Another UFA sighted in Arizona

My post on Lampasas,TX has created quite a stir when Atmoz, a climate scientist unknown person at the University of Arizona, tried to demonstrate that the temperature spike shown in the GISS data at Lampasas, TX, was not due to the relocation next to a building and asphalt parking lot, but rather some problem with the GISS algorithm to do homogeneity adjustment to the data.

Steve McIntyre posted a tongue in cheek notice of Atmoz’ theory of unidentified faulty algorithm at NASA (UFA). Arizona already has the parking lot weather station operated by the Atmospheric Science Department of the University of Arizona.

Enter serendipity. Warren Meyers’ son Nicolas, has been actively surveying Arizona USHCN stations for his school science project. My inbox had a new station from him today, Miami, AZ. So I decided to take a look at it.

As is typical when an MMTS sensor gets installed by NOAA/NWS to replace the traditional Stevenson Screen, it got closer to human habitation, and in this case, a LOT closer. Too close I’d say:

miami_az_mmts.jpg
click for full sized and additional images at surfacestations.org database

So I though I’d take a look at the raw GISS temperature plot for Miami, AZ to see if the move would show a spike, it did:

miami_az_giss_raw520.png

From NCDC’s MMS database, they have a map showing station moves. This is at the Magma Copper Mine in Arizona, and the station used to be further away from the administration buildings near the pit:

miami_az_locations_map.jpg

Seeing a similar scenario to what occurred in Lampasas, TX, where a rural station was moved from a cooler location to a much warmer one, I decided to do the same sort of comparison on the GISS temperature plots as I did before:

RAW GISS DATA:

miami_az_giss_raw520.png

HOMOGENIZED GISS DATA:

miami_az_giss_homogen520.png
Note that I changed the color to red using a hue shift to prepare for the next step, to see the original GISS data, click on the image.

HOMGENIZED GISS DATA OVERLAID ON RAW DATA:

miami_az_giss_raw-homogen520.png

Notice that after the GISS homogeneity adjustment, the past temperatures go down, with the present acting as a hinge point, thus making the slope of the temperature trend rise. The new slope is purely artificial, and appears to be an artifact of data adjustment by NASA GISS on this rural station. This is the second instance of this happening, the first being seen in the GISS Lamapasas, TX data adjustment for homogeneity.

In both cases, the abnormal spike coinciding with a station move near the present time remains in the record, and that is what the homogeneity adjustment is supposed to catch and remove as I understand it.

In a comment on the subject, Steve Mosher offers an explanation:

In Hansen 2001 Hansen says he uses nightlights to determine
if a station is Rural in the US and population everywhere else.
Miles city population is less than 10K which makes it rural,
BUT, nightlights ( satellite imagery taken in 1995)
indicates a brightness factor for Miles of 26! effectively making it urban.

I concur, there appears to be a flaw in the GISS nightlight methodology and adjustment algorithm. I look forward to seeing GISS investigate, and if this problem is indeed verified, a dataset correction.

320 Comments

  1. Kristen Byrnes
    Posted Feb 17, 2008 at 11:22 AM | Permalink

    He’s not a climate scientist, he’s a student.

  2. Posted Feb 17, 2008 at 11:35 AM | Permalink

    Kristen, point taken but it’s hard to know for sure, he like some other critics we know publish under an anomymous name. I’ve edited the text accordingly.

  3. steven mosher
    Posted Feb 17, 2008 at 11:35 AM | Permalink

    Hey Anthony,

    You should add nightlights to your Spreadsheet.

    Also, Remember Quayle study on MMTS showing a significant cooling bias?

    What if you put a MMTS next to a CRS for 20 years.

    http://ams.confex.com/ams/pdfpapers/91613.pdf

  4. steven mosher
    Posted Feb 17, 2008 at 11:35 AM | Permalink

    Hey Anthony,

    You should add nightlights to your Spreadsheet.

    Also, Remember Quayle study on MMTS showing a significant cooling bias?

    What if you put a MMTS next to a CRS for 20 years.

    http://ams.confex.com/ams/pdfpapers/91613.pdf

  5. Posted Feb 17, 2008 at 11:38 AM | Permalink

    Mosh, if you are using a laptop and touchpad, that might explain your double postings, of which there are a lot. I’ve already deleted some. Try a mouse, it is what I use now because I had the same problem with my laptop touchpad.

    I agree, nightlights need to be added to the spreadsheet. Can you point me to a source?

  6. steven mosher
    Posted Feb 17, 2008 at 11:41 AM | Permalink

    Miami is a 19.

    Source is here

    http://data.giss.nasa.gov/gistemp/station_data/station_list.txt

  7. steven mosher
    Posted Feb 17, 2008 at 12:03 PM | Permalink

    Nightlights is a proxy for urbanization. Miami arizona, Population

  8. steven mosher
    Posted Feb 17, 2008 at 12:03 PM | Permalink

    Nightlights is a proxy for urbanization. Miami arizona, Population

  9. Brooks Hurd
    Posted Feb 17, 2008 at 12:48 PM | Permalink

    It would seem to me that the fact that a temperature station would be affected by its surroundings would be evident to most people who work around temperature indicators. Most people, at some time in their childhood, have stepped from cool green grass onto sun baked asphalt with bare feet. It does not require a huge jump of deductive reasoning to realize that moving a temperature monitor from a most locations to an asphalt parking lot next to a tin building will cause an increase in the temperature around the monitor, which will be reflected in an increase in the observed temperature. This leads me to the conclusion that we are either dealing with people who are incredibly naive, or we have people moving some of these temperature monitors with a motive for their actions.

    There is a third possibility; that we are dealing with idiots. I do not believe that this is a reasonable possibility. Whatever the reason behind these moves, thank you Anthony and all those who contribute to your efforts for bringing these issues to everyone’s attention.

  10. deadwood
    Posted Feb 17, 2008 at 1:10 PM | Permalink

    Brooks@9:

    The argument that the GISS apologizers use is that the volume of stations and the correction factors in the averaging algorithms compensate for site conditions, moves, etc.

    I think it is evident from these two stations that these corrections are not working – at least for these two stations.

    This is where the work of Mr. Watts and his volunteers can provide valuable reality checks for the scientists and policy makers who are using this data to support the multi-trillion dollar economy wrecking carbon fantasy coming soon to America.

    Whether Watts’ work will support or refute GISS remains to be seen, as does whether a refutation, assuming the study turns this way, will sway the alarmists or policy makers.

  11. Stephen Richards
    Posted Feb 17, 2008 at 1:15 PM | Permalink

    Anthony

    Love your what’s up site. Help and old idiot please ( bad combination that): When we are talking of 150yr anomaly trend of approx 0.55°C how the hell does GISS do a sensible adjustment for stations with a binary change of 2°C ? I do not see any possible way of adjusting temps with sufficient reliabilty to describe anomalies of the current order even a 1°C anomaly.

  12. Posted Feb 17, 2008 at 1:22 PM | Permalink

    RE11, Stephen, If I could offer an explanation, I’d be Jim Hansen. Right now, much of this is not well understood outside of his circle at GISS. Perhaps they have a valid method, perhaps it is flawed. I honestly don’t know yet.

    From my world view, being just a guy whose specialty is weather instrumentation and weather graphics/presentation, and finding such discrepancies illustrated above, I’m betting on the latter.

  13. Jon
    Posted Feb 17, 2008 at 1:23 PM | Permalink

    Perhaps one issue with the lights methodology is that this isn’t strictly speaking an UHI effect at all.

    Its a point beyond dispute that it is warmer nearer to buildings (at night, in the winter, etc) than away from them. This is true for a single cabin out in the wilderness too.

  14. steven mosher
    Posted Feb 17, 2008 at 1:37 PM | Permalink

    Anthony,

    Other proxies for Urbanization

    http://www.ngdc.noaa.gov/dmsp/download_sprawl.html

    And data with a pixel size of 1KM as opposed to the 2.7KM of hansens 1995 nighlights

    Perhaps Nighlights ( 95) as a proxy for urbanization can be tested?

  15. Steve McIntyre
    Posted Feb 17, 2008 at 1:44 PM | Permalink

    In the datafile http://data.climateaudit.org/data/ushcn/details.dat, I’ve collated GISS info on lights and brightness to USHCN identification numbers – try using that.

  16. Joe Black
    Posted Feb 17, 2008 at 1:45 PM | Permalink

    “I do not see any possible way of adjusting temps with sufficient reliabilty to describe anomalies of the current order even a 1°C anomaly.

  17. Stephen Richards
    Posted Feb 17, 2008 at 1:50 PM | Permalink

    Anthony

    Thank my little surface stations you are not Hansen II. That would be too much to bear and the current Jan 2008 cooling would be due to CO² global warming :)

  18. Posted Feb 17, 2008 at 1:53 PM | Permalink

    RE15, thanks Steve, like usual you are a step ahead of me, much appreciated.

  19. steven mosher
    Posted Feb 17, 2008 at 1:56 PM | Permalink

    Hansen2001 cites Imhoff for establishing the relationship between nightlights (1995)
    and urbanization. It turns out there are many studies most behind green walls

    http://www.ngdc.noaa.gov/dmsp/publications.html

    One abstract, looking at the problem after hansen

    Henderson, M., Yeh, E.T., Gong, P., Elvidge, C.D. Baugh, K., 2003, Validation of urban boundaries derived from global night-time satellite imagery. International Journal of Remote Sensing, v. 24, 595-609.

    Night-time imagery from the Defense Meteorological Satellite Program (DMSP) Operational Linescan System (OLS)
    has been proposed as a useful tool for monitoring urban expansion around the world, but determining appropriate
    light thresholds for delineating cities remains a challenge. In this paper we present a new approach.
    We used DMSP stable lights and radiance-calibrated images to delimit urban boundaries for San Francisco,
    Beijing and Lhasa, cities with different levels of urbanization and economic development, and compared the
    results against boundaries derived from high-resolution Landsat Thematic Mapper (TM) imagery.
    Unthresholded DMSP images exaggerate and shift the extent of these urban areas.
    We then calculated light thresholds that minimized the discrepancies between the
    DMSP- and TM-derived urban boundaries for each city.
    Our comparison highlights the difficulty of using DMSP data across areas with
    disparate urban characteristics, but suggests the possibility of calibrating this
    data source for monitoring growth of cities at comparable levels of development.

    Lots of papers, lots of data ( on things other than population that could be proxies)

    Question: Did Hansen use Unthreasholded DMSP images? Shifting and exagerating the extent of
    Urban Areas. Simply, there are some sites, like Miles City, that LOOK RURAL from the ground
    but from space, they get a peri urban brightness index. Unthreasholded DMSP?

  20. Joe Black
    Posted Feb 17, 2008 at 2:08 PM | Permalink

    “I do not see any possible way of adjusting temps with sufficient reliabilty to describe anomalies of the current order even a 1°C anomaly.”

    I am still waiting for an adequate definition of just exactly what system is being measured and predicted.

    It would seem that some are trying to measure the air temperature differences from year to year in a band 4 to 5 ft above the earth’s (over land) surface in places with similar and constant land use. These data are combined with (randomly selected?) sea surface temperature differences and somehow combined to estimate the Earth’s energy balance.

    Why not include the effects of land use changes and human population increase? People seem to be dying in cities when the temperature in the UHI goes high (see France a couple of years ago). Why not control the in-city temperatures? Orchard temperatures?

    Why not compare the subsurface land temperatures (many ag stations seem to have historical data) changes rather than that of the air (in a limited height band)? Why not measure the ocean temperatures in depressions where the densest water might be lurking? Why not just try to measure and control the energy balance in the upper atmosphere? Where’s the rainfall history? Soil moisture levels? Is global warming global if all temps changes are not the same?

    It seems to me that most of the “climate change” discussion is similar to the committee of blind folks examining the Elephant. A good system definition couldn’t hurt.

  21. steven mosher
    Posted Feb 17, 2008 at 2:28 PM | Permalink

    here’s a quick look at the problem of using nightlights to define Urban

    http://www.geography.du.edu/sutton/SuttonHome_files/PDFsOfPapers/UrbanSprawlinRSE.pdf

    http://www.geography.du.edu/sutton/SuttonHome_files/PDFsOfPapers/AmbientPopDenPEandRS.pdf

    Question: can it define “rural”

  22. Demesure
    Posted Feb 17, 2008 at 2:32 PM | Permalink

    Some nice “adjustments” in France by the GISS. Looking at the step changes with regular intervals (8 years for Dijon or Marseille), you don’t need stations’ photographs to know it’s UFA.

  23. steven mosher
    Posted Feb 17, 2008 at 2:39 PM | Permalink

    ok.. no more posting after this…

    http://www.geography.du.edu/sutton/SuttonHome_files/PDFsOfPapers/CEandUSmodPopDen.pdf

  24. Demesure
    Posted Feb 17, 2008 at 2:56 PM | Permalink

    Here is a plot of the above Miami station (ID=722780040) with the difference “homogeinity adjusted” – “not adjusted” in the second graph: again, one must wonder where the “adjustment” step changes (3x 24 years) come from !

