, that’s a question for Steve. I’ll make him aware -A
James, just continue to enjoy all that false stuff and life in general.
As per all those false assertions, tricked figures and all what the owner is trowing. Aren’t you happy to live in a country where you can lie and you even get a fan club out of it. And the best you can make a living out of this.
Just look at the scales from the graph, it’s wrong. The data is not the same, c’mon can you see that ? are blind or what. Just take a station mentionned above, pull the figures for yourself, and you will see that the processed data is cooler than the raw data. You will also see how the owner of this website is tricking you by comparing max temps with average monthly temps.
Double check the numbers for yourself and open your mind.. You badlly need it.
I live well and i don’t need climate audit for that. I base my life and knowledge on facts and the way they are defined – not the way people are twisting the reality of things as i see here.
With this kind of work, Climate Scientists are to science what kazoo players are to classically trained musicians.
For its own good, Science needs to step on the shoddy workmanship we’re seeing from the AGWers or they will be tarred with the same brush. Credibility is all the science has.
Regg,
You are correct that my comparison between the state mean graph and the station max graphs was not correct.
However, you other assertions are nonsense. The raw max data at Ajo shows a downwards trend, while the adjusted max data shows an upwards trend. The corrections are clearly bad. The graphs were taken straight off the USHCN web site. If you think they are bad, then we are in complete agreement.
“REPLY: I did all the metadata and imagery, that’s a question for Steve. I’ll make him aware -A”
Do you mean to say you put you name on a “paper” and did not check the work of your coauthor? That’s outrageous.
REPLY: No, you are jumping to a conclusion. I’m saying that it’s a question for him, and that Julienne keeps asking me as if I’m the only one with my name on it. In any paper with more than one author, the work gets divided. The best person to answer is the one who did that specific portion. In this case, Steve. -A
also from above in case you don’t see it
REPLY: Mike, you are behind the curve. You are arguing using USHCN Version 1 references, which are no longer used. NCDC data that we used here is USHCN Version 2 with a whole different set of adjustments. -Anthony
Regg says:
July 10, 2010 at 1:37 pm
“[…]will also see how the owner of this website is tricking you by comparing max temps with average monthly temps.[…]”
Are the adjustments up or down over time, that’s the interesting question. I don’t expect YOU to answer that, you’re only here to drop a comment to cause some confusion. Simple FUD tactics.
One would expect adjustments to subtract temps as time passes because of the UHI, but often WUWT found examples where it was exactly the other way round. I think it’s one more case here.
Do your part to elect enough Republicans or Tea Party candidates for Congress to make Democrats the minority in both the House and Senate. That will empower …
That will empower the party you prefer. That is all.
This is not a political website. The moment it becomes one, I will leave. This site is good because it sticks to science and events.
Not to mention an enormous number of visitors here are from outside the US. We know this is a global issue and the nature of the US government is pretty much irrelevant in the global scale of things.
Anthony: I’ll take you on your word on Reply 1. As for Reply 2, thanks. I will read this link later, probably tomorrow:
http://www.ncdc.noaa.gov/oa/climate/research/ushcn/
Mooloo says:
July 10, 2010 at 3:17 pm
“[…]Not to mention an enormous number of visitors here are from outside the US[…]”
Yes, and i find the political fight in the U.S way more entertaining, way more informed and way more relevant than whatever the European AGW Bloc government does (which amounts to always the same, always the party line of Crazy Connie).
BTW…the link I posted was for USHCN version 3 which shows different results than the data in this posting. I don’t know if this is the “latest” version, but I would suspect it is more accurate than version 2 or version 1. So perhaps you could redo the comparison with at least version 3 data…
REPLY: Thanks for pointing that out, but I’m unaware of USHCN Version 3. Version 2 only came out last year in BAMS. Further, I can find no references to version 3 via Google, or at CDIAC, NOAA, or NCDC. Care to share? – Anthony
Regg
The scales on the graphs are the same, but they are normalised.
There are two ways to make a trend tip up. One is to lower the temperatures on the left side. The other is to raise the temperature on the right side. In this case, they did both.
Julienne
The graphs are from here :
http://cdiac.ornl.gov/epubs/ndp/ushcn/ushcn_map_interface.html
Anthony, this is what the site says:
U.S. Historical Climatology Network Description
The U.S. Historical Climatology Network (U.S. HCN) was compiled in response to the need for an accurate, unbiased, modern historical climate record for climate change research. The Carbon Dioxide Research Program of the U. S. Department of Energy and the National Climatic Data Center (NCDC) of the National Oceanic and Atmospheric Administration (NOAA) established a network of 1219 stations in the contiguous United States for the specific purpose of compiling a data set suitable for detecting and monitoring climate change over the past two centuries. This network, known as the U.S. Historical Climatology Network (U.S. HCN), and the resulting data set were initially documented by Quinlan et al. (1987) and made available free of charge through the Carbon Dioxide Information Analysis Center (CDIAC). The data presented in Quinlan et al. (1987) extended only through 1984. In 1990 NCDC and CDIAC revised and updated the HCN data records through 1987 (Karl et al. 1990). In addition, using the techniques of Karl et al. (1988), NCDC generated temperature files in which the biases introduced by urbanization effects were removed.
The new revision 3 (Easterling et al. 1996) data represent the best available data from the United States for analyzing long-term climate trends on a regional scale and may be used for studies attempting to determine the climatic impacts of increased concentrations of greenhouse gases. The data for most stations extend through December, 1994, and a majority of the station records are complete for at least 80 years. Unlike many data sets that have been used in past climate studies, these data have been adjusted to remove biases introduced by station moves, instrument changes, time-of-observation differences, and urbanization effects.
At this revision (Revision 3) the data set has been extended through the end of 1994 and several stations have been added (54) and deleted (52). Twenty-five of these additions/deletions involved stations who’s official station name and number changed, but who did not actually move. These revisions have resulted in a network with 1221 stations (Figure 1). Seems it currently ends in 1994 but I don’t know why that is. It doesn’t seem to show the same upswing though that the Version 2 data showed here does…although 1989 has the highest annual maximum temperature, the next highest was in 1934…