When is a USHCN airport station not at the airport?

In his early 60’s schtick “Hippy Dippy Weatherman” comic George Carlin used to say “Why do they always give the temperature at the airport? Nobody lives there!”

In the case of the airport weather station in Susanville, CA, he’d be right. Here’s the airport station:

2007 Photo by surfacestations volunteer Russ Steele

Only one problem. Even though NCDC’s metadatabase and GISS both label this Station as “Susanville Airport”…

…the station at the airport above is not the USHCN station. This is something we discovered in quality control for our recent paper.The first clue? The lat/lon of 40.4167, -120.6631 isn’t anywhere near the airport, but at the Lassen County Courthouse downtown. WUWT?

What’s even more odd is that the alternate name is “Susanville 2SW”. This generally means the station is two miles southwest of the U.S. Post office in the center of town, an old locator method employed by NCDC for a very long time. But that isn’t anywhere close either. This weekend I figured out the puzzle with a road trip. 

In the NCDC metadatabase equipment list, they show that the station is MMTS based, and not using airport equipment:

But they show this station as having an airport identifier:

I did my preliminary work to locate the actual station using the lat/lon and the location cues which said “Sheriff’s office” so this courthouse location made sense:

So I programmed my GPS, I set off last Saturday afternoon, fully expecting to find the USHCN station there. Boy was I wrong.

After about a half hour of searching, I couldn’t find any evidence of the equipment. I even went up the hill behind the courthouse to get a vantage point so I could look at the roof with binoculars to see if the MMTS and rain gauge were on the roof. No luck.

There were no Sheriff cars parked at the substation behind the courthouse, but that isn’t unusual for a weekend and many deputies take the cars home. So I was stumped. I decided to head back to the airport for a look just in case it was maybe near the flight services office building. No luck there.

Thank goodness for free WiFi. I decided to try some searching at the local McDonald’s which offers free WiFi.

Searches for the NCDC B91 forms were no help either. Two stations showed, one current (the one I was seeking) and one that was closed:

And the B91 report simply said “Sheriff’s Dept”:

Susanville CA B91 Feb 2011 (PDF)

Great. I was stumped. Had the station recently closed maybe? There was no indication as such in any of the NCDC metadata, it all said “current”. There were no recent lat/lon changes either.

However there was one clue in the description. While both the old and new descriptions fit the courthouse location indicated by the most current lat/lon, and the courthouse is in fact near the northwest edge of the city, something didn’t seem quite the same in the two descriptions:

So I decided I’ll go back to basics, and simply Google “Lassen County Sheriff’s Department”. This is what came up:

Hmmm. Clicking on the GE marker led me to wholly different place about a mile northeast of the courthouse. I reprogrammed my GPS and set off for the new location.

Eureka, I found it:

Note the a/c heat exchanger and the backup generator. But of even greater concern was all the standing water in the ditch full of cattails near the MMTS. Note the lush lawn:

Susanville is generally considered “arid”, and almost desert. When you look at the location NOAA has a rule in the observing handbook that we cited in our 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., 116, D14120, doi:10.1029/2010JD015146.Copyright (2011) American Geophysical Union.

National Weather Service Observing Handbook No. 2 [National Weather Service (NWS), 1989, p. 46], which states that “The equipment site should be fairly level, sodded, and free from obstructions (exhibit 5.1). It should be typical of the principal natural agricultural soils and conditions of the area…

Hmmm, from the air, it doesn’t look that typical or representative to me:

Big patch of green and parking lot in the middle of an arid landscape.

And compared to the original airport location shown in the photo at top? Lawns versus dry scrub. Big change.

After doing all my photos and GPS reading, I wrote my report yesterday. Here’s the results. The station has been moved a fair amount, and the metadata and NCDC’s metadatabase does not reflect the most recent move. The lat/lon published there is from the second location. Here’s the location chronology mapped on Google Earth:

Here’s my report, and the explanation for how the station ceased to be located at the airport. The difference was the equipment, AWOS versus ASOS:

Site Survey

Susanville, CA USHCN COOP NUMBER: 048702

Location at Sheriff’s office at north edge of town, near the jail. 1415 Sheriff Cady Lane, Susanville, CA 96130

