During the past 2 months, there has been an active discussion in the blogosphere about the quality of the Tucson U of Arizona weather station. While I have never claimed that the blogosphere is a substitute for academic journals, the thread of Tucson discussions seems to me to be a good example of internet discussion at work, doing something that couldn’t be done very well , if at all, in more formal venues. In this case, I’ll take partial credit for initiating this particular topic as, in a post on June 11, 2007, I observed that the Tucson – Univsersity of Arizona station ranked #1 out of all 1221 USHCN stations in the 20th century trend (using the time-of-observation adjusted version).
The identification of substantial non-compliance with WMO policies at the site with the largest trend in the USHCN network has prompted a small blizzard of commentary on the Tucson station. I’ll document some of these recent discussions here [Note with an update covering discussions since this post added on Sep 4, 2007].
Atmoz, July 24, 2007
Only two days after Anthony reported non-compliance at the Tucson station, Atmoz, an anonymous climate scientist doing business out of the Uriversity of Arizona, defended the Tucson site at his blog here. He argued that the parking lot was “hardly used” and suggested that the cars in the photo had been there for over three years:
The lot is hardly used. The cars and trucks seen in the (at least 3 year old) satellite image are quite possibly the same as the ones that I can see right now.
In response to a comment by the surfacestations.org surveyor that the site was “ringed with 4-10 storey buildings”, he argued that there was only one 10-storey building, as though this were grounds for condescension:
However, there are a few gems hidden in that link. In the caption to that image, it says that the station is “ringed with 4-10 story buildings.” Granted, there is at least one building 10 stories tall nearby, seen best in tucson looking NW2. The building I’m in, directly to the north, is 4 stories. The buildings to the south and southwest are also around 4 stories. The building to the east is, at most, 7 stories tall. Here’s a tip for estimating the size of buildings… count the windows!
He also compared the temperature history from Tucson U of Arizona to Tucson airport, arguing that the histories were roughly similar, a point that I’ll return to.
Herman and Jones, Aug 6, 2007
On Aug 6, two U of Arizona climate scientists defended the Tucson station at Roger Pielke Sr’s blog here. To a considerable extent, their defence was that poor conditions at the U of Arizona weather station was not incompetence by U of Arizona personnel, but had been endorsed by NOAA:
When the station was first located at the current site, the location and instrumentation were approved as an official COOP station by NOAA. When the current electronic instrument package was installed, the local NWS station provided direct input and final approval. ..
Like Atmoz, they compared results from the U of Arizona station to the Tucson airport station, saying that the results were roughly similar. Unlike Atmoz, they acknowledged that there had been tremendous urbanization in Tucson in the 20th century, suggesting that “a good part of this rise is due to the Urban Heat Island Effect” and that while both stations showed trends:
How much of this is due to the growth of Tucson and how much can be attributed to other causes is impossible to determine at this time, but the trend is clearly present in both data sets.
Atmoz Revisits, Aug 7, 2007
On Aug 7, 2007, Atmoz responded to the above post at Pielke Sr. In this case, he located a couple of photographs of the Tucson airport ASOS station, as evidence that this was a “good” site and unaffected by urbanization.
He did not show a Google Earth image of Tucson airport, which (if I’ve got the right location), and the overhead gives a different impression than Atmoz’ pictures. [UPDATE: This has subsequently been confirmed to be the right airport, but the ASOS location within airport property is located slightly outside the frame of this image to the northwest.)
Both Atmoz and Herman and Jones compared Tucson U of Arizona to the airport site. As it happens, Comrie (BAMS 2000) is a specific discussion of the urban heat island effect in Tucson itself and deserves to be read in full by interested readers. It involved detailed surveying of the type seemingly sneered at by too many climate scientists, surveying in this case done by undergraduates. (A persistent theme here is that just because a task is somewhat menial doesn't mean that it doesn't have to be done.) This peer-reviewed study (not cited by IPCC) reported"
Tucson's urban heat island is ~3°C over the last century, with >2°C of this warming in the last 30 years. The annual average urban warming trend over the last three decades is 0.071°C yr-1 with the strongest effect in March and the weakest effect in November.