  25. steven mosher
    Posted Feb 17, 2008 at 3:37 PM | Permalink

    RE 24. Since miami is rated as a Nighlights 19, it falls into the PERI URBAN category
    (8-88) As such, Miami will be adjusted to match the trend data of the “rural” stations
    around it ( within 500 or 1000km)

    The adjustment is a curve fit ( with one or two legs) The adjustment is made in discrete
    steps.. which is why you se the staircasing.

    In short, this station gets 1 adjustment ( for UHI) and that adjustment is trend adjustment
    that is “fed” in by steps.

    get it?

  26. Demesure
    Posted Feb 17, 2008 at 4:13 PM | Permalink

    @25
    OK Steven for the discretization of trends. Now that I know the software programming explanation (not really a scoop in fact), I would like to know the physical explanation ;)

  27. John F. Pittman
    Posted Feb 17, 2008 at 4:40 PM | Permalink

    Why is one done on about 7 years steps and the other about 14 year steps? Why use a step that changes frequency?

  28. Jim Edwards
    Posted Feb 17, 2008 at 4:49 PM | Permalink

    Brooks Hurd, #9:

    I offer an alternative [completely speculative] hypothesis for a pattern of poor site selection for MMTS retrofits.

    1. Science / academic types who can’t identify the working end of a screwdriver saw a neato hi-tech sensor and advocated for a budget appropriation to upgrade the temperature measurement network. Half of them didn’t understand the distance constraints of the sensors, the other half [w/ zero idea how much it costs to build anything...] naively believed the budget would include enough money to run utilities, et. Al. to old locations.

    2. A budget was passed and speeches were made about a great investment in data quality for the future.

    3. Government procurement and contracting regulations took over.

    4. Some astute electrical contractor noticed the MMTS cable length limitations and reported it back to the contract administrator, hoping to get a costly change order added to the original contract.

    5. The contract administrator did some phone calling and multiplication and realized that a bunch of bureaucrats and politicians had already taken a lot of credit for a project that was at least 50% underbudgeted.

    6. A low to mid-level, non-technical bureaucrat made an executive decision to have the contractors move sensors inappropriately close to government owned, contracted, or friendly buildings. He’s never heard of UHI.

  29. thomas
    Posted Feb 17, 2008 at 4:59 PM | Permalink

    when were the lampassas and miami stations moved? what year

  30. Posted Feb 17, 2008 at 5:13 PM | Permalink

    Last year I surveyed a station where the curator noted that he used to walk several hundred feet, in all kinds of weather, to take readings from the old screen, because he’d been told that distance was important. Then, when the MMTS arrived and he no longer had to go outside in the wet and the heat and the cold, they stuck the sensor just twenty feet from the door of the radio station.

    He marveled at the irony.

  31. steven mosher
    Posted Feb 17, 2008 at 5:17 PM | Permalink

    RE 26. Shrugs. dont ask me to defend their stuff. Do I look like your cousin Vinnie?

  32. Posted Feb 17, 2008 at 5:20 PM | Permalink

    I recalled a USHCN station I visted last year that was really far removed from civilization and time. Cedarville, CA in the far NE corner of the state. See my writeup here.

    I plotted the GISS provided USHCN data and the homgenized data from the GISTEMP website and the graph for that is shown below:


    click for a larger image

    Note the missing year at 1957, 1915 should have not plotted also, but for some reason my program insists on doing so.

    But the important thing is that once again, GISS has made the past colder and the present is unchanged. Even more odd, the far past prior to 1900 is adjusted upwards, warmer.

  33. steven mosher
    Posted Feb 17, 2008 at 5:49 PM | Permalink

    RE 27. The “adjustment has two legs” Or rather a “midpoint” at
    which the adjustments are made.

    ( they are trying to MATCH slopes if you look at the math in the code)

    Nudging stuff.

    Prior to 2001 paper, this Midpoint was FIXED as 1950.

    So, you calculate the SLOPE of all the rural stations nearby
    your Urban station between 1880 and 1950. Then you force the urban
    slope to match the rural slope. Then you do the same for 1950 to the Present

    in his 2001 paper, hansen notes that he makes this “hinge point”
    variable… What does this mean? the hinge point varies in time.
    So you search for a hinge point that minimizes error kinda sorta
    ( still struggling with this of code)

    It’s a way of kinda sorta saying that 1950 wasnt the year we all went
    urban.

    Sorry I cant be more specific, yet. gimme time

  34. steven mosher
    Posted Feb 17, 2008 at 6:11 PM | Permalink

    RE 15 and 18. SteveMc and Anthony.

    I took SteveMcs data ( lights stations etc)

    Then I asked: What would Deming do.

    Lights= DARK
    Pop= R
    Airport = NO (x)
    Bright =0.

    My rationale. Nightlights could confuse urban with Peri urban
    But dark is dark. Pop=Rural. This is a TIGHTER restriction than hansen.
    For purely rural it has to be lights =0 and Population

  35. steven mosher
    Posted Feb 17, 2008 at 7:13 PM | Permalink

    RE 32. I like the idea of combining nightlights =0 and population =R and Airport = false and CRN =1 &2. to pick out

    Good sites.

    Right now, you got 7 of those!. that is 7 sites in the whole US fulfill these criteria.

    Anthony, nobody can accuse you of cherry picking anymore. In the beginning folks said you would not travel
    outside your comfort zone. I think that assertion has been Starbucked.

    So, How about a list on Surface stations of MOST WANTED.

    Here is what I mean.

    Your critics say you dont highlight good sites. Ok.

    TARGET THE GOOD SITES. go after them. by my count there 210 sites or so that meet
    the criteria.

    1. Lights =0
    2. Population = Rural.
    3. Airport = No

    Just a thought.

  36. jeez
    Posted Feb 17, 2008 at 8:39 PM | Permalink

    It is probably not a coincidence that these downward adjustments seem to peak around the 30’s. I’m not saying Hansen is doing this intentionally, but if he “believes” that it is now warmer, unconscious bias would have him preferentially settle on algorithms that confirm the 30’s was not as warm as today.

  37. sigma
    Posted Feb 17, 2008 at 8:54 PM | Permalink

    snip

    If as I understand it there are sufficient “good” sites available, why does the data need to be contaminated with “adjusted” data.
    This of course raises a problem that the lies at the heart of the AGW theory, ie that the “observed” warming which is partly the result of a model calculation is consistent with another model that predicted it (GCM’s). In my field of atmospheric study our model outputs are compared to the raw “real world” measurements. If there is a difference then either the model or the data is wrong (or both). But you dont use two models and say that because they are consistent then they are therefore real. I come back to my main point, why use contaminated data in the first place. Personally I doubt that it is due to any agenda, but for my mind it is sloppy practice. If the “real world” data is faulty, dont use it, and dont try to fix it with another model.

  38. eric mcfarland
    Posted Feb 17, 2008 at 9:20 PM | Permalink

    This entire discussion lacks context. If you were really auditing (and not spinning) you’d contact the agency and find out why they placed the device in the shown location. Also, the lack of foundation and authenticating evidence is troubling. Consequently, this discussion is half cocked … at best.

  39. Dennis Wingo
    Posted Feb 17, 2008 at 9:26 PM | Permalink

    On a related topic, the oil companies are putting climate monitoring systems on their offshore oil wells.

    http://ioos.noaa.gov/

  40. steven mosher
    Posted Feb 17, 2008 at 9:40 PM | Permalink

    RE 38. this discussion is half cocked. You’re the expert there take over

  41. John A
    Posted Feb 17, 2008 at 10:06 PM | Permalink

    Dennis,

    Any chance they could be mounting the sensors on the boom that flares off the excess gas?

  42. jae
    Posted Feb 17, 2008 at 10:07 PM | Permalink

    I really dig the Algorithms; they help the science “move on.”

  43. Joe Perth
    Posted Feb 17, 2008 at 10:23 PM | Permalink

    First post ever on CA but I have to confess to being a long-time lurker. Enjoy the site and technical discussions.

    Here’s a question on the MMTS Temp Sensor shown in the image and the accompanying temperature graph. The biggest upkick in the graph is in the late 1990’s whereas the buildings look like they have been around a fair while. They look rusty. Does anyone know when these buildings were erected or when the paved areas were installed? Does their erection coincide with the upkick in temperatures? I know how long it takes a shed in Australia to go rusty and it’s longer than 10 years, unless coastal. This level of micro-detail may be required to fully comprehend the temperature history of each location and influences thereon. Big job. Is anyone doing the same QC on weather stations elsewhere in the world?

  44. hswiseman
    Posted Feb 17, 2008 at 10:30 PM | Permalink

    RE 35:
    I remember Gavin commenting here in an ultra-rare appearance (I think in a set-to with Anthony)that he only needed a handful of good sites to get a valid surface temperature grid. If a couple hundred “good” sites can be audited and confirmed it would go a long way to isolating and identifying UHI effects and validating (or falsifying) temperature adjustment formulas. If the Man wanted solid data, we would have it. The fact that the observational science is so weak, and the the howls of protest so non-existent, tells you everyting you need to know. Real scientists would have chopped off the fingers of anyone who messed with their instruments they way Anthony’s work has revealed. Needless to say, everyone in climate science has all their digits.

  45. deadwood
    Posted Feb 17, 2008 at 11:03 PM | Permalink

    hswiseman@44:

    Anthony notes that the station was moved. I don’t see the date, but he attributes the sharp rise to the move. So the age of the mining building is not relevant.

  46. deadwood
    Posted Feb 17, 2008 at 11:04 PM | Permalink

    Sorry that should have been addressed to Joe@43

  47. Joe Perth
    Posted Feb 17, 2008 at 11:25 PM | Permalink

    Deadwood 46.

    Point taken but it would still be nice to have dates of all relevant events such as station shifts, bitumenisation of parking lots, new buildings etc displayed on the temp graph. Sudden spikes in temperature would then stick out like, well, you know what.
    snip

  48. Joe Perth
    Posted Feb 17, 2008 at 11:25 PM | Permalink

    Deadwood 46, 45.

    Point taken but it would still be nice to have dates of all relevant events such as station shifts, bitumenisation of parking lots, new buildings etc displayed on the temp graph. Sudden spikes in temperature would then stick out like, well, you know what.

    BHP is the company here. I used to work for them. I know how focussed they are on accurate temperature records. A real business driver. Sarcasm off.

  49. Anthony Watts
    Posted Feb 17, 2008 at 11:28 PM | Permalink

    RE45, Deadwood, the NCDC MMS database says that is was moved to that location (near the admin buildings) on 1996-02-21 and then again to it’s closer present location on 2007-06-26.

    For Eric McFarland, we’ve had the discussion before with NWS/NOAA etc about why they place things where they do and the answer is always the same: time, equipment, money, convenience of the observer.

    Call up the COOP manager of WSFO in Phoenix, Keith Kincaid, and ask him yourself, then report back. Then please tell us why you think the location next to the metal building is acceptable. At the same time ask why NOAA’s own 100 foot rule was violated. It is not so much the “why”, but the fact that the station is in violation of a specification that is the most important.

    As for “…the lack of foundation and authenticating evidence is troubling.” Troubling to you perhaps, since its obvious that all you have to offer is barbs and that you’ve not done your homework. Any of the above can be authenticated by anyone, it is all open for inspection.

    For starters, read the survey report in the surfacestations.org database for this station. You’ll find the contact info for the surveyor, GPS coordinates, and everything you need to verify the location.

    Please, go do that and report back. If there are errors or omissions they will be corrected.

  50. G Alston
    Posted Feb 18, 2008 at 1:43 AM | Permalink

    #33 Steve Mosher (It’s a way of kinda sorta saying that 1950 wasnt the year we all went
    urban.)

    This strangely corresponds to the original expose of the use of whitewash on the
    screens. Isn’t that about the time estimated for the changeover to paints other
    than whitewash? What else happened physically that would cause 1950 to be used as
    a point?

  51. John A
    Posted Feb 18, 2008 at 6:18 AM | Permalink

    Anthony:

    Referring to the first photograph – what the buildings made of or cladded with?

  52. BarryW
    Posted Feb 18, 2008 at 6:28 AM | Permalink

    Re 35

    Airports have sizes just like towns have population, so while I agree that ruling out airports would give the most stringent restriction, you’d be ruling out small airports that have little traffic or concrete. Tower classification might be a way to winnow out the ones that are likely to be heat islands.

  53. steven mosher
    Posted Feb 18, 2008 at 7:14 AM | Permalink

    re 52, the change from whitewash to latex started in 1979

  54. Andrey Levin
    Posted Feb 18, 2008 at 7:16 AM | Permalink

    Is it possible to compare station data with satellite measurements?

  55. steven mosher
    Posted Feb 18, 2008 at 8:19 AM | Permalink

    stevemc. here is an odditty. in your tables brightness =0 is sometimes dark and sometimes dim!

    h2001 says that 0-8 is dark, 8-88 is dim and 89+ is bright

    Steve: These aren’t my classifications. They are concordances of two different NASA GISS information sets.

  56. steven mosher
    Posted Feb 18, 2008 at 8:42 AM | Permalink

    ok i’m looking for the place where hansen decides what is dark, dim and bright. some places with a brightness
    index of 0 are classified as bright, and some with indexs of 21 are dark

  57. Anthony Watts
    Posted Feb 18, 2008 at 8:56 AM | Permalink

    RE53, that’s when the official change was published in the observers handbook, but I beleive many observers changed to Latex well before that.