Survey By: Anthony Watts 08/13/11 at 2:30PM

USHCN Reported Coordinates of Site: Susanville AP (40.4167, -120.6631, 4184feet 1263m)

Measured GPS Coordinates of site: Lat./Long. 40.425298, -120.649119 Elevation 4211 feet 1283m

Site description and known history:

The MMTS sensor and rain gauge are located southwest (~30 feet) of the main Sheriff’s office building on an irrigated lawn. The MMTS temperature sensor is within 5 feet of a drainage ditch that has a significant stand of cattails in the summer, indicating year round water, likely runoff from lawn irrigation. A large asphalt parking lot and administrative building is approximately 70 feet to the west of the sensor. Another concrete walkway and parking lot is approximately 100 feet east of the sensor. An asphalt roadway is approximately 80 feet to the south of the sensor.

Curator Notes:

The current site manager is the Lassen County Sheriff’s office. The sensors were changed from CRS/LIG thermometers on October 24th 1996 to MMTS when the station was moved to the Sheriff’s office near the courthouse. The station was moved to this location because the Susanville airport weather observation system was upgraded by the FAA to an AWOS system, which is not as full featured as ASOS and does not report to the NOAA weather wire for regular surface observations as ASOS equipment does.

On March 15th 2004, the station was moved to the new Sheriff’s office north of town near the jail.

See Google Earth Map of all three locations.

Site Survey Notes:

  1. The lawn was well watered and well trimmed. This indicates a possibility of induced moist enthalpy bias in Tmin which is sensitive to increased nighttime humidity

  2. Standing water observed in the nearby drainage ditch with cattails, downslope from sensor, indicating excess water drainage is likely.

  3. A/C heat exchanger observed operating at SW corner of building ~30 feet away.

  4. Name of station remains as “Susanville Airport” in NCDC database even though the USHCN station has not been located there since 1996.

I don’t blame volunteer Russ Steele for the misidentification of the station in his initial survey. Back then, metadata was even worse than it is now, and all indications then (and now from the name) is that the station was at the airport. A simple administrative decision by the FAA to deploy AWOS instead of a full ASOS station resulted in this entire USHCN station move debacle.

The difference is summed up here, red highlight mine:

The Automated Weather Observing System (AWOS) units are operated and controlled by the Federal Aviation Administration (FAA) in the United States, as well as by state and local governments and some private agencies; the American National Weather Service (NWS) and Department of Defense (DOD) play no role in their operation or deployment.

These systems are among the oldest automated weather stations and predate ASOS. They generally report at 20-minute intervals and do not report special observations for rapidly changing weather conditions. There are several varieties of AWOS depending upon the sensor systems which are installed; the most common type is the AWOS-III, which observes temperature and dewpoint in degrees Celsius, wind speed and direction in knots, visibility, cloud coverage and ceiling up to twelve thousand feet, and altimeter setting. Recently, additional sensors which have become available for AWOS systems include present weather, freezing rain, and thunderstorm (lightning).

The Automated Surface Observing System (ASOS) units are operated and controlled cooperatively in the United States by the NWS, FAA and DOD. After many years of research and development, the deployment of ASOS units began in 1991 and was completed in 2004.

These systems generally report at hourly intervals, but also report special observations if weather conditions change rapidly and cross aviation operation thresholds. They generally report all the parameters of the AWOS-III, while also having the additional capabilities of reporting temperature and dewpoint in degrees Fahrenheit, present weather, icing, lightning, sea level pressure and precipitation accumulation.

Besides serving aviation needs, ASOS serves as a primary climatological observing network in the United States, making up the first-order network of climate stations. Because of this, not every ASOS is located at an airport; for example, one of these units is located at Central Park in New York City and another is located on Cabbage Hill near Pendleton, Oregon, for the sole purpose of providing climatological observations.

Mystery solved, and the surfacestations.org database has been properly updated with the correct station metadata and photos.

The temperatures between 2007 and 2011 at GISS, well that’s another mystery:

2007:

2011:

0 0 votes
Article Rating

Discover more from Watts Up With That?

Subscribe to get the latest posts sent to your email.