They contrasted three "urban" temperature stations: Tucson U of Arizona (Tucson NWSO), Tucson Airport, Campbell Ave Exp Farm) with four "rural" stations: Anvil Ranch, Cascabel, Oracle, Santa Rita Exp Range. I've downloaded daily data for these 7 sites from the GHCN daily information and calculated monthly averages (note that this is without all the Karl and Hansen adjustments). As you see, there is a remarkable difference between the urban and rural sites. Given that Comrie (2000) considered Tucson airport to be an "urban" site, the supposed similarity between Tucson U of Arizona and Tucson airport doesn't seem like a particularly relevant piece of information for justifying the use of this particular series.
Ben Herman, August 16, 2007
On August 16, 3007, Ben Herman (whose post at Roger Pielke's was noted up before) sent an email to Marc Morano, responding to one of Morano's emails reporting on, among other things, Hansen's "Y2K" error. In this email, he expanded on an issue relating to "aspirated" screens mentioned in passing in his post at Pielke's:
From: Ben Herman
Sent: Thursday, August 16, 2007 4:52 PM
To: Morano, Marc (EPW)
Subject: Re: L. Lohan Laments distracting from g. warming - UN Accused of Fudging Data - NBC Embarrasses Itself w/ 'Deniers' Segment - DiCaprio's 'horror' film - Round Up
Hi Marc, it's me again. I was doing some thinking about the NASA recalculation of annual temperatures and something occurred to me that I had forgotten about, and probably most people never heard about.
Sometime in the mid 1980's NOAA changed their official thermometers and went with a new company. During the next several years, here in Tucson, the official NOAA reporting sit at the Tuc. Airport started having an unusually large number of record high temperatures mostly in June and July, the hottest time of the year here. These records were being broken when no other stations within a 1000 mile radius were not breaking records. To make a long story short, it was found that the new thermometers had faulty aspirators, which were no more than small fans to blow fresh, outside air over them constantly. Thermometers without
aspirators could read several degrees too warm, as we knew. There were other station acknowledging the same difficulty, and my guess is that there were many other stations that were not publicized. I believe all official NOAA sites eventually were eventually converted over to these new thermometers. Apparently, the aspirators worked well when first installed, and some at least, gradually deteriorated. I have now way of knowing whether or not they all had problems, But I do know there stations other than Tucson that did have problems.
At some time in the mid 90's these thermometers were quietly. apparently, replaced, but temperature data for the time they were in operation, was never adjusted, at least it wasn't here in Tucson. The record highs we set during that period, are still in the books.
You can see a paper describing the Tucson situation at [Gall et al 1991] “Journal of Climate” Vol 5, page 657″ titled “The Recent Maximum Temperature Anomalies in Tucson: Are They Real or an Instrumental Problem?” url
This issue could be important if the problem was widespread across the U.S. as it occurred during the period of rapid temperature rise of the late 80′s to mid nineties. The fact that no one has mentioned it yet, to my knowledge, is not surprising as NOAA certainly did not advertise the problem, and in fact, probably tried to cover it up here in Tucson, at least. Read the paper, it’s not very long and no equations, just a nice piece of detective work to try and solve why our temperatures here in Tucson had suddenly gotten warmer. I, by the was was not an author, but was involved in some the temperature comparisons of the old thermometers with the new ones.
Have a good day,
If one now goes to the current NASA station plot for Tucson U of Arizona, it appears that, despite the publication of this problem in peer-reviewed literature, Harman’s claim that NOAA (and thus NASA) have not adjusted for the aspiration flaws appears correct.
UPDATE (Sep 4, 2007)
I posted further on previous discussion of the HO-83 thermometer on Aug 22 here here.. The paper cited above (Gall et al, 1992 entitled The Recent Maximum Temperature Anomalies in Tucson: are they real or an instrumental problem? ) stated:
Since 1988, what appears to be an abnormal number of maximum temperature records has been set at the NWS office in Tucson, Arizona (TUS). We present several analyses that indicate that the recent measurement system at TUS is indicating daytime temperatures that are 2 to 3 degrees too high. It appears that the instrument is not appropriately aspirated so that, during the day, temperature readings are significantly warmer than air temperatures, while at night they are slightly cooler. The system at TUS is similar to one that has been installed at many NWS sites around the country. We speculate on the impact that this system may bave on the climate record if the errors noted at Tucson are similar at the other sites.
