Who are you going to believe, my B-44 or your lyin eyes?

NCDC needs to incorporate information into their metadata database from four sources that they currently ignore and/or keep from the public.

Guest essay by John Slayton

Southern California’s San Gabriel Mountains have long been a natural habitat for scientists.  Some of their activities are well known–the observatory on Mt. Wilson and Michelson’s measurement of the speed of light come readily to mind.  But there are others. One was the touristy Thaddeus Lowe, who built an electric railway into the mountains to an observatory at Echo Peak, near another peak which now bears his name, Mt. Lowe.  Mr. Lowe’s enterprise soon changed hands, but years later, from 1926 to 1936, there was a COOP weather station on Mt. Lowe.


I hike around there from time to time, and I got curious a while back about the exact location of that station.  So I consulted NOAA’S MMS site to get a satellite view of the terrain.  But the terrain that came up was very strange.  Turned out that the coordinates given were, as I remember, about a hundred miles to the east, in the direction of Twenty-Nine Palms.  I e-mailed NOAA, and they promptly corrected the error.  We residents of Los Angeles County are happy to have our mountain back.  And NOAA thereby demonstrated a willingness to listen to laymen and correct mistakes.

Alas, the MMS site is no more.  It has been replaced by HOMR, the ‘Historical Observing Metadata Repository.’  HOMR mostly contains the same data as the earlier MMS site, but there is at least one significant improvement.  In the earlier site, all the reported locations were treated as fictional points on the map.  Fictional, because all measured coordinates are in fact approximations.  In the pre-GPS days the approximation could be pretty rough and records typically recorded low precision measurement.  The new HOMR site addresses this by dividing recorded locations into two groups:  ‘precise’ and ‘imprecise.’

Stehekin, Washington, provides a good example of how this works.  This is a fascinating vacation community at the north end of Lake Chelan.  You can’t drive there, but there are three ways to get into the area.  You can take a boat up from the south end of the lake; you can fly in on a pontooned plane; or you can use the equestrian/hiking trail.  (John Hultquist could probably give good advice about this….)  I chose to walk.

Stehekin has had a weather station since 1906.  It was located for many years at the Buckner Homestead, a site that has been turned over to the federal Park Service and is now maintained by the Buckner Homestead Heritage Foundation.  You can look Stehekin up on HOMR at: http://www.ncdc.noaa.gov/homr/#ncdcstnid=20027918tab=MSHR

Use the location tab and you will find 4 locations considered to have ‘high precision’ measurements.  The Buckner Homestead is not among them.

Under the map is a button to click to view the ‘low precision’ locations.  Clicking that will bring up three more locations.   Buckner Homestead is number 7.  It is represented by a rectangle approximately 1 by 2 kilometers, containing roads, open land, forested land, buildings, and a major river.  The station was somewhere inside that rectangle.  Not near as precise as we would like, but a very honest presentation by NOAA of the precision of their records.

We would like better precision, of course, so I went back to visit the site and look for some remnant of the station.  (Took the boat, this time.)   Looked all over, found nothing.  Ran into a fellow out there doing some work around the old house, who turned out to be Herb Sargo.  Mr. Sargo is the grandson of long time observer Harry Buckner, and the great-grandson of M.E. Field, the observer from 1906 to 1908.  He is also the president of the Buckner Homestead Heritage Foundation.  I don’t know where you’d find a better authority, and he was able to point out the location of the station “…within probably four feet in any given direction, as verified by numerous family photos.”

I have shared this information with the Spokane weather service office and they have been kind enough to reply.  They note that our information is not entirely consistent with their rendition #1 drawn map.  I am not surprised; over a period of 50+ years things get moved
around.  But I am perplexed by their policy on the metadata database:

As you know, the latitude/longitude precision goes only to DDMM on these historical forms. This is what would be represented in the station history database. A location provided to greater precision than possible with equipment during that time could garner a remark in the station’s record, but would not be fully reflected in the database.

I can only read this to mean that their database is not about the station itself, but only about the station’s official records.  Certainly those records would be the first place to look if I wanted to know about station metadata, but the database should be more than a compilation of old federal forms.

NOAA should fully integrate information from other sources, including the following:

1. Reliable Observers with Personal and First-hand Knowledge
Stehekin Buckner Homestead certainly meets this test, as do many others.  In some cases one feels almost a sense of urgency, as many potential informants are retiring.  (Example:  The gentleman who  pinpointed the Buckeye city yards location for me remarked that he was probably the only person left there who could do that.)  Sadly, some are getting up in years and will not be with us much longer.

2.  Historical Photographs
San Luis Obispo’s COOP station has a record going back to 1885.  Observations were made from 11 known locations, and historical photographs show instrumentation from 7 of these.  We have photos from (approximately) 1885, 1894, 1910, 1929, 1930, 1937, 1938, 1970, and the current location.  (Serendipitously, for several years the station was close to the university gymnasium.  The CRS shows up in the background of athletic team photos published in the school yearbook.)  All of these locations can be pretty much pinpointed and placed in HOMR’s “precise” category.  And should be.

3.  Semi-archeological Evidence
As stations are moved, some things get left behind.  The result is a physical trail testifying to station history.  For example I took a swing around Texas and neighboring states earlier this year.  At 5 sites I was able to pinpoint former locations because part or all of the MMTS posts had been left when the thermometer was removed.  (That would be Eureka Springs, Clarksville, Mexia, Alpine, Poteau.)  In one site (Alpine radio station) the concrete base of the old CRS remained.  And in two sites entire instruments were left behind (MMTS in Flatonia, CRS in Hooker).
So we know from the physical remains exactly where these stations were located.  Yet in at least four of these sites the former locations are treated by HOMR as “imprecise.”

4.  Yer Lyin’ Eyes.
The station in Falfurrias, Texas, was for some period of time located at the local airport.  It was then moved, according to HOMR, “ 4.7 MI W“.  Trouble is, the new reported coordinates (and the location shown on the maps) moved it south, not west.  On my first visit to Falfurrias  I was not able to even access the reported site, as it was way out in a large acreage that appeared to be fenced and gated on every side.
The Brownsville Weather Service office was very helpful, however, and two of their men drove all the way up to Falfurrias to give me access to the current site, which is located in a hunting lodge several miles from the reported coordinates.  True coordinates for the thermometer are 27.2326, -98.20040.

If you’ve read this far, I would recommend you Google these coordinates, because the satellite photo resolution is a bit extraordinary.  You can see, not only the top of the MMTS, but also a cast shadow that clearly shows  both the thermometer and the post it sits on.


(Of course you could also look at our ground photos in the surfacestations gallery.)
When I contacted NOAA to try to get the location corrected, this is the answer I received:

The latitude and longitude provided in HOMR come directly from the National Weather Service B-44 rendition… The most current rendition for Falfurrias TX was  effective 5/15/2013 and the coordinates on that rendition are 27.1354, -98.1202. I cannot make changes to the coordinates without a new rendition from NWS….”

Which brings me back to the title of this post….


NOTE: John Slayton is one of the most prolific and thorough volunteers with the surfacestations project. His work has helped us solve many mysteries on station siting. For novice readers, B-44 refers to the “B-44 form” used by NOAA to catalog the details around stations, such as this one from Santa Cruz, CA. Unfortunately, B-44 forms are considered “off limits” to the public by NCDC, for “privacy reasons” even though we can now look down directly into the back yards of COOP observers with advanced tools like Google Earth. I had once asked NCDC for access to the B-44 forms, and was denied. Maybe with instant look down capabilities in the hands of the public using the lat/lon coordinates provided by HOMR, the time has come to drop such silly policies to put the B44 forms online for research purposes. – Anthony

About these ads
This entry was posted in Weather_stations and tagged , , , , , . Bookmark the permalink.

16 Responses to Who are you going to believe, my B-44 or your lyin eyes?

  1. GlynnMhor says:

    I wonder how long ago the parking lot and associated buildings were installed, and whether the station had been more free of ‘urbanization’ in times past.

  2. David Ball says:

    Thank you to John Slayton for your efforts.

  3. Larry Ledwick (hotrod) says:

    They could easily create an appendix of annotations for the station locations without altering their current data base. As GPS becomes more prolific and widely used, many of these sites could easily be verified for location and updated just like the original surface station project documented quality and local environment of official stations. By compiling historical photographic evidence and verification of location by modern methods they would add value without having to “change” their current data base so it provides an accurate unadjusted snapshot of the original official information, but interested parties could go to the appendix of updates and find more precise location data and its provenance if needed.

  4. Larry Ledwick (hotrod) says:

    Ooops bit of a run on sentence there —-

  5. JohnWho says:

    Sadly, NOAA does not appear to care about accurate data.

    They seem to be in the “as long as we produce something and receive our paychecks” mode of operation.

    Besides, it is the predictions/projections that are important, not good data and proper science. /sarc

  6. DocMartyn says:

    This Thanksgiving my wife and I went for a little break to Padre Island, Corpus Christi, Texas.
    At the end of the pier there is a weather station and tide gauge.
    I really liked the way the temperature sensor was placed right behind the solar panel.



    Wonder if the temperature is a bit odd?

  7. 23skiddo says:

    Potential location accuracy on land would have been much better than a minute of angle, which is roughly about one nautical mile. The best sextants could, under ideal conditions, locate a point relative to the equator to about 0.2 minutes of angle, or roughly 370 meters (if I calculated the arithmetic properly). Using typical land-surveying techniques available at the time could easily have produced a far more precise location than even that (+/- inches). The coastal survey conducted extremely precise surveys of the coast of the US – important to anyone interested in sea level changes, since these are independent of tide gauges. So, it is likely that the nearest minute approach to station location was methodologically based, and depended on what was considered to be “necessary” precision for the purposes for which the station was established.

  8. Kev-in-Uk says:

    Kudos to Mr Slayton for his efforts – and of course our host Mr A Watts!! Seriously though – this is exactly how real science advances. i.e. through query and checking and indeed re-checking…..NOT by automatic ‘peer reviewed’ acceptance!
    Well done to all those who carry the genine and honest skeptical flag forward!

  9. jorgekafkazar says:

    Yes, kudos to John Slayton and Herb Sargo and the Brownsville Weather Service office and the two of their men who drove all the way up to Falfurrias. And to A. Watts! Your efforts are much appreciated.

  10. John F. Hultquist says:

    Hi John,
    Great post – and because you brought it up:

    “. . . or you can use the equestrian/hiking trail. (John Hultquist could probably give good advice about this….) I chose to walk.

    I think “to walk” is good advice. Although I’ve spent some days (maybe 20) working on the north side of Rainy Pass on the PCT on its way to Canada (and on a log bridge on Cutthroat Trail), I have not been to Stehekin. They do not call this region “Rainy” because of its sunny nature and riding and caring for horses in the rain is not my idea of fun.

    The trail head is here: 48.504839, -120.719236

    A 3 page report is here:

    Nancy says hi!

  11. Mike McMillan says:

    “Measure with a micrometer, cut with a chainsaw.”

    How much does precision matter when the temp data goes through the wood chipper of GISS/HadCRUT/NOAA/NCDC/USHCNvX “adjustments”?

  12. John Day says:

    >John Slayton is one of the most prolific and
    >thorough volunteers with the surfacestations project.

    I really admire Slayton’s zeal in getting the station location information. Reminds me of the dedicated volunteers in the NWS CWOP, who furnish their own equipment and time to provide data for mesoscale forecasting. There are over 7000 stations reporting daily, most sending reports every 10 minutes (144 reports a day).


    I’ll confess that the self-contained, wireless stations that are available today makes this relatively easy for us. But before these automated stations were available volunteers had to access their equipment in pretty remote places and often had to make their periodic eyeball observations under nasty conditions, logging the data manually.

    So even more credit to Mr. Slayton, who must still do it the hard way.

  13. Brian says:

    Anthony wrote, “…I would recommend you Google these coordinates, because the satellite photo resolution is a bit extraordinary.”
    A common mistake is to consider on-line “satellite view” images as actual “view from a satellite” images… the majority of the data is, in fact, AERIAL photography, taken at relatively low altitudes with massive cameras on the underside of… airplanes… There is NO WAY that the image Anthony reproduced was shot from a satellite, through 60 miles or more of image-distorting atmosphere. In land areas where real-estate or government interests justify the expense, aerial photography is periodically done, usually at two or three altitudes. Multiple images are stitched together with software, producing a somewhat “seamless” image. As you “zoom out” from the maximum “zoom in” view, you’ll see changes to the higher altitude “cover” shot, then a transition to an actual satellite view.

    Oh, and with iPhone maps and Google maps, if you express the latitude and longitude as fully qualified decimal degrees, these map-software engines will locate the exact spot. Where I live, many rural street addresses “do not compute” or become “destinations” that are significantly far away from the real location. If I go to a particular rural location, “drop an pin” (using my iPhone) and then extract the address for that pin, and then use that street address as a destination argument, my native iPhone mapping application routes me to a spot more than a mile away from where I was when I dropped the pin. But, if I extract the latitude and longitude (using a real GPS app, in my case, MotionX GPS) and use that as an argument in my native iPhone mapping application, it takes me to the exact spot. For example, this exact string, without the quotes:
    “21.509675, -157.829397″ (Latitude in decimal degrees, a comma, a space, a minus sign, and the longitude in decimal degrees) is the tiny island of Mokoli’i, a few hundred yards off of Oahu, Hawaii. The key to success in using latitude and longitude as an argument in the “Search or Address” field is using the minus sign in front of the latitude digits (instead of “S”) [a plus sign is implicit for "N" latitude] and separating the latitude (D.dd) from the longitude (D.dd) with a comma, a space, and the minus sign instead of “W”. Cut-n-paste the above latitude, longitude digits into your favorite mapping software’s argument-input field, just as if you typed in a street address. Here’s another example: “1858 3rd St, Langhorne, PA USA” … Apple or Google will drop a pin on the street, or, perhaps on the house at this address… but use the argument, “40.181278, -74.888928″ and you’ll see it is a spot in the BACK YARD of the same address. You can even drop “40.181278, -74.888928″ into the search field on the Google main page, and the first result will be a map with the spot correctly positioned. Select “satellite” or “hybrid”… zoom out, notice when the image changes from aerial to (true) satellite (often taken on a different day or different year, changes in weather, season, or visible construction become obvious) and then zoom in, you’ll (hopefully) see the transition from (true) satellite back to aerial, and then, as you zoom in, you might see another layer of aerial photography, taken from a lower altitude. Notice weather, cloud cover, sun-shadow angle, moving vehicle positions, perspective changes, indicating that you’re not zooming in and out of a particular photograph, but changing from one photograph to another. Zoom out so you see the whole continent, pick a really rural spot, far from valuable real-estate, and zoom in, and in and in, and you’ll see the true satellite image until the zoom fails. There, no one is willing to pay for aerial photography.

    REPLY: Another common mistake is to assume I wrote something when it says “Guest Essay by John Slayton” – Anthony

  14. phodges says:

    REPLY: Another common mistake is to assume I wrote something when it says “Guest Essay by John Slayton” – Anthony

    LOL… that’s the most amusing thing I’ve read on here today…thanks!

  15. “Sadly, some are getting up in years and will not be with us much longer.”
    Document the location with GPS and video and archive photos and the old guys themselves, and have this notarized. Preferably, bring the notary with you and have him taken on video during the recording of the thing. Use aerial pictures of the time, cross-match them with modern satellite pictures and GPS. Once they are recorded, and with archive photographic proof (triangulation with mountains and background features, etc), they will not be able to deny it. It does not matter if the exact location was established recently, as long as the ground location can be authentified.
    Do they think that earth coordinates have moved in the last 100 years? Well, with plates tectonics, you never know, don’t’ch’ya?

  16. juan slayton says:

    Thanks for your comment. I had noticed the changing appearance of some sites as you zoom in and out, but it didn’t occur to me that they were integrating satellite and aerial views. I should have been paying attention; when 3 tabs (I think they were ‘map,’ ‘satellite,’ and ‘birdseye/aerial’) coalesced to one (‘map,’ toggling with ‘satellite’) the clue was definitely there.

    Of course, our point is that the station and its location is plainly visible to anyone who will take the time to look at the published picture. This is valid, regardless of the picture’s source. Hey, I’d be happy if it was taken from a balloon. Like this 1910 picture of San Luis Obispo:

    js : > )

Comments are closed.