52 Comments
Inline Feedbacks
View all comments
August 15, 2011 12:57 pm

Derek Sorensen says:
August 15, 2011 at 11:03 am
Re: the two graphs at the bottom of the post. I’ve been taking snapshots every couple of months of the NOAA data with the intention of looking at how things change over time although I’ve not done any real analysis yet, but looking at what appear to be very significant differences between those two graphs I might turn my attention to GISS instead.
Are the historical GISS data available anywhere online?

You would have to ask around to see if anyone has a copy of each and every Station record from GISS. However I wouldn’t bother because GISS just folds, spindles and mutilates the data that NOAA comes up with. For example just wait until next month when GISS switches over to GHCN v3 as an input. You will see changes to station records from around the world just like we do/did when NCDC switched from USHCN v1 to USHCN v2. That switch of USHCN versions is probably what caused the change in the GISS graphs.

August 15, 2011 1:07 pm

To Anthony:
I mentioned this to Derek above but the differences in the GISS graph would come from the underlying input datasets. Up to May 2007 GISS used USHCN version 1. After May 2007 to 2009 GISS used the US sites in the GHCN v2 dataset until USHCN v2 came out based on the Menne 2009 paper. Remember that GISS uses USHCN for the US sites in their analysis so in 2007 you had USHCN v1 in 2011 there is USHCN v2

Editor
August 15, 2011 1:29 pm

For anyone who uses Google Earth, all the GHCNv2 stations are available in a download that gives a fully interactive fly round of all the stations, coloured by temperature trend. Clicking on a station marker brings up the raw and adjusted data in a graph. Overview and link to download here: http://diggingintheclay.wordpress.com/2010/10/08/kml-maps-slideshow/
Also available as GHCNv3 Beta although there were issues with the US stations due to a change in the station numbering, which I am not sure was resolved. Just waiting now for the GISS version.

Chris
August 15, 2011 2:06 pm

I am impressed by your fortitude Anthony.
It has been mentioned before that the accuracy of temperature measurement is one thing, but where the measuring instrument itself is sited, and any protection for the instrument are just a couple of variables that can, and have produced wildly differing results.
Perhaps one of the problems that we have today, is the potential for a temperature measurement accuracy of, at least +/- 0.5 degC variation. In the past, an accuracy of five times that figure may well have been acceptable, so in order to maintain continuity, the same variability ought to be applied to historical figures.
It is not meaningless to introduce our satellite instrumental figures, but it surely must assist by utilising a certain degree of known discrepancy from past records from our modern data to arrive at a reasonable conclusion.

August 15, 2011 2:27 pm

As far as urbanity designations for that station go for its listed lat/lon, its urban via GRUMP, impermeable surfaces, and 1930-2000 population growth, but rural via nightlights (only 19 brightness).
When run through the pairwise homogenization process, NCDC significantly reduces the 1960-present minimum trend from 0.24 C per decade to 0.03 C per decade. The max trend is mostly unchanged, going from -0.10 C per decade to -0.11 C per decade. There is some weirdness in the very early part of the century that appears not to be corrected (that dramatic drop circa 1900), likely due to the lack of nearby stations to check against that far back.
As mentioned earlier, the main difference between the two GISS graphs is the use of USHCN v1 vs. USHCN v2.

rbateman
August 15, 2011 3:06 pm

Ah, Susanville. Nice place for a windy day, all exposed to the elements.
I’m surprised they don’t just stick a sensor out in the Black Rock Desert to the East where it will remain untouched for the next 10,000 years. Sort of like a NE Calif. version of Winnemucca, NV.

Walt
August 15, 2011 3:20 pm

Garbage in gives garbage out! You can get any result you want by massaging the raw data.
That is what the problem is with these stations and their historical records.

August 15, 2011 5:18 pm

It’s worse then we thought !!!!!!!!!!!!!
Great work again Anthony !!