On Aug 20, Atmoz showed that the ASOS station was located slightly outside the frame of the Google Earth image shown above, which I responded to in a separate post on Aug 22 here, from which the comments below are derived.
In his Aug 20 post, Atmoz omitted the caveat in which I allowed for the possibility of an incorrect location and using climate science methods typical of Schmidt or Hansen, criticized me without linking to climateaudit. He showed the Google image shown below, arguing that this vindicated the original rustic impression that he sought to convey in his original picture and stated that “The airport ASOS is located over natural terrain”, making it appear, especially with the omitted caveat, that I had grossly mislocated the airport, a refrain picked up commenters at Tamino (commenters who do not cavil at Mann’s use of Paris precipitation data in a New England gridcell).
I went back to Google Earth to try to reconcile these pictures. Below is a wider view that encompasses both images. The ASOS station location is marked by the pin. It’s at the far end of one of the runways shown in the original overhead that I showed, slightly out of the original view, but hardly in a rustic location, as the station is surrounded by urban settings in all directions and is close to large asphalt runways (known to have an impact from the Asheville CRN experiment.) So yes, this site is slightly outside the original frame, but not by much, and it is hardly in a pristine rustic setting. Comrie included it as an “urban” site which he distinguished from non-urban sites and there is nothing in this wider overhead to cause one to disagree with Comrie’s classification.
In subsequent discussion at his blog, Atmoz in his continuing effort to argue that the above post contained a scandalous error, as I reported here, Atmoz then fabricated a variation of my original graphic in which he inserted a green arrow that he said was my original attempt to locate the ASOS station – a green arrow obviously totally absent in the above graphics and inconsistent with the original caveat. Atmoz:
[Response:I've included an image here of the relative location of the station (red arrow) and where Steve showed the fictitious “airport asos station” (green arrow).
Atmoz also argued that if you stepped back even further, you could see desert to the south of Tucson International Airport using the image below. If you step back even further, the Earth will appear as a little blue marble. So what? As far as I’m concerned, Tucson International was Comrie 2000 as an urban site and there’s no reason to dispute that.
I take it that the premise of Atmoz’ objections is his view that the Tucson U of Arizona site is flawed, not because of being located over asphalt – something which is an embarrassment to his department at the U of Arizona, but because of a more generalized UHI – which has nothing to do with his department. Atmoz argues that the time series of Tucson Intl is “similar” to that of Tucson U of Arizona and thus the asphalt is not a problem. At this stage, no one has attempted to separate out the faux contribution of asphalt to the Tucson U of Arizona station; perhaps that will happen another day. For now, there is surely no problem in observing that the station with the largest trend in the entire United States is in a defective location. Yes, there may be multiple defects at the Tucson U of Arizona site: perhaps the generalized urban heat island is as much to blame as microsite asphalt. Again, so what? In addition, I’m unimpressed by arguments demonstrating “similarity” to the Tucson Intl record. Here we have peer-reviewed publications stating that the Tucson Intl record was flawed because of its use of HO-83 thermometers at the ASOS station. According to an email from a climate scientist, to his knowledge, the HO-83 error at Tucson remains uncorrected in NOAA records. Thus similarity of the Tucson U of Arizona record to a record known to be flawed is hardly a convincing argument for the validity of the Tucson U of Arizona record.
Andrew C. Comrie Mapping a Wind–Modified Urban Heat Island in Tucson, Arizona (with Comments on Integrating Research and Undergraduate Learning) , Bulletin of the American Meteorological Society Volume 81, Issue 10 (October 2000) pp. 2417–2431 url
Gall et al, 1991. The Recent Maximum Temperature Anomalies in Tucson: Are They Real or an Instrumental Problem?” url