    RE51, it appears the are cladded with corrugated steel siding.

  58. steven mosher
    Posted Feb 18, 2008 at 9:40 AM | Permalink

    Anthony, I still struggling with the source for 2 things.

    1. Hansen’s criteria of dark,dim,bright. I think he takes this from Imhoff’s study but I have
    no data source. What we have is a file that GISS reads in with A,B,C ( unlit, dim, bright)
    He says he follows Imhoff and sites two Imhoff papers in Hansen2001. I cant get at the Imhoff
    papers. Simply, at some stage in this process nightlights information was turned into 3
    categories unlit/dark, dim, bright. I have not found the code for that.

    2. Brightness index. This is in the input files…. BUT

    A. what is the source.
    B. the code doesn’t access this field as far as I can see.

    I know I read something about this index somewhere but cant recall.

    If you are working on checking nightlights then the Imhoff papers ( see H2001 or the other links I gave)
    is a good place to start

  59. Dennis Wingo
    Posted Feb 18, 2008 at 9:42 AM | Permalink

    Any chance they could be mounting the sensors on the boom that flares off the excess gas?

    John A

    Yer funny.

    I am quite sure that Dr. Hansen and other NASA scientists will execute the most rigorous due diligence to establish the quality of the data from these stations.

  60. Posted Feb 18, 2008 at 10:20 AM | Permalink

    RE58 Mosh I’ll see if I can get at the Imhoof paper, I have ways.

    RE59, mounting a climate monitor on an oil platform is basically a CRN5 right off the bat. If they mount it at work level, then depending on which way the wind blows that day will detrmine if the sensor gets fresh sea air or air that has passed across the platform first which may also include waste heat and exhaust from operations.

    This idea is just nuts. Buoys do a much better job without being heat islands.
    Here is the original press release from Shell and NOAA (never thought I’d see those two together): http://ioos.noaa.gov/pdfs/NOAA_ShellOceanObs.pdf

    gotta love this bullet point:

    Upgrade all weather stations on four Shell platforms to include direct transmission to NOAA’s geostationary satellites (GOES), and emergency power ensuring an uninterrupted stream of information even if the platform is evacuated.

    So if the platform is on fire, they’ll be able to get the weather data. Thank goodness. Probably more interested in hurricane wind monitoring than climate here, which is the most common evac reason.

    NOAA met buoys do a better job of this without being thier own heat island like an oil platform:

    http://www.vos.noaa.gov/MWL/aug_07/thunderbay.shtml

    This pact with Shell seems more like a feel good project.

  61. Bruce
    Posted Feb 18, 2008 at 10:22 AM | Permalink

    Do any of Hansen’s “corrections” ever go against the AGW cult? Or do they all make the past cooler so the present seems warmer.

  62. steven mosher
    Posted Feb 18, 2008 at 11:05 AM | Permalink

    RE 60… I have something EVEN better

    http://ams.allenpress.com/archive/1520-0450/38/6/pdf/i1520-0450-38-6-806.pdf

    A vegatative index.

    Now, Take SteveMc tables of stations.

    Eliminate the Dim and the Bright, just use the dark

    Eliminate stations that are not Rural ( there are few small towns that show up)

    Eliminate Stations that Have a brightness index >0

    Elimate Stations that are closed.

    Elimate stations on the coast.

    That leaves you with 197 stations.

    IN that group you have surveyed 47. there is 1 CRN1, and 5 CRN2.

    So, if you want to use multiple methods to pick the the best, you have around 200.

    NOW,

    wind the clock back to JohnV studys of CRN123R… That needed to be CRN123DARK
    because dark is the feild hansen uses to determine rural in the US, not the rural feild.

    complicated

  63. MarkW
    Posted Feb 18, 2008 at 11:23 AM | Permalink

    Anthony,

    What if the station is put on one of the legs, well below the platform?

  64. steven mosher
    Posted Feb 18, 2008 at 12:22 PM | Permalink

    RE 61, Sites are made cooler in the past by the following logic.

    1. Only Dim and Bright sites are supposed to be adjusted. These sites
    are assumed to be infected with UHI.

    2. Each Bright or Dim site is then Compared with its rural neighbors, within 500km
    and then 1000km if required.

    URBANITY is determined by 1995 data. If A site is urban in 1995 it is always treated
    as such, even in 1910.

    Lets suppose that a Bright site ( supposedly Urban) showed a warming trend of 2C over early part of the century.
    Suppose its rural neighbors only showed a 1C warming trend. The URBAN SITE would then be cooled to
    match the Unlit or rural site.

    So, take Vegas for example, it would listed as BRIGHT in 1995. This would mean it is treated
    by the algorithm as URBAN throughout its history. And in 1930 when it had 5,000 people it’s record
    would be adjusted to match the record of surrounding Rural sites, under the assumption that if Las vegas is
    BRIGHT and Urban in 1995 it was Urban in 1930.

    So sites that are rural today are assumed to have always been rural. They are used to adjust sites
    that today are urban, but may have been rural in the past.

    By cooling the past the warming trend increases and the anomaly drops

    Interesting question: how many records in 1934 have a cooling or warming adjustment applied. How many records
    in 1998 and 2005 have a cooling or warming adjustment applied

    The initial USHCN approach was to collect population data for every place from the historical census
    but the data coverage was only about 70% I believe

    ( I find the refernce if you like,)

  65. Anthony Watts
    Posted Feb 18, 2008 at 12:46 PM | Permalink

    RE64, Jeez Mosh, did you sleep last night?

    Now the curtain is being drawn back to see the machinations of the algoritms operating in the dim light.

  66. eric mcfarland
    Posted Feb 18, 2008 at 1:24 PM | Permalink

    I have submitted a request to NOAA on this particular device/location. Let’s see what I get back in response … May take a couple of weeks … in all events.

  67. Mike B
    Posted Feb 18, 2008 at 1:55 PM | Permalink

    I have submitted a request to NOAA on this particular device/location. Let’s see what I get back in response … May take a couple of weeks … in all events.

    Would you mind sharing the nature/content of your alleged request? ;-)

  68. Anthony Watts
    Posted Feb 18, 2008 at 1:59 PM | Permalink

    RE66 Eric, I welcome that info, thank you. My experience so far has been that NWS/NOAA tends to ignore these requests, such as the one I submitted to the MIS in Key West which has gone unanswered for weeks.

    Some of my surveyors have gotten responses, so there is a chance you will too.

    I truly hope you will get a response.

  69. steven mosher
    Posted Feb 18, 2008 at 2:24 PM | Permalink

    re 65. sleep? ha. The human disguise I wore when we met was successful!

    I nap.

  70. steven mosher
    Posted Feb 18, 2008 at 2:43 PM | Permalink

    re 66. Eric why wait two weeks for a response. All the data you need is online
    or described in the papers.

  71. Sam Urbinto
    Posted Feb 18, 2008 at 5:00 PM | Permalink

    2 C? I’d put an error of more like 5-10 C.

    Mosh: “What if you put a MMTS next to a CRS for 20 years.”

    By then AGW will have killed us all; we can’t afford to wait. Think of the children.

    Or a better answer, for somebody to put the two together at once (along with another CRS nearby which is placed to be accurate as to climate and need no adjustments as a control) they would have to actually be interested in learning if the other station is accurate rather than just interested in massaging the past to make now “warmer”.

    But you already know what I think of the anomaly trend in the first place. A by-product of instrumentation changes and the averaging tmean over and over process. Even if tmean is accurate and meaningful, it’s still just some number being attributed to the daily temperature changing over time on an averaged month-to-month basis.

  72. steven mosher
    Posted Feb 18, 2008 at 6:01 PM | Permalink

    RE 71. I posted the study around here somewhere. I’m napping now. Dont bug me

  73. John A
    Posted Feb 18, 2008 at 7:03 PM | Permalink

    Re #57:

    RE51, it appears the are cladded with corrugated steel siding.

    So the temperature sensor is put in a south-facing, iron clad solar oven. Is anyone here surprised that temperatures are rising?

    I know that Steve is anxious for us commenters to calm down with our conversation on this blog, but I’m infuriated by the low quality of science coming from GISS.

  74. eric mcfarland
    Posted Feb 18, 2008 at 7:19 PM | Permalink

    This particular instrument may be a COOP instrument, or so I am told. It would have ultimately been selected on the basis of a survey. In all events, that’s a very tentative response. I have been directed to communicate with Asheville, which I will do next week.

  75. sigma
    Posted Feb 18, 2008 at 7:32 PM | Permalink

    A question for Steve (perhaps slightly off topic). As i understand it, the method of correction for UHI effects is to adjust using a nearby rural site. Now for the purposes of calculating trends and other correlations, do they (IPCC) et al then use the full set of stations that includes the rural and urban sites that have been corrected using the rural sites (I dont know). If they do, then surely the temp variables cease to be independent, and from my basic knowledge of stats, will introduce enhanced correlations (pos and neg). In simple terms, isnt this the same as doubling the number of sites that show a similar trend when in fact they should be independent?

  76. eric mcfarland
    Posted Feb 18, 2008 at 8:24 PM | Permalink

    http://www4.ncdc.noaa.gov/cgi-win/wwcgi.dll?wwDI~StnSrch~StnID~10100054#DIGITAL

    ??

  77. eric mcfarland
    Posted Feb 18, 2008 at 9:01 PM | Permalink

    76:
    By my reckoning, the site may actually be across the road. Am I missing something?

  78. Geoff Sherrington
    Posted Feb 18, 2008 at 9:02 PM | Permalink

    The header aerial photo of Magma copper mine is an excellent example of why lights and population might not correlate well, especially for a 24/7 operation. There’s a lot of energy consumption per unit area also and this mostly ends up as heat.

    It highlights the problem with surface temperature adjustment using general formulae. It has been shown time and time again to produce results open to question. Temperature is obviously one of the building bricks of global climate change. Until it is got right, there is not much point in pressing on. Surely, station-by-station methods have to replace generalisation.

  79. Anthony Watts
    Posted Feb 18, 2008 at 10:06 PM | Permalink

    RE77, Eric,

    What you are seeing is the fact that NOAA does not record the lat/lon in their NCDC MMS database with great accuracy. Many entries are often simply wrong, and our volunteer surveyors have to become CSI’s to figure out where the station actually is located.

    In some cases, we’ll find stations hundreds of yards or miles away from the locations described by the database.

    One such station was Bly, OR which is a USHCN station #350854 at 42.43333 -121.1 or so you’d think.

    Those coordinates put you north of town, and I spent a whole Sunday afternoon last fall looking for it. After finally giving up, I drove south to my next destination, Lakeview, on Highway 140.

    As I drove along the highway I spotted a Stevenson Screen in the front yard of a farmhouse, came to an abrupt halt, backed up and drove in. The owner met me and I asked him “is this the official NWS COOP station for Bly?”. He said “yep” and I asked if I could talk to him about it. He was actually very happy to do so, and even showed me the newborn calf in his barn he was so proud of.

    Turns out, the previous observer had died, and they took over the observing. The NWS created a NEW data entry in the MMS for Bly, calling it Bly 4 SE, at lat/lon 42.2209 -120.5792 where the previous station was called Bly 3NW. Those designations denote the distance and direction from the post office, which NWS uses as a benchmark.

    They gave the new station a COOP ID of 350856 but NOWHERE in the NCDC MMS database do they show any connection or continuation between Bly 3NW #350854 and Bly 4SE #350856

    The complete survey of the new station is here

    So as many of my volunteer observers can attest, the NCDC MMS database while useful, is not always updated, accurate, or relevant to the station at hand. In somes cases it is dead on, in others, it leaves a trail of confusing clues for the surveyor to interpret and follow.

    Case in point: NASA GISS isn’t even aware of the new station relocation, as they still list Bly 3NW but not Bly 4SE. From the GISTEMP plot, it is clear that GISS thinks the station is closed, see the plot here:

    http://data.giss.nasa.gov/cgi-bin/gistemp/gistemp_station.py?id=425725970090&data_set=1&num_neighbors=1

    Go ahead search for other stations nearby in GISS, the new one isn’t there. It rather tickles me that GISS doesn’t know this and I do. But, how could they? NCDC makes no mention of the continuance at the new location. They only way I found out is by being there to inspect the site in person.

    Some stations are never found, like the one we keep chasing in Dover, DE. One of my volunteers, Russ Steele, spent a whole day looking for it. The lat lon put it way north of town but the location description puts it elsewhere. The state climatologist there we asked for help hasn’t found it, yet it still produces data. We think it is at the state highway garage but no employees queried there know anything about it…which isn’t surprising, since it often falls to one person to do the job and other employees may not be aware of it. The state climatologist hopes to pay a visit soon to settle the issue.

    So the moral of this story is: don’t take the station metadata at NCDC for gospel. Sometimes digging and CSI work is required to find what has been lost in record keeping or simply entered wrong.

  80. eric mcfarland
    Posted Feb 18, 2008 at 10:17 PM | Permalink

    79: Wow. Many foundational problems … indeed. I appreciate your hard work … but, for the sake of my own mind, I shall continue to dig.

  81. Posted Feb 18, 2008 at 10:38 PM | Permalink

    RE80, Digging is good. I shovel daily.

    I know it seems incredulous that there could be such problems. I know many think that the kinds of problems we find must surely be made up or exaggerated. I also know firsthand what is the reality of the current network.