August 15, 2011 5:29 pm

Hello Anthony
Today I read about GHCN files in your article “Pielke Sr. on the quality of global surface stations”.
I’m very puzzled at all about the LAT/LON information given in the GHCN data.
From the GHCN station where I live (WMO 617 10763), the description says 49°30’00.0″N 11°03’00.0″E. That point is located in woods, more than 30 kilometers north of the real location at Nuernberg Airport 49°30’10.81″N 11° 3’18.49″E (exact Google Earth coords; you can see the station there).
As in GHCN TAVG v.3 it’s location is tagged as “COOL CONIFER”, which applies to the wrong location. In real it is open meadowland next to the runway (but with good distance).
The next ominous point is the begin date: 2007-10-02 => Current
https://mi3.ncdc.noaa.gov/mi3qry/identityGrid.cfm?setCookie=1&fid=45254
This is clearly not true.
The Nuernberg Airport station exists at the same location since 1943, data is available since June 1953 and validated data since 1961.
On first skimming through the NOAA GHCN data since 1961, this data is clearly NOT consistent with the real values I’ve received from the DWD (German Weather Office).
There are huge differences between NOAA and DWD. I’ll work it out the next days.
If you’re interested, please send me an email. It’s too complex for this site.
And if you need help exploring southern Germany GHCN stations, let me know :-;

KenB
August 15, 2011 5:45 pm

Gee, I suppose the whole thing is a matter of trust. Trust in equipment, trust in the data, trust in the use and trust that no one is manipulating or cooking the system. But then you have the gleeful fact that some of the “high priests” deliberately turned up/made sure a congressional hearing was artificially hot to swing millions or was it billions of research dollars to their agenda, in such a climate is it so hard to think that similar fiddling doesn’t happen with “adjustments” to GISS, when it is basically the same agenda and not at arms length from the gleeful pretenders?
Nah, no scientist of any real standing could be swayed by media and money – sarc!

Eric Gamberg
August 15, 2011 6:32 pm

Anthony,
In the interests of Surface Stations Quality Control, I suggest that you review ALL of the info now in the Cottonwood, SD folder to see if the rating for that station is properly made.
Doing my own quality control, I realized that I had likely made an error (in the haste of trying to find ~8 stations in a long day of driving, with limited experience and cryptic notes). I stopped by again at a later date and found the (relatively sheltered) MMTS instrument that was indicated in the NOAA database.
REPLY: Will do, thanks – Anthony

Steve McIntyre
August 15, 2011 7:41 pm

Anthony, back when you were starting this, I did an exploratory post here http://climateaudit.org/2007/05/30/some-northern-california-station-plots/ looking at Susanville and its neighbors in keeping with then adjustment methods.

August 15, 2011 7:51 pm

Anthony –
The effects you mention regarding the screwup in station positioning goes towards the accuracy (not precision) in the temperature data used by Hansen, I gather. From a Hansen-rise of 0.7C, I thought your previously identified problems with data accuracy [UHIE undercorrections, urban, low-latitude and warm location biases, plus computer interpolations (biased by the urban, low-latitude and warm location biases), and other adjustments of a warming, non-random nature – have I missed some, like the “superior” Arctic data?] suggested the Hansen/GISTemp long-term temperature record was about 0.3 or 0.4C too warm. Is that still the lay of the land?
This post is about another occurence of the same station-position problem, correct? Is the total warming effect of your studies still in the 0.3-0.4C range? HadCruT is about 0.12C cooler than GISTemp, though the same planet is being measured (far as I know). The difference, if attributable to CO2, strikes me as enormous if you are trying to figure out the radiative power of said CO2. Does Jones/UEA/HadCru agree that Hansen’s “superior” Arctic temperature readings give a more accurate vision of the global temperature?
What a lot of work, self-education and understanding of how the powerful, personal agenda-driven political elites get their stuff done has the climate debate created!

POWinCA
August 15, 2011 8:42 pm

You can use all the fancy statistical techniques you want, but always remember that the data has been collected by the village watchman who writes down whatever the Hell he pleases.
Nice investigative work Anthony. Some time I’ll have to tell you the story of an ARRA project in my neighborhood that magically vanished from the Recovery.gov website just before the number of jobs “created or saved” was supposed to be published.

cotwome
August 15, 2011 8:59 pm

It’s interesting in the first image labeled ‘2007 Photo by surfacestations volunteer Russ Steele’, the ground under the instruments is noticeably darker than the ground surrounding the site. Most likely due to preparing the area for the weather station, but it has obviously darkened the ground. Land use change!

juanslayton
August 15, 2011 9:10 pm

Eric Gamberg, Anthony and other interested persons,
There seems to be a problem in getting station changes promptly posted to MMS. For example the stations Weeping Water, NE, and Milbank, SD, appear to be closed. But when I spoke with the last reported observers last week, they told me that the station had merely been moved, they told me to where, and when I went to the indicated sites, the equipment was there and I got pictures. (In Milbank it had been moved twice, first to the police department, and currently to the wastewater treatment plant.)
So while we are reviewing things, we need to take a look at all the stations that appear to be recently closed, since they might have merely been moved. (Eric, if you have time, contact me at juanslayton@dslextreme.com. I’m curious about your experience in Pawnee City.)

itronix
August 15, 2011 9:32 pm

Balazs says:
August 15, 2011 at 11:58 am
“Why NOAA does not have live video camera at each site”…
Not NOAA but available for some airports and routes.
http://akweathercams.faa.gov/sitelist.php

John F. Hultquist
August 15, 2011 10:58 pm

I live in an environment called ‘sage brush – steppe’ with an average yearly precipitation of about 9 inches (230 mm). Photo by Russ Steele (first photo) shows typical vegetation. Much of the local valley is irrigated and many of the roadside ditches carry such water throughout the summer. The county highway department regularly removes the built-up sediment and cattails.
The water in the irrigation ditches is recent snow. Twenty to thirty seconds with my hands in the water is extremely unpleasant. I do not know the actual temperature. Also, the growth of cattails is on the edges of the running water – sort of like a tunnel. Water within the tangle of vegetation may be a bit warmer but it is still cold.
Why place a MMTS next to obvious water in a ditch with cattails? One needs to dig a trench to lay a cable. It looks to me that in the photos Anthony posted the idea was to put a little distance between the buildings and the sensor – but not to cross or disturb the ditch and its vegetation. Look at a bunch of the Surface Station photos and one gets the impression that these ‘ad hoc’ circumstances and decisions are much more important than regulations when it is time to start up the back-hoe and begin digging.

wermet
August 16, 2011 12:09 am

Steve Garcia says: August 15, 2011 at 12:42 pm
1. Cattails in the desert?
———————-
Sure, why not? In the Mojave Desert in CA, I would stumble across small water sources complete with cattails in some of the most unlikely locations. One summer, during the “dry season”, on the eastern edge of the Panament Valley, just a few of miles west of Death Valley National Monument, I found an abandoned mining shack next to a moderately large (~1/2 acre) “swamp” full of cattails. There was even an old willow tree (still alive) between the shack and the swamp. Even though this swamp was on the valley floor, none of the greenery was visible if you were more than ~1/3 mile away.
These types of surprises were one of the reasons that the Mojave Desert was one of my favorite places to explore.

Brian H
August 16, 2011 2:11 am

As KenB reminds us, “cooking” the political climate input is in the DNA of climatologists. Once noble cause corruption is in control, data is just a malleable tool to achieve a consensus of papered-over ignorance.

Shona
August 16, 2011 2:17 am

seeing all the money spent on this you’d have thought they’d have had grad students checking allthis stuff. But of course they are not interested in actual measurements.

August 16, 2011 2:30 am

boballab says:
August 15, 2011 at 12:57 pm
… I wouldn’t bother because GISS just folds, spindles and mutilates the data that NOAA comes up with …

Thanks, boballab; I’ll stick to what I’ve been doing then.

Gary Swift
August 16, 2011 7:12 am

“Shona says:
August 16, 2011 at 2:17 am
seeing all the money spent on this you’d have thought they’d have had grad students checking allthis stuff. But of course they are not interested in actual measurements”
None of the temperature records were ever intended to be used for the purpose of calculating global climate change. For local weather prediction, aviation, agriculture, and land management, there wasn’t ever a need to be so picky.
They were never supposed to be accurate enough to show a trend on decedal scales.
Sarcasm alert: However, if you average this in with a bunch of other stations then you can get .01 degree accuracy with 95% confidence.

John Marshall
August 16, 2011 8:04 am

I like your show of determination and effort. Well done Anthony

clipe
August 16, 2011 3:00 pm

Started out looking for a replacement for [i]Weatherpixie[/i] widget.
Somehow ended-up here.
[IMG]http://i22.photobucket.com/albums/b331/kevster1346/KSVEwx.jpg[/IMG]
[IMG]http://i22.photobucket.com/albums/b331/kevster1346/ICAOSVE.jpg[/IMG]

REPLY:
This is because the station doe snot report to NOAA – so the report is blank – Anthony