    The only way to know, is to experience it yourself, may I suggest you try surveying a station in your area? I could suggest one if you give me a city or a regional area.

    or just pick one from the list: http://www.surfacestations.org/USHCN_stationlist.htm

    The list will be updated soon, so if unsure about a station just ask.

  82. John A
    Posted Feb 18, 2008 at 10:59 PM | Permalink

    I forgot to add: …and that steel railing to protect it from traffic should warm the sensor up nicely as well.

  83. Posted Feb 18, 2008 at 11:16 PM | Permalink

    Eric,

    As one of Anthony’s volunteers, let me share some of my experiences in Dover Delaware and Panguitch Utah in finding surface stations. First the Panguitch story can be found here and here. The station had been moved to a local citizens garage in 2004 and the NCDC metadata had never been updated in three years. After we returned home to California it took me about ten phone calls to find someone who knew where the shelter had been stored.

    While on vacation in Dover Delaware visiting my wife’s sister and brother-in-law we went in search of the Dover surface station. I put the coordinates from the NCDC data in my Garmin GPS set. However, I was a bit concerned as the coordinates on Google Earth pointed to a plowed field. The description in the MMS data base said the surface station was at HWY DEPT ON RTE 113 WITHIN AND 1.2 MI SE OF PO AT DOVER, the listed address. Regardless of all the new highways in Dover, the interconnections and confusing signage, we finally located the Highway Department. But, no one knew where the weather station was. A kind individual, a local IT guy, call three different offices who he thought would know where the station was. No joy!

    We finally located a Davis weather station on the corner of the Central District Headquarters building at 930 Public Safety Blvd., but everyone we asked claimed it was only for DeDOT use, no connection to the NWS. So, we decided to follow the GPS, but there were no access roads leading to the location given by the GPS in a plowed field and we could not find any evidence of shelters near by farm buildings.

    After we returned to California Anthony, called David Legates the State Climatologist and he was equally baffled as to the station’s location, and as far as we know he is still looking for Station 72730.

    Here is a link to my original post on the Dover site with some pictures.

    Hunting for surface station can be a real challenge, especially when the data in the NCDC metadata is wrong, or worse misleading. My wife and I have invested many hours in hunting for surface stations. It has turned into challenging game of hide and seek for us. The NWS hides the station and we find them, even when the clues in the MMS metadata are wrong. We are planning a cross county trip in April to seek out more hidden stations. Stay tuned.

  84. Raven
    Posted Feb 18, 2008 at 11:37 PM | Permalink

    Would someone working for NCDC have the same problem finding the stations or do they have access to records (i.e. personal records) that would make this job easier?

  85. eric mcfarland
    Posted Feb 19, 2008 at 12:15 AM | Permalink

    83: From your Dover article, it appears that you are locating stations based, ultimately, upon your opinion:

    “After looking at the location of the Department of Transportation on Google Maps, I can only conclude that the current surface station is the one on the DeDOT Hq Building, regardless of the claims to the contrary by the Assistant Weather Emergency Coordinator, and her supervisor.”

    How often do you do that … is it a regular practice?

  86. Posted Feb 19, 2008 at 12:46 AM | Permalink

    Eric@85,

    No it is not regular practice, this was an opinion posted on my blog. Did you read the update? If go to the surface station Dover listing you will find it is empty. The exact location of the Dover stations is still an open issue. One thing we do know, it is not where the MMS metadata said it was. We are still looking for it with the help of David Legates.

  87. MarkW
    Posted Feb 19, 2008 at 6:07 AM | Permalink

    Eric,

    Since the station remains unsurveyed, hence there is no data for it in Anthony’s survey, why does Russ’s opinion of where it might be matter?
    When the station is finally located, then we will find out if Russ’s intuition is right, or wrong. That won’t affect the survey, just Russ’s ego.

  88. Joe Black
    Posted Feb 19, 2008 at 6:20 AM | Permalink

    Eric mcfarland says:

    February 19th, 2008 at 12:15 am

    83: From your Dover article, it appears that you are locating stations based, ultimately, upon your opinion:

    “After looking at the location of the Department of Transportation on Google Maps, I can only conclude that the current surface station is the one on the DeDOT Hq Building, regardless of the claims to the contrary by the Assistant Weather Emergency Coordinator, and her supervisor.”

    How often do you do that … is it a regular practice?


    I’d respectfully suggest your get off your high horse, get your butt in gear and go survey a few stations. When you have done ten or so, please come back and provide your $0.02 about the process.

  89. Anthony Watts
    Posted Feb 19, 2008 at 10:32 AM | Permalink

    RE84 Raven,

    NCDC has a whole wealth of information not available to the public, including detailed driving directions to the station, phone numbers of the curator and street address of the stations.

    They won’t share this info because of “privacy concerns” so leave us to figure it out by logic and deduction. What is funny is that while NCDC cites “privacy concerns” they publish the photos of private citizen observers along with their station info in their quarterly COOP newsletter and also on many NWS WSFO websites.

    Dr. Roger Pielke Sr. just gave the surfacestations.org project a huge endorsement.

  90. Sam Urbinto
    Posted Feb 19, 2008 at 11:50 AM | Permalink

    Eric’s just asking questions too quickly. Obviously, many people wouldn’t believe that the surface network isn’t as organized as it should be, and that we have the siting issues, and different agencies not working together as effectivly as they could. Unless you’re familiar with how the government works or how climate science works. :) Think of it this way Eric, if you are told a station
    in India is at Latitude: 22.111155 Longitude: 80.222255 about 7 KM SE of Lamta around 400 KM SE of Bhopal, and you go there and there’s no station; is it your fault it’s not where it’s supposed to be and nobody knows where it is? If it’s not there it’s not there; what else would you do but ask around, look around, and guess? So it’s not “opinion” it’s a guess based upon a lack of information by the people that should have it perhaps just a tad more accurate. If you found 3 sensors within 5 KM of the supposed location, of course you’d guess one of them would be it, and not know for sure; all you could do is run down the logic path and see if anything made sense (like balancing the location against the reported anomalies) and other such detective work…

    Anthony: “It rather tickles me that GISS doesn’t know this and I do.”

    I’d more state it as this. How can the station be properly adjusted for by GISS (or anyone for that matter) if nobody knows exactly where it is and what’s around it and when and how and why the station was moved? This is unacceptable; how can anyone even claim the data is accurate much less that it has meaning if something so basic is questionable at best? This is more an unknown than anything else.

    But of course, location stations, auditing them and taking pictures is not necessary, in fact, it’s a waste of time by morons out to disprove global warming and too stupid to understand the entire system is perfect and done totally properly on all levels. The science is settled.

  91. steven mosher
    Posted Feb 19, 2008 at 2:42 PM | Permalink

    re 89. Anthony. I’m proud to know you.

  92. eric mcfarland
    Posted Feb 19, 2008 at 6:14 PM | Permalink

    There’s also this possible location:

    http://maps.google.com/maps?hl=en&tab=wl

  93. eric mcfarland
    Posted Feb 19, 2008 at 7:54 PM | Permalink

    oops:

    http://maps.google.com/maps?f=q&hl=en&geocode=&q=33.4n,+110.9w&ie=UTF8&ll=33.401567,-110.900009&spn=0.007416,0.012167&t=h&z=16&iwloc=addr

  94. Paul Linsay
    Posted Feb 19, 2008 at 8:29 PM | Permalink

    #93, I took a tour of the place once, an open pit copper mine. They took us around in air conditioned busses and all the mining equipment had air conditioned cabs, otherwise you’d burn to death out there. It was even hotter than Tucson. It’s a great place for setting temperature records.

  95. Paul Linsay
    Posted Feb 19, 2008 at 8:35 PM | Permalink

    The link got dropped Inspiration Mine

  96. Mike Rankin
    Posted Feb 19, 2008 at 9:32 PM | Permalink

    Re: 89, Anthony Watts

    The ClimateScience writeup you noted was most interesting. The thesis by Ashley Brooks showed a different method of evaluating surface locations. Obviously, most volunteer observers will not be able to provide the scope of analysis as presented there, but it does give some ideas of the broader definition of site location and impacts on observations. The thesis also gave me more information of how reanalysis looks at such observations over time and location moves.

  97. Anthony Watts
    Posted Feb 19, 2008 at 9:32 PM | Permalink

    RE93 Eric, A note about mines, especially open pit mines, they migrate with the mineral seam.

    So it is not at all surprising to find that where there may have once been a mine office and temperature sensor nearby, there are now just mine tailings.

    Most USHCN climate stations of record have several moves in their history, even without the influence of a moving mine operation.

  98. Posted Feb 19, 2008 at 9:44 PM | Permalink

    RE96 Mike,

    All of Ashley Brooks Indiana surveyed stations have been incorporated into the surfacestations.org project with the blessing of her thesis supervisor. I agree she did a splendid job, but you are correct in pointing out that type of survey is complex and time consuming and perhaps not suitable for a volunteer network.

    On the plus side, many, but not all, of her land use and station representivity analysis techniques can be applied to the surfacestations project after the surveys are complete. There will be time after the main survey is done to revisit that.

  99. Brian
    Posted Feb 19, 2008 at 10:10 PM | Permalink

    I was looking at some stations here in MN, and I found the same problem with the charts. I went to surfacestations.org and looked at some of the sites that had been surveyed during July and Aug last summer. I compared the plots listed with the survey compared to what GISS has listed on there site. I found the same problem. Cooler past(2000 and earlier)by about 0.25 C.

    But interestingly, I also found some that were warmed considerably(2000 and earlier) by about 0.5-1.0 C. Can anyone explain that?

    All plots after 2000 looked to be unchanged.

  100. Bill Drissel
    Posted Feb 20, 2008 at 9:21 AM | Permalink

    Easily confused engineer asks:
    1. Is it true that stations have moved from grassy knolls to parking lots next to buildings?
    2. Did measured temps at these stations rise after the move?
    3. Did anyone ever decide to “adjust” data from 1900 to 1930 DOWNWARD to “compensate” for the temp rise in Q 2?
    4. If answer to Q 3 is yes, how did they get any scientist over 12 yoa to certify the data?

    Please help,
    Easily Confused in Texas

  101. Joe Black
    Posted Feb 20, 2008 at 11:19 AM | Permalink

    Bill Drissel says:
    February 20th, 2008 at 9:21 am………

    …..Please help,
    Easily Confused in Texas

    Easy. It’s “Climate “Science””. Quite like a Social “Science”.

  102. henry
    Posted Feb 20, 2008 at 11:37 AM | Permalink

    Quick question to Anthony – when the survey is done, does a copy of the current temp chart get placed in the record? That would make it easier to check to see if “adjustments” were made to the raw data.

  103. Sam Urbinto
    Posted Feb 20, 2008 at 1:05 PM | Permalink

    Bill: There’s multiple things at play here. The past adjustments down would of course cause a trend rise. (Another sub-question is the drops and rises over the last 30 years of year to year or multi-year that are equal to the trend/current anomaly). The change in the instruments/techniques of sampling the daily temperatures at each location. (In addition to the sub-question of station moves and siting biases). In addition to the question of thinking of these things in terms of a global temperature. And all the odd adjustments using strange mystical methods.

    But hey.

  104. Joe Black
    Posted Feb 20, 2008 at 1:39 PM | Permalink

    But hey.

    The phrase that must not be uttered. LOL

  105. Joe Black
    Posted Feb 20, 2008 at 1:53 PM | Permalink

    henry says:
    February 20th, 2008 at 11:37 am

    does a copy of the current temp chart get placed in the record?

    Not always, but see:

    http://gallery.surfacestations.org/main.php?g2_itemId=1587&g2_page=2

  106. henry
    Posted Feb 20, 2008 at 3:18 PM | Permalink

    I read the Miami, AZ blog enrty, and understand what took place to data AFTER it was submitted. But could we look at other stations surveyed at the start of the program to now, and see if there have been any changes to the raw data?

    Anyone know what the first station surveyed was (and how long ago)?

  107. Joe Black
    Posted Feb 20, 2008 at 3:33 PM | Permalink

    and see if there have been any changes to the raw data?

    It’s kind of unclear exactly what you mean by raw data. The raw data is contained by the daily/weekly log sheets made at each station ultimately submitted to NOAA (and its predecessors). GISS (NASA) uses the USHCN data corrected by NOAA as its “raw” data and corrects from there. Steve McI induced GISS to make corrections to its published data at some poiint last year.

    See:

    as an example

  108. steven mosher
    Posted Feb 20, 2008 at 3:54 PM | Permalink

    RE 103

    but Hey.

    but hey is the straw that sticks to your backside after

  109. eric mcfarland
    Posted Feb 21, 2008 at 1:31 PM | Permalink

    Does this link work? If so, map it … and really zoom in … no buildings?

    http://mi3.ncdc.noaa.gov/mi3qry/search.cfm?PleaseWait=OK

  110. eric mcfarland
    Posted Feb 21, 2008 at 1:37 PM | Permalink

    I was afraid it would not be direct. Look under COOP # 025512 … which should be the number for the Miami site. You can then map the station from right off of the page. Watts previously did it … but did not focus in … which shows no buildings adjacent to the site.

  111. eric mcfarland
    Posted Feb 21, 2008 at 7:22 PM | Permalink

    “However, back in the mid 90’s offices were provided with hand held GPS units and sent out
    to accurately locate the Coop stations and update the database with the correct data.” More to come.

  112. MarkR
    Posted Feb 21, 2008 at 7:33 PM | Permalink

    eric. Google Earth shows a different location for the NOAA coordinates (but hey etc,) a little way from Mr Watts site. Mr Watts has taken pictures of the equipment so I imagine his coordinates are correct. Even so, the google Earthed NOAA coordinates show an area surrounded by buildings.

    Well done for effort eric. Why not take your camera out over the weekend, and do some of your own sites. (I’m in the UK so am excused)

  113. eric mcfarland
    Posted Feb 21, 2008 at 7:46 PM | Permalink

    112: But, the buildings are not directly adjacent … by any means … as the pictures depict.

  114. Earle Williams
    Posted Feb 21, 2008 at 8:07 PM | Permalink

    Re #112,

    eric mcfarland,

    What bone exactly are you gnawing on? The coordinates of record are given in whole degrees, minutes, and seconds. Official coordinates date back to prior to 1986, when the station converted to MMTS. The best precision manageable is 1 arc-second lat or lon.

    The fact that the official coordinates mislocate the pictured MMTS unit is no big surprise. The presence of the MMTS adjacent to buildings is a suprise. What I challenge you to make sense of is what has been the position of the MMTS relative to the official coordinates of record since 1986? There are clues in the metadata.

    What is lacking of course in the metadata is datum and PDOP for the GPS location that was established in 1996. Don’t be surprised if you get a 62 meter offset if there is a NAD27 / WGS84 mixup.

    There certainly is a lot of mystery lurking in the metadata. Imprecision in the coordinates of record isn’t one of them.

    By the way, Google Earth uses a WGS84 cylindrical projection. Presumably Google Maps does too. The white trucks in the top picture show up nicely.

    Magma Mine

  115. MarkW
    Posted Feb 21, 2008 at 8:52 PM | Permalink

    eric appears to be convinced that the stations must be where NOAA data says they are. Even if they can’t be seen.
    Apparently the latest equipment upgrade included Romulan cloaking devices.

  116. ferris
    Posted Feb 21, 2008 at 8:55 PM | Permalink

    Re. #114

    Don’t forget, Selective Availability (SA) was turned off in the year 2000. SA would have been a large source of error in the 1990’s.

  117. Earle Williams
    Posted Feb 21, 2008 at 9:18 PM | Permalink

    Re #116

    ferris,

    Yep, unless differential correction was applied. Given the dearth of helpful information in the location metadata I don’t think one can expect there to be better than 200 m accuracy or precision. But by golly there’s a point on the satellite picture not next to any buildings! ;-)

    It gets interesting if you dig into the location metadata, which I’m sure our TIR is doing now. For example in the 1996-2001 time period the MMTS is listed as being 180 feet(?)/meters(?) away at a bearing of 230 degrees. Oddly enough a building is listed as being 180 feet(?)/meters(?) away at a bearing of 230 degrees. Juxtapose those two and I’m kinda detecting a MMTS adjacent to a building. Things that make you go hmmmm!

    We’ll see what our TIR comes up with.

  118. eric mcfarland
    Posted Feb 22, 2008 at 11:48 AM | Permalink

    117: It’s “the” point … chief.

  119. Earle Williams
    Posted Feb 22, 2008 at 12:11 PM | Permalink

    Re #118

    eric mcfarland,

    The point? Wuzzat? That NOAA records don’t adequately reflect sensor location and surrounding conditions? You’re finally coming around to agree with Anthony Watts then. If you’re belaboring some other point you may wish to actually voice it with some degree of specificity.

    Cheers,
    Earle

  120. eric mcfarland
    Posted Feb 22, 2008 at 4:01 PM | Permalink

    Even the exact coordinates from the survey form put you off in the bush:

    http://maps.google.com/maps?f=q&hl=en&geocode=&q=33.401983n,+110.869483w&ie=UTF8&ll=33.401983,-110.869483&spn=0.001854,0.003042&t=h&z=18

  121. eric mcfarland
    Posted Feb 22, 2008 at 4:11 PM | Permalink

    Steve is now moderating my comments … all of a sudden? Is that universal? What are your thoughts on why none of the coordinates place the station in question where the pictures places it — i.e., next to the buildings?

  122. Sam Urbinto
    Posted Feb 22, 2008 at 4:22 PM | Permalink

    No, here’s what’s happening is that whatever datum you’re using is inherently going to have some error, and there are many station moves that don’t get listed, or were listed incorrectly. The “official” location may have had the GPS set to the wrong datum. There are often conversion errors if you convert back and forth from DD, DMS, UTM or MGRS. A 1:50000 map can be off by 50 meters. Etc.

    If the station is supposed to be at X/Y/30 -Z/A/60. If that’s rounded, there’s a lot of land between 29-31 and 59-61 If it’s wrong, who knows.

    Even if exactly there at 30 60, the margin of error on something like Google Earth may not point directly at it; giving you terrain or buildings/no buildings that are not actually at the site. GE has known to be incorrect. Some areas have less resolution or are older images.

    Or in other words, what GE shows is not proof you are looking at where the sensor is.

    eric, you might want to tone it down some and think a little more about these things, or ask. People might think you’re behaving like a troll.

  123. Severian
    Posted Feb 22, 2008 at 4:32 PM | Permalink

    Snippage material to follow:

    C’mon Sam, you know the ugly truth. The “What are your thoughts on why none of the coordinates place the station in question where the pictures places it — i.e., next to the buildings?” question is answered by the fact that the evil “deniers” are falsifying the photos! It’s obvious that the surveyors are taking fake sensors with them to put them next to buildings just so they can take a photo and attempt to support ExxonMobil by lying about all of this. That’s why I can’t survey any stations personally, my car is too small to hold a Stevenson screen or a MMTS sensor!

    Good grief.

  124. BarryW
    Posted Feb 22, 2008 at 4:41 PM | Permalink

    Re 121

    Your also dealing with mapping distortions from the sat/aerial photographs to the earth, and who says GE is totally correct? My house address was mapped a block away from where it really is.

  125. Sam Urbinto
    Posted Feb 22, 2008 at 5:24 PM | Permalink

    eric: If you have too many links (or certain other elements) the spamcatcher grabs it and it has to be taken out of the queue. John A explained it, you can probably find his post here someplace.

    Severian: lol

    BarryW: Ah, that’s the point, yes; it could be incorrect data, or it could just be the margin of error. If you’re land navigating with a 1:50000 map, 50 meters is a lot of room; satellites, well…. Mine house is pretty much right on, (at least down to a certain margin) but it depends on a lot of factors what you get. On the other hand, I’ve done map lookups on various online services where the point shown was not even on the same street and was way far off in distance…. But of course, after the restaurant must have been in that location, an Internet site map said it was… :)

  126. BarryW
    Posted Feb 22, 2008 at 6:28 PM | Permalink

    Sam U

    You’re not only not working with a flat surface, you’re working with an “oblate spheroid”. Try computing a back azimuth on a bearing and see how close you get to where you started. Drove my office mate nuts.

  127. Severian
    Posted Feb 22, 2008 at 6:41 PM | Permalink

    This whole subject is giving me nightmarish flashbacks to my first job out of college, working on the Pershing II missile. Shades of gyro-compassing, geodesy, UTM vs WGS variants, oblate spheroids, etc. Ugh.

    Suffice to say that there are a lot of different errors that can creep in, even things that are not errors but are artifacts of different coordinate systems and different earth shapes, ala UTM vs. WGS. vs. lat/long and such without attributing any of this to deliberate malfeasance.

  128. Earle Williams
    Posted Feb 22, 2008 at 6:58 PM | Permalink

    I was snipped too, a follow-on to my #117 that offered nothing but an explanation of the TIR abbreviation. Our host is making clear there are limits to the amount of incivility he will tolerate.

    eric mcfarland, I am still at a loss as to what your issue is. If you haven’t dug through the metadata like I suggested, then I’ll walk you through it now. The NOAA records show that the official location was reported in whole degrees, minutes, and seconds. Got to the MMS link that you looked up Miami with originally and click on the Location tab. There you will see that the official location is given as N 33 24′ 16″, W 110 52′ 12″. There are no decimal bits beyond the 16″ north or 12″ west. So, unsurprisingly, there is going to be some mislocation because the coordinates are not any better resolved than 1 second latitude or longitude.

    The good folks at NOAA were aware of this limitation and used these coordinates as a reference point. They then provided details as to where the MMTS is located with respect to this reference point. In the Location tab for the MMS Miami, AZ, site scroll down and click on the entry that begins in 1996-2-21. This gives you the location metadata on record for the time period covering 1996 to 2001. Now scroll down and click on the Obstructions/Exposures link. This is the one with the following text in blue:
    MMTS 230/180 BLDG 130-150/140/14 180/130/10 230/180/12 270/190/22 HILL 280-310/90-90/26

    This information tells you where various things are in relation to the reference point. Note the first two words: MMTS 230/180. This means that with respect to the official coordinates the MMTS is located 180 feet to the southwest (a bearing of 230 degrees). Now read a bit further, to the buldings part. Note the second bulding set of numbers, the 230/180/12. They indicate that a buliding 10′ high is located 180 feet to the southwest.

    So, I’ll spell it out here in case you didn’t follow me. The MMTS is located next to a building. NOAA has documented that the MMTS is located next to a building. If you plug the offical coordinates into Google Earth (or Mapquest, etc) your point will not be next to a building. The metadata explains the discrepancy. The photos taken by the observer provide visual confirmation of the instrument location.

    Sorry about the lack of links, but as you’re aware the MMS site doesn’t provide click-through linking.

  129. eric mcfarland
    Posted Feb 22, 2008 at 7:12 PM | Permalink

    I plan to keep looking into this issue until I am satisfied. I hope that does not hurt any feelings. In any case, and in all events, Watts and Friends could have and should have disclosed this data problem (if that is … all it is) from the get go — particularly where they are parading a particular site around as a bad apple … and no coordinates (official or otherwise) appear to place it properly (including their own … for whatever reason … GE included). I think that has to be taken as a fair comment, at least. And, BTW, I never as much as hinted at malfeasance. I think that was a paranoid invention in the mind of some around here who seem to revert to using the word “troll” when they can’t think their way out of a problem. Pretty weak approach … in my estimation.

  130. Alan S. Blue
    Posted Feb 22, 2008 at 7:24 PM | Permalink

    The main table for the USHCN simply doesn’t have enough decimal places for latitude and longitude. It may seem like it, but it is a conversion from degrees-minutes-seconds or something.

    If you pop the coordinates for “Seattle, Urban” (47.65°N, 122.3°W) into GE, you get a spot on the University of Washington’s practice football field. I would hope there’s no MMTS placed where GE says it is. The master list has:

    457458 SEATTLE-URBAN SITE 1.5 S WA 47.65 -122.3 72793007 19

    Search for “Seattle” in the NOAA search engine, and there’s 10 sites. There’s only one that’s at “47.65N”, and that one’s at the Seattle Yacht club. The longditude is -122.3166 though, which is a slight shift. Let’s see. (47.65°N, 122.316667°W) That’s… in the middle of a bay.

    But the name says “Seattle Yacht Club”. That’s here: (47.64545°N, 122.3086°W) – according to maps.google.com anyway. The amusing thing is, the next-door neighbor to the yacht club is NOAA Fisheries. They have some good sites for a station. The yacht club is wall-to-wall asphalt. There’s a freeway abutting the SYC property, which will at least ensure that the air is well-mixed.

  131. eric mcfarland
    Posted Feb 22, 2008 at 7:26 PM | Permalink

    128:
    “They indicate that a buliding 10′ high is located 180 feet to the southwest.” To the southwest of the coordinates … or the actual instrument in question … or no one can say? And when you say it is next to a building, does your reading indicate that they are directly adjacent … as depicted in the picture?
    Thanks!

  132. eric mcfarland
    Posted Feb 22, 2008 at 8:04 PM | Permalink

    Also … lest we forget … Watts said that it was specifically NCDC’s data that is/was wrong. There was not “this and that” about degrees, GE, and what-not.

  133. K
    Posted Feb 22, 2008 at 8:10 PM | Permalink

    Like others I am having a hard time figuring out what point(s) eric is trying to make about station locations. Or even what point can be possibly be made.

    The GPS system was designed to not be precise for civilian use (SA #116). I retired in 2000 and didn’t follow the matter afterwards. I didn’t even know that SA was turned off in 2000. My daughters car has GPS and I recently wondered how the it could be so accurate, sometimes within 50 feet (but 50 yards is the better bet.) I didn’t wonder enough to investigate.

    We have three references. Government provided public location(AGENCY). The SurfaceStation.org location (SURVE). Online access to Google Earth and/or maps(SATS).

    We don’t know which AGENCY readings were taken before 2000 or how. SURVE readings were taken after 2000. And what is the accuracy of SATS; we do know it often disagrees with GPS devices at the site. I personally know it placed my hotel in Sinai over a mile out in the Red Sea last year. (I’ll try to compare my daughter’s GPS readout at my house location to SATS this weekend. Might zoom down on Greenwich too; that one better be right).

    Plus there are, apparently, secret instructions that allow AGENCY to actually visit and operate the stations.

    And readers here understand the problem of precision; i.e. knowing a location to the nearest degree and minute often isn’t enough to find it. In fact degree, minute, and second isn’t a gimme in some surrounds.

    The obvious conclusion? Surfacestation.org must be mistaken when coordinates disagree.

    Know what? Exact coordinates don’t really matter. What matters is whether the real station was surveyed. i.e. did that station produce the data we attribute to it?

    PS. I wrote the above about two hours ago. Since then some excellent comments – #127, 128, 130 have been made. And reading eric at #129, 131 and 132 has convinced me he does not intend to understand.

    BarryW #126: I remember first seeing the term ‘oblate sphereoid’ about 1970. We had a lot of fun with it too; I think part of the Indian Ocean was said to be eight miles too high, or too low, or anyway it wasn’t quite as it should be.

  134. M. Jeff
    Posted Feb 22, 2008 at 8:20 PM | Permalink

    #133 your links should be to http://surfacestations.org/

  135. BarryW
    Posted Feb 22, 2008 at 8:34 PM | Permalink

    Re K

    Hey, I finally got to use “oblate spheroid” and back azimuth in a sentence outside of work!

  136. K
    Posted Feb 22, 2008 at 8:34 PM | Permalink

    #134 Actually I didn’t intend that it be a link. I simply typed. Some software, somewhere, somehow detected the pattern and converted it to a link.

    Is that a feature of blogs? I don’t know. Or of some blogs? Or of my browser – doubtful. Or my ISP – even more doubtful.

    Sorry for any inconvenience. I’ll try to remember and avoid.

  137. Julie KS
    Posted Feb 22, 2008 at 8:46 PM | Permalink

    I don’t think it’s really a problem of lat-lon precision. I think it’s in many cases a question of the intent of the technician who GPSed the instruments, and of course human error enters into this too.

    Eric, you are over-thinking this. In my (admittedly limited) experience, I have seen given coordinates that marked the temp sensor exactly (within the margin of error of my GPS unit). I have seen given coordinates that marked the street location nearest the temp sensor (Blanco, TX). I have seen given coordinates that marked the front door of the Ag Extension office, which had a weather station out back by the staff parking lot (Fort Stockton, TX). I have seen given coordinates that marked the precipitation instruments, which turned out to be quite a distance from the temp sensor, which was right up next to the house it was wired to (Dubois, WY).

    If you look at the metadata, there is only one set of coordinates, but the instruments are not always grouped together. It’s difficult to know which instrument the coordinates are referring to without visiting the site.

    I think in many cases the technicians doing the GPS marking were only trying to establish a general location in the climate zone, and help lead other technicians to the vicinity of the stations. Their intent was NOT to enable auditors to measure the distance between the Nimbus unit and the barbecue grill.

    And speaking of GE, it’s such a great tool for this, but as Anthony said, resolution for the images over much of the US is so poor that the instruments can’t be seen. Only after the instrument’s location has been verified on the ground can GE be used to analyze the site, in many cases.

    Also, Eric, the obstruction metadata refers to the distance and azimuth from the actual instruments.

  138. eric mcfarland
    Posted Feb 22, 2008 at 8:59 PM | Permalink

    137: So, the “building” should be about 180 ft away? That is, according to the metadata as posted by Earle?

  139. Julie KS
    Posted Feb 22, 2008 at 10:46 PM | Permalink

    The example Earle gave you is for 1996-2001. He walked you through how to look at the obstruction data. Go see. You’ve been instructed; it’s time for you to do your own work, grasshopper, or you will never understand.

  140. eric mcfarland
    Posted Feb 22, 2008 at 11:37 PM | Permalink

    139: sounds a bit like a dodge to me …

  141. eric mcfarland
    Posted Feb 23, 2008 at 12:15 AM | Permalink

    Attention: All of the NCDC’s systems will be down due to a scheduled power outage from
    Sat, Feb 23 at 7am EST until approximately Sun, Feb 24 at 8am EST.
    ——————————————————————————–

    [Skip to end of header ]

    Phenomena Other Considerations Map Remarks Files Related

    Identity Updates Location Other Party Data Products Data Programs Equipment

    Station Name: MIAMI
    Country: UNITED STATES
    State/Prov: ARIZONA
    County: GILA
    Latitude: 33.4044 (33°24’15.84″N)
    Longitude: -110.87 (110°52’12.0″W)
    Elevation: 3560.00 FEET ( GROUND )
    POR: 1948-07-01 => Present
    Climate Div: 04 – EAST CENTRAL
    IDS: COOP NUMBER = 025512 NCDC STATION ID NUMBER = 10100054 NWSLI = MIAA3
    ( Not Locked )
    No Info Src Permissons

    Tab Remarks: View (0)

    2007-06-26 – 9999-12-31

    Station Location (click on data for more detail)
    Lat / Lon
    [ 33.4044 (33°24'15.84"N) /
    -110.87 (110°52'12.0"W) at primary location ]

    Geographical Information
    Geo Region: AMERICA, NORTH
    Country: [ UNITED STATES ]
    State: ARIZONA
    County: GILA
    NWS Region: WESTERN
    WFO:
    Climate Division: EAST CENTRAL
    Time Zone: [ MOUNTAIN ]

    Elevation
    GROUND: [ 3560.00 FEET ]

    Relocation: None
    Obstructions/
    Exposure: [ BLDG 105/18/9 ]
    [ TOWER 255/160/36 ]
    [ HILL 280-310/90-90/26-26 ]
    [ BLDG 230/180/12 ]
    [ BLDG 180/130/10 ]
    Other Regions:

    Topography: [ SMALL LEVEL AREA IN MIDST OF MINING OPERATION.ROUGH BROKEN MESAS SURROUNDING. ]
    Location Description: None

    2007-06-26 – 9999-12-31

    http://www.ncdc.noaa.gov/common/ncdcfoot.html

    Downloaded Saturday, 23-Feb-2008 01:13:32 EST
    Last Updated Tuesday, 04-Dec-2007 11:25:52 EST
    Please see the NCDC Contact Page if you have questions or comments.

    ——————————————————————————–
    Dynamically generated by http://mi3.ncdc.noaa.gov – Created by [NCDC] – Version 5.3.4 – September 20, 2007

  142. eric mcfarland
    Posted Feb 23, 2008 at 12:22 AM | Permalink

    So, the closest building should be 9 feet tall, out 105ft, at 18 degs (N-Slight East). Do I pass?

  143. Julie KS
    Posted Feb 23, 2008 at 8:18 AM | Permalink

    Obstruction Type: BLDG
    AZM: 105
    Distance (units): 18
    Elevation (units): 9

  144. Julie KS
    Posted Feb 23, 2008 at 8:30 AM | Permalink

    Click on [ BLDG 105/18/9 ] to get what I posted above.

  145. M. Jeff
    Posted Feb 23, 2008 at 9:14 AM | Permalink

    I also need a personal tutor, where do I sign up?

  146. Earle Williams
    Posted Feb 23, 2008 at 10:38 AM | Permalink

    Re #141

    eric mcfarland,

    Congratulations! Passing grade on finding the data. Failing grade on having any coherent argument. Passing grade on dogged insistence to find SOMETHING to complain about in Anthony Watts’ post. Failing grade on acknowledging the plain and simple fact that is the entire point of the post: the location of the MMTS at the Miami, AZ, station is not suitable for measuring climate.

    According to the metadata which you have found, there have been several adjustments to the location. So tell me, has the sensor itself moved, or just the official record of its location? How on earth would one know based on the metadata?

    Finally,

    Also … lest we forget … Watts said that it was specifically NCDC’s data that is/was wrong. There was not “this and that” about degrees, GE, and what-not.

    No eric, he didn’t. He said that the station has moved and because of the move the recorded temperature is significantly higher. YOU are the one that brought up this and that about degrees, GE and what-not.

  147. Earle Williams
    Posted Feb 23, 2008 at 11:15 AM | Permalink

    Ric Locke makes an interesting point in the 19 versions thread.

    Warning: do NOT, repeat NOT, trust Google Earth or Microsoft Live Earth at the detail level. I know some of the people who do data capture for them, and what you see is for all practical purposes a Photoshop exercise — texture mapping image data on questionable surface models.

    I don’t know to what extent we can rely on the image to discern whether or not the TRUE position of the MMTS should be as shown by Google, in the bushes as shown by the observer’s GPS coordinates, or at the official coordinates provided by NOAA. Welcome to uncertainty.

  148. Alan S. Blue
    Posted Feb 23, 2008 at 1:22 PM | Permalink

    A Google Map with pins for various bits. As Earle says: none of these things line up.

    Hand Google Maps the tabulated official coordinates (33.4044 -110.87) and you’d get the green pin.

    Hand Google Maps the handheld GPS coordinates reported by an observer (33.401983, -110.869483) and you get the purple pin.

    Start from the map and weasel down to the location of the MMTS from the pictures submitted by the observer, and you get the blue pin. The “Google Maps Coordinates” of the MMTS.

    Fiddling with the TOWER/HILL/BLDG instructions doesn’t seem to get me to any of the pins. Focusing on the TOWER line, for instance, puts me in a pile of rock. Even if the MMTS hasn’t moved, it seems like the current tower is not the one referenced. And, this being a mining area, perhaps even the HILL reference moved. I dunno.

  149. eric mcfarland
    Posted Feb 23, 2008 at 1:28 PM | Permalink

    See 79:

    “What you are seeing is the fact that NOAA does not record the lat/lon in their NCDC MMS database with great accuracy. Many entries are often simply wrong, and our volunteer surveyors have to become CSI’s to figure out where the station actually is located.

    In some cases, we’ll find stations hundreds of yards or miles away from the locations described by the database.”

    You can’t fault me for spotting an obvious wrinkle … and one that does not offer an obvious explanation to a reader like me. I still have questions about the “find” on this site … especially since the ncdc database provides such great detail … which leaves me to believe that great effort is put into recording location information with a certain degree of accuracy.

    Also, the simple fact that the site has been around for a long time makes it good for hist. temp. work … even if it does sit on the rim of hell itself.

  150. eric mcfarland
    Posted Feb 23, 2008 at 1:45 PM | Permalink

    “Relocation: None” ???

  151. Alan S. Blue
    Posted Feb 23, 2008 at 2:25 PM | Permalink

    Eric. Please look at the map I posted and identify all the potential thirty-six foot towers. The one tower I see doesn’t seem like it is thirty-six feet, but taller. Google maps will let you draw a line – and measure the line. Yes, there are inaccuracies, but it will give you a good sense of “Is this spot 160 feet from that tower?” But the observer who actually did visit the site was able to talk to the curator – the person actually entering the data daily. Is the curator using the wrong instrument? Did the observer go to the wrong spot? Is the reasonably detailed information under “Exposure” flat wrong? Are the coordinates insufficient? The observer took pictures E,S, & W of the MMTS he observed as well. I don’t see any towers in those pictures, except perhaps the ‘large tower’. Which isn’t south-of-west (255) of this MMTS (more like bearing-300), doesn’t seem like it is thirty-six feet, and is more like 466 feet away.

    Are you arguing “Anthony Watts is inaccurate when he says the lat/long is inaccurate”? Or “Surface Stations.org is using incompetent observers?”

    I, for one, am missing what your point is. If they’re talking with the curator, how can they possibly be getting the wrong device? I can see messing up the SEATTLE-URBAN one. There’s probably 15 MMTS-like devices within the possible lat/long error. (There’s an entire 70,000 person university campus, a NOAA laboratory, a NOAA ship-base…) But for MIAMI? There aren’t going to be that many curators around a site like that. If he’s running multiple devices, he should know about it.

    If they aren’t getting the wrong device, then discussing how detailed the NCDC database is is just extending the amount of tabulated data that is either erroneous, insufficiently accurate, outdated, or otherwise not pointing to the actual device providing data.

  152. Joe Black
    Posted Feb 23, 2008 at 2:51 PM | Permalink

    Welcome to uncertainty.

    Can’t we get the politicians to pass a law to eliminate uncertainty?

  153. EW
    Posted Feb 23, 2008 at 3:36 PM | Permalink

    Concerning the ROW, I would like to see someone trying to locate a Czech station Milesovka, 11464 according to its description in GHCNv2 inventory ;-)

    In the description, there is elevation -999 (they don’t know) and the topography is described as “Mountainous valley or at least not on the top of a mountain”. The station, however is located on the very TOP of a mountain of the same name (838 m) and was built there in the year 1906 as an observatory for lightning studies, so that a believer could spend a lot of time running through various neighboring valleys, seeking in vain for the station. The station has never been moved.
    The GPS on a Czech web mapper says 50° 33′ 18.11″N, 13° 55′ 53.56″E, the GHCN says latitude: 50.55 longitude: 13.93

  154. EW
    Posted Feb 23, 2008 at 3:38 PM | Permalink
  155. eric mcfarland
    Posted Feb 23, 2008 at 6:01 PM | Permalink

    151:
    The curator notes don’t actually identify a witness by name … or what was said. The notes do indicate, however, that the device has been in its present location for the past 20 years.

  156. eric mcfarland
    Posted Feb 23, 2008 at 8:10 PM | Permalink

    “In both cases, the abnormal spike coinciding with a station move near the present time remains in the record …”
    The only problem here is … that site has not moved … or at least not for 20 years … and this aint no move from cool-rural to hot-urban … if any. So, the whole punchline in this story is screwy. Snake oil?

  157. Posted Feb 23, 2008 at 8:29 PM | Permalink

    RE149 Eric writes:

    … especially since the NCDC database provides such great detail … which leaves me to believe that great effort is put into recording location information with a certain degree of accuracy.

    Hi Eric, You might not realize this. This NCDC MMS is a database of field gathered data. The way NCDC was setup from its inception is to use government forms, paper forms, typed or handwritten paper forms. Do you remember Corporal Radar O’Reilly, the company clerk in MASH, always needing the colonel to sign some form? Well our government is still like that. Bureaucracy subsists on forms. Nothing gets logged, approved, collated, or transcribed unless its on the correct form. Don’t have the form? Tough noogies. Got the form but it has handwritten scribbles on it and coffee stains? Ok. We’ll figure it out, so long as it is on the correct form.

    They still do that today. The “great detail” you speak of comes from something called a B44 form, which is filled out in handwriting and/or typewriter by the local NWS COOP manager when the site is setup and inspected.

    Here is an example of one for the Fire Station at Marysville, CA

    http://gallery.surfacestations.org/main.php?g2_view=core.DownloadItem&g2_itemId=31477

    That form gets mailed by snail mail to NCDC in Asheville. There it is logged then sent for transcription by a human being into the database. Sometimes the transcription is perfect, sometimes it isn’t. In the case of Dover, DE we believe a transcription error is the reason the lat/lon for the station shows up in the middle of a field on Google Earth but clashes with the description of the station location at NCDC MMS suggesting it is at the highway department.

    You can’t conceive of the idea that a station could be lost? Huh. Well look at the story in Los Angeles times the other day where the Los Angeles school district can’t find millions of dollars worth of computers. Of how about the annual GAO reports that come out where our government can’t find millions of dollars worth of inventory? Vehicles, airplanes, parts, you name it, lost in the black hole. But there’s a form somewhere that says it exists.

    Ok back to our story. I’m normally not allowed to look at these B44 forms (NCDC does not put them online, nor will they provide them on request, citing “privacy concerns”) but I do have a set for California given to me by the former state climatologist that are about 15-20 years old. I’ve posted the ones online that are for public places like the fire station, in their entirety, but not for any private observers who run stations at residences.

    Note also that every monthly data set that is submitted by an observer to NCDC is also on a paper form (the B91 form) and transcribed at NCDC, just like its been done for over 100 years. Here is one:

    http://gallery.surfacestations.org/main.php?g2_itemId=28535

    here is another:

    http://gallery.surfacestations.org/main.php?g2_itemId=28542

    Note they have handwriting, typed, and the missing data on the B91 forms sent to NCDC each month. Remember, observers are human, they have days off, weekends, sick days etc. so what you see of missing data in the two samples above is not uncommon.

    That missing data for the station temperature gets “filled in” by something NCDC invented called FILNET, but that is another story.

    The point of all this is, the system of COOP station reporting of location and of data is largely dependent on human recording and transcriptions. If you want to believe that such a system would have a low error rate, and that we just keep finding the wrong stations from the “right” data provided by NCDC, you are certainly welcome to think that.

    I used to think that going in. How could something so heavily depended on by scientific research be that badly messed up? I soon learned that the reality is different from the expectations.

    In your comment 151 you point out that

    The curator notes don’t actually identify a witness by name …

    Yes that is correct, and that is done for a reason. You see when this project was first started, there was a big hullabaloo about how the army of volunteers were going to “invade the privacy” of the private COOP observers.

    See the POLICY that volunteers for the surfacestations project must agree to before they signup:

    http://www.surfacestations.org/rules.htm

    Note how heavy it is on protecting observer privacy, even though our friends at the NWS publish names and photos of the observers standing in front of their stations.

    http://www.wrh.noaa.gov/mtr/cpm/2001awards.php

    You see, this project isn’t well liked by some in official positions, and roadblocks have been thrown at it. See this:

    http://wattsupwiththat.wordpress.com/2007/06/30/noaancdc-throws-a-roadblock-my-way/

    and this

    http://wattsupwiththat.wordpress.com/2007/06/30/update-noaancdc-privacy-concerns-questioned-by-alert-blog-reader/

    and finally this:

    http://wattsupwiththat.wordpress.com/2007/07/07/noaa-and-ncdc-restore-data-access/

    So I have to operate this project (on my own time with zero budget) on the assumption that we must maintain that privacy even though the NWS doesn’t.

    Double standard there, but it is what it is. It was the only way the project could move forward.

  158. Posted Feb 23, 2008 at 8:32 PM | Permalink

    Oh forgot to mention:

    FUBAR

  159. M. Jeff
    Posted Feb 23, 2008 at 9:07 PM | Permalink

    Hopefully temperature records are of higher quality than Social Security records. Excerpts from http://www.wsmv.com/news/15357541/detail.html

    … According to a government audit, Social Security had to resurrect more than 23,000 people in a period of less than two years. …

    … The audit said the lack of documentation in the Social Security computer makes it impossible for the government’s auditors to determine if the people are dead or alive. …

  160. BarryW
    Posted Feb 23, 2008 at 9:36 PM | Permalink

    Re 157

    Not unique. We had similar problems working with the FAA analyzing flight data, and we were their federally funded research and development center (FFRDC)! It’s taken years to get them (and the airlines) to cough up data. And from their point of view there are thousands of lawyers circling waiting for a chance to grab any and all data to use against them if there’s an accident. Wouldn’t be surprised if NWS had similar paranoia.

  161. Peter D. Tillman
    Posted Feb 23, 2008 at 10:56 PM | Permalink

    I should chime in with a few observations on the Miami (AZ) station, since I used to work at the Miami operation — the Geology office is 2 buildings east of the MMTS location. These buildings were put up, ims, in the 1930’s. The facility just S of the MMTS is the electrolytic refinery, and one didn’t want to park his car nearby, due to the acid mists from same. Which is why the steel buildings are so rusty. Perhaps not the best choice for siting a precision instrument. The black gravel in the parking lot is (almost certainly) smelter slag from the International Smelter, which is the next large cluster of buildings NE of the Miami operation. Good for soaking up the Arizona sunshine. The concrete structure next to the MMTS is the main vault for the operation’s records, in case the building burns down. It is massive poured concrete, a fine heat reservoir.

    All in all, not the best choice for a weather station.

    On the positive side, the best Mexican restaurant in Arizona (and quite possibly the world) is five minutes away.

    Cheers — Pete Tillman

  162. eric mcfarland
    Posted Feb 23, 2008 at 11:56 PM | Permalink

    No FUBAR details or warnings in the surface station survey detailed instructions for finding sites: http://gallery.surfacestations.org/main.php?g2_view=core.DownloadItem&g2_itemId=4843

    That’s odd …

  163. Anthony Watts
    Posted Feb 24, 2008 at 1:00 AM | Permalink

    RE162, Well I tried.

    Congratulations, Eric you are the new TCO, or the old one in a new wrapper.

    Folks maybe you can talk sense into him, or just ignore him, either way I’m done.

  164. Severian
    Posted Feb 24, 2008 at 9:57 AM | Permalink

    TCO? Not sure I understand that acronym. Must have missed it previously.

    But, you’re right, Eric has all the characteristics of Hoffer’s “True Believer.” I’m not sure what, exactly, his entire point is, as he’s not been able, or been unwilling, to elucidate it clearly. It would appear that he’s desperately looking for anything he can find that’s not perfect in your data or arguments so he can say “aha!” and use that as an excuse to ignore the demonstrated problems with the surface station data the AGW believers rely on so heavily.

    So, Eric, you’ve bounced off the walls for days on end now, are you ever going to come up with a conclusion or clear statement of exactly what you’re going on about?

  165. BarryW
    Posted Feb 24, 2008 at 10:47 AM | Permalink

    Re 164

    TCO was the blog name of one of the denizens here who kept getting more and more “trollish” as time went on. I’m amazed at the “Troll-erance” of the folk here with Eric. Remember-“don’t feed the troll”

  166. EW
    Posted Feb 24, 2008 at 2:58 PM | Permalink

    He’ll try to use your comments to show that the volunteers don’t know what they’re doing, can’t find stations, putting them in the wrong spot, etc.

    No wonder, if the USHCN oe GHCN database is like it seems to be. I don’t know what percentage of the USHCN stations isn’t where it should be or the description doesn’t fit, but there are 13 Czech stations, in the heart of Europe, all official and active (I mean that they are professionally kept by state or research institutions) and their GHCN database descriptions aren’t in many cases correct (elevation, lat/long, location description, data truncated in 90’s…) and God only knows how (and IF) any station moves were entered. There’s that puzzle of Smetschna/Praha-Ruzyn and similar ones…

    Therefore I’m rather skeptic about the data from the rest of the world, when even the Central Europe entries are plagued with errors that might be easily resolved by asking the responsible institutions for new data or database corrections.

  167. eric mcfarland
    Posted Feb 24, 2008 at 4:55 PM | Permalink

    True believer … you apostles of Watts cannot even see that the the entire premise of the original post is wrong. The stupid station has not even been moved … or not at least for 20 years. So, how does this “move” show up “in recent time” in the data? That’s just a starting point. Also, if the ncdd data were so FUBAR, Watts would sensibly give general warnings or cautions in the instructions on “how to find” the stations. However, he does not … does not even as much as hint that the ncdc metadata might be amiss. Instead, he instructs the survey participant to use the data … and if you can’t see the implications of that … your not thinking. There’s something screwy going on with this “survey” business.

  168. Severian
    Posted Feb 24, 2008 at 6:08 PM | Permalink

    Ah, I see Eric…if the stations aren’t where they say they are, that’s Anthony Watts’ fault. Glad you cleared that up.

  169. Craig Loehle
    Posted Feb 24, 2008 at 6:29 PM | Permalink

    About 10 yrs ago I visited Harbin China, pop 2million or so. When we drove around it at night it was DARK, as in only the rare storefront had dim lights on, no streetlights, a little light leaking out of windows. It would show up on nightlights I bet as only semirural (some houses).

  170. Craig Loehle
    Posted Feb 24, 2008 at 6:50 PM | Permalink

    It is useful to remember that the weather stations were established to provide local weather info, and a mile move is still local for that purpose. If the move was of no consequence for the purpose back in 1947, why was anyone going to be pedantic and document everything? It is only now that we wish they had, like people may wish, after they became famous, that they had kept a diary when they were young. On the other hand, it seems crazy to ignore the records that WERE kept and run an algorithm that misses the details to “fix” it.

  171. Craig Loehle
    Posted Feb 24, 2008 at 8:19 PM | Permalink

    For Eric: The point of Anthony’s project is NOT that the stations are not where the records say they are. The point is that they are not where the standards say they should be, which is NOT next to buildings, on roofs, next to incinerators, and in parking lots. If you read some of the stories of searching for the sites, the accounts clearly illustrate how the sites were found and that in fact the site is the correct one.

  172. eric mcfarland
    Posted Feb 25, 2008 at 12:54 AM | Permalink

    “So I though I’d take a look at the raw GISS temperature plot for Miami, AZ to see if the move would show a spike, it did:”

    What move?

    When?

    Pin it to the alleged “spike”…

  173. eric mcfarland
    Posted Feb 25, 2008 at 1:57 AM | Permalink

    Also, I do not accept on faith the claim that ncdc’s data is amiss. Mr. Watts is essentially doing hit pieces on the temp. stations, and now he also wants to claim that the metadata is bad … well, the burden is his to show that. Simply saying “trust me” is not sufficient. Looks like he wants to have his cake and eat it too. In short, the data is good when it suits him … and bad when it does not.

  174. MarkW
    Posted Feb 25, 2008 at 6:59 AM | Permalink

    K,

    eric is hoping that if pulls at enough threads, he will eventually unravel the sweater.
    The whole issue of GPS locations is just another thread to be tugged at. Eventually he will tire of it’s irrelevancy.

  175. MarkW
    Posted Feb 25, 2008 at 7:12 AM | Permalink

    Severian,

    You don’t understand. To eric, the fact that the record doesn’t mention a move proves that the station has not been moved. The fact that you can’t find a station at the location mentioned in the records is because of the romulan cloaking technology that was added during the last equipment upgrade. (NOTE: The records don’t include any mention of this cloaking technology in order to protect the privacy of the station attendant.)

  176. Steven Mosher
    Posted Feb 25, 2008 at 10:20 AM | Permalink

    I think everybody is misunderstanding eric. He has some concerns about the accuracy
    of the survey, so lets address the concerns in a systematic open fashion and leave emotion
    on the sidelines. Eric has two concerns. I will detail them and if he agrees that I have
    expressed them adequately we will address them in a systematic fashion.

    I will take 167 as eric’s statement of Concern:

    “True believer … you apostles of Watts cannot even see that the the entire premise of the original post is wrong. The stupid station has not even been moved … or not at least for 20 years. So, how does this “move” show up “in recent time” in the data? That’s just a starting point. Also, if the ncdd data were so FUBAR, Watts would sensibly give general warnings or cautions in the instructions on “how to find” the stations. However, he does not … does not even as much as hint that the ncdc metadata might be amiss. Instead, he instructs the survey participant to use the data … and if you can’t see the implications of that … your not thinking. There’s something screwy going on with this “survey” business.”

    There are TWO issues here.

    1. Issue number 1 is a factual issue about the station moves. This is a factual concern.
    This statement is open to confirmation or disconfirmation. We will deal with the factual question
    Later. As eric concludes :”The stupid station has not even been moved ”

    2 Issue number 2: The surface stations instructions. “if the ncdd data were so FUBAR, Watts would sensibly give general warnings or cautions in the instructions on “how to find” the stations. However, he does not … does not even as much as hint that the ncdc metadata might be amiss. Instead, he instructs the survey participant to use the data … and if you can’t see the implications of that … your not thinking.”

    Lets start with Issue #2 because it is the easiest to resolve. Instructions are written to
    capture the vast majority of situations a reader will face. These instructions were written early on in the process, when a few stations had been surveyed. Most then, and most now, are found
    near to where they are expected to be. Essentially, if you expect 90% of the sites to be
    where the NOAA data say they are, your best instructional text would say: “go where Noaa
    tells you it is.”

    As the effort has progressed and the surveys have become more numerous oddities
    pop up, a more complcated list of instructions is probably required. So, lets work through
    a fair set of instructions to survey volunteers, without attributing malice or stupidity
    or motive.

    Eric, would you approve of this as an adition

    1. On occassion you may not find the station located at the location indicated by the Metadata.
    In this case you should do the following to locate the station and confirm it’s Provenance.

    A.
    B.
    C.
    D.

    If you can help with A-Z that would be a way forward and the instructions and surveys would
    be improved.

    So eric, if we can agree on a set of instructions for surveyors, then I’ll bet than Anthony
    will ammend the instructions. Especially if the instructions are helpful. But what you suggest,
    A mere warning that the NOAA data may be wrong, really isnt an instruction. So, lets work together
    and craft an ammendment to the surface stations instructions. What should an observer do if the station is not located at the position indicated by NOAA. Since over 50% of the stations remain
    unsurveyed your skills may actually help us locate these MIA stations.

    When we settle this, which should be rather simple. We will address the factual issues.
    So, lets get the instructional issue put to bed. Propose your changes to the survey instructions
    and lets see if we can all agree on that. Otherwise we have one fight about facts and another
    fight about motives ( would have, should have, could have). Prpose some improved instructions
    that you think will help a volunteer surveyor find a site that is not where the metadata says it should be.

  177. Alan S. Blue
    Posted Feb 25, 2008 at 11:04 AM | Permalink

    Steven, he hasn’t read the actual instructions. Which mention several routes to finding the station. The recommended is finding the curator – the person actually sending in the data – online first to get permission. The instructions are peppered with “may” and “try” and present several different techniques for finding stations. This implies to non-troll-level intelligences that finding a station is not as straightforward as slavishly following the metadata – if they actually read the instructions.

  178. M. Jeff
    Posted Feb 25, 2008 at 11:47 AM | Permalink

    #176

    “I think everybody is misunderstanding eric.” stereotypes everybody. However, you may be correct, even Eric may not understand Eric.

  179. Steven Mosher
    Posted Feb 25, 2008 at 12:27 PM | Permalink

    re 177. Alan. Lets not write about whether or not eric is reading impaired. Eric has suggested
    that the instructions should have some additions made. Let him make that proposal if he is sincere
    and then move on to the real issue. But if we cant clear up the silly issue about the instructions,
    then we can’t hold much hope for the other discussion.

    The instructions are fine. Could they be made more perfect? In retrospec yes. So
    Eric suggests that Anthony should have some kind of warning in the instructions.
    So, since eric is a reasonable rational fellow, he will suggest one. And then, we can
    reccommend him for a job at the New York Times.

  180. Eric McFarland
    Posted Feb 25, 2008 at 12:32 PM | Permalink

    Blessed are the peace makers. I have some work to do for the day … but will chime in later.

  181. Posted Feb 25, 2008 at 1:11 PM | Permalink

    RE176, I’ll be happy to improve/amend the instructions, if something can be worked out that will help the surveyors, accurately and easily locate stations, I’m all for it. Mosh you are correct that the instrctions were a first effort, dated in June for first version and the some changes were made in July. Since then it hasn’t been updated. It probably needs to be since we’ve learned quite a bit about situations related to the NCDC database not being accurate or always up to date.

    Some additional situational awareness imparted to the surveyors would certainly be helpful.

  182. Anthony Watts
    Posted Feb 25, 2008 at 1:38 PM | Permalink

    Eric McFarland @ 167:

    The stupid station has not even been moved … or not at least for 20 years.

    Eric, were you referring to this move that never happened?


    Miami, AZ Map1 above – 1965-1996, click for larger image
    NCDC says this location (red dot) is:
    Location Description: CITY SERVICE CO. LAB, WITHIN & .8 MI N OF PO AT MIAMI, AZ


    Miami, AZ Map2 above – 1996-2007, click for larger image
    NCDC says this location (blue dot) is:
    Location Description: MAGMA COPPER COMPANY WITHIN AND 0.8 MI N OF PO AT MIAMI, AZ

  183. bender
    Posted Feb 25, 2008 at 2:04 PM | Permalink

    Cursed are the trolls. Including those who make waste while attempting to make peace.

  184. Alan S. Blue
    Posted Feb 25, 2008 at 2:09 PM | Permalink

    There’s also two somewhat different sets of instructions that should be collected. The PDF download should basically be the strict superset of instructions – including all of the easy methods listed directly as well as all of the ‘what if that fails?’ information of the PDF.

  185. Sam Urbinto
    Posted Feb 25, 2008 at 2:18 PM | Permalink

    Oh, please, like anyone taking the time out to do some land nav is so stupid they don’t realize the station may not be in the same place it’s reported to be.

    The instructions:

    “About 10% of the time, the sensor in question may be at a location other than the officially plotted location. The sensor location may be different because of measurement error, margin of error for the reading, a station move that is undocumented, or some other reason. In no case will Anthony Watts be responsible if you have to go look for it. But he can help in other ways, for example these hints that you may not have thought of.

    As one example of an issue that may come up even if the station is where it’s tagged to be located, if the reported latitude and longitude are only whole numbers, 1 second of accuracy is about 25 meters, so if a station at 33 24’15″N 110 52’12″W is actually rounded down on both, for example, 15.999 12.999; that’s about 40 meters away. In the case of Miami, AZ, for example, 15 is actually 15.84 which is 21 meters, even if the original number is correct, and accurate on the ground to no margin of error. But what if the North seconds got transposed as 25 instead of 15? That’s 250 meters.

    The written description of the site may help to locate the temperature equipment and verify it is the correct set (in the case of many nearby sensors) (for example,[ BLDG 105/18/9 ] means a building 9 feet tall, 18 feet away, on an azimuth of 105 degrees). However, the only way to verify that the sensor you’ve found is the correct one, and in any case the best and quickest way to locate it in the first place, is to find the curator and get verbal confirmation of the site. Please be aware that policy and civility are that the person’s name is not to be divulged; this should be stressed to the curator, that they will remain anonymous. In no case should you try and make the curator sign a sworn affadavit that the temperature sensor is the correct one, nor infer that they are a lying scumbag weasel purposely showing you the wrong temperature sensor.”

    Something like that.

    You’re not only not working with a flat surface, you’re working with an “oblate spheroid”. Try computing a back azimuth on a bearing and see how close you get to where you started. Drove my office mate nuts.

    Just get a MGRS map of the area at 1:25,000, a protractor and a compass. You can verfiy your position by getting the location of two others and triangulating. But of course, you’re talking about looking at a 3D surface as a 2D map. And we’re also assuming a perfect sphere 6378 KM in radius. Yep.

    Stupid oblate spheroid! :D

  186. Eric McFarland
    Posted Feb 25, 2008 at 4:40 PM | Permalink

    182:
    “In both cases, the abnormal spike coinciding with a station move near the present time remains in the record,”

    Where?

    How did you pinpoint the move?

    When exactly did the move happen?

  187. bender
    Posted Feb 25, 2008 at 4:52 PM | Permalink

    Ummm. 1996-02-21?

  188. Anthony Watts
    Posted Feb 25, 2008 at 4:55 PM | Permalink

    RE186, right there in the graphs provided in the post GISS RAW, GISS Homgenized. Its called a “step function”. Our friend Atmoz knows what to look for:
    “However, a station move should manifest itself as a near step function over a maximum of two years (in the yearly data). ”

    Of course I’m sure you’ll dispute that too.

    So you are not going to say anything about the moves and your “stupid station” statement? Too uncomfortable?

  189. Eric McFarland
    Posted Feb 25, 2008 at 5:09 PM | Permalink

    I guess that’s why the curator’s notes say that the station has been in its present location for 20 years.

  190. bender
    Posted Feb 25, 2008 at 5:16 PM | Permalink

    #193 What are you smoking? 2008-1996=20?

  191. Eric McFarland
    Posted Feb 25, 2008 at 5:24 PM | Permalink

    So, the metadata is right and the curator is wrong?

  192. Anthony Watts
    Posted Feb 25, 2008 at 6:16 PM | Permalink

    RE191 So you are not going to say anything about the moves and your “stupid station” statement? Too uncomfortable?

  193. Eric McFarland
    Posted Feb 25, 2008 at 6:27 PM | Permalink

    192: The issue is … when did the station move. The curator says not for 20 years. Previously, you said the metadata is FUBAR. Now … you’re hanging your hat on the data (without actually answering the question raised) to make your questionable graph appear ok. I am comfortable with continuing to point that out …

  194. Eric McFarland
    Posted Feb 25, 2008 at 6:30 PM | Permalink

    Not to mention:
    Obstructions/
    Exposure: [ BLDG 105/18/9 ]
    [ TOWER 255/160/36 ]
    [ HILL 280-310/90-90/26-26 ]
    [ BLDG 230/180/12 ]
    [ BLDG 180/130/10 ]

  195. Earle Williams
    Posted Feb 25, 2008 at 6:32 PM | Permalink

    It’s a curious conundrum. Curator says it’s been there 20 years. NOAA says it moved there in 1996. I suppose it’s possible they could both be right, if there was some overlap from 1986 to 1996. Who’s to know? Who is correct? If I were Eric McFarland I would say anyone but ‘Watts and company’ because I’m desperate to disparage the surfacestations work. So if the instrument has been there for 20 years that means the NOAA data is grossly inaccurate and cries out even more for … wait for it … bona fide auditing of the station metadata.

    Bender #190
    Eric McFarland is moking this: curator notes at surfacestations.org

  196. bender
    Posted Feb 25, 2008 at 7:03 PM | Permalink

    “When did station X move?”
    “What was the magnitude of change in bias associated with the move?”

  197. K
    Posted Feb 25, 2008 at 7:05 PM | Permalink

    There really isn’t any issue here. But it can tell us a little about gathering material.

    Poor Question: How long has X been there?

    Better Question: When was X put there? Mo Better: When was X put right here? (questioner indicates exact location).

    Better yet: Do you have a record of when X was put there?

    The first is vague. Vague questions often get vague answers of guesses simply because the question implies the answer isn’t very important.

    The second asks for a date. People asked for a date reply tend to reply in a date format. (provided they can.)

    The last asks for a fact. The Curator has a record or does not.

  198. Eric McFarland
    Posted Feb 25, 2008 at 7:05 PM | Permalink

    201: When did it move … bender?

  199. Anthony Watts
    Posted Feb 25, 2008 at 7:09 PM | Permalink

    Eric I didn’t say the metadata was FUBAR, you did. You missed the joke of #158, which said nothing about metadata. But, thats OK, you’ve missed a lot.

    Eric, I’m comfortable too, more comfortable that you know. The volunteer surveyor found the correct station, the station has been moved, the station siting is out of compliance, the metadata and the curator both have problems, the metadata isn’t as accurate as you’d like, and the curators memory may be faulty. Welcome to reality, perfection does not live here.

    So your point is now this:

    “…to make your questionable graph appear ok.”

    Its not my graph, but I like your statement, it shows the depth of research on your part. You mean the graphs from NASA GISS, right? These identical ones?

    RAW http://data.giss.nasa.gov/cgi-bin/gistemp/gistemp_station.py?id=425722780040&data_set=1&num_neighbors=1
    HOMOGENIZED http://data.giss.nasa.gov/cgi-bin/gistemp/gistemp_station.py?id=425722780040&data_set=2&num_neighbors=1

    Yes the NASA GISS temperature graphs are questionable, thats the whole point of this thread. Bad location + siting compliance failure = bad data. But your circular logic and lack of reasearch keeps you from seeing that.

    “I just don’t believe that Watts plays fair … or scientifically. He’s got a clear bias … of which I do not totally approve. We need answers, clarity, rigour … not snake oil.”

    Wow. After all the detailed explanations that have been offered? And just where is the snake oil Eric? Where is the “unfairness” to you? I’ve laid out all the facts, I’ve been patient as have others, and we’ve all tried to help you understand.

    So far you have not done anything but hurl accusations, and you’ve missed basic key points, and won’t even acknowledge that you have. Is that your idea of fairness?

    Go survey a station, get that report from the Phoenix COOP manager you promised, and make those recommendations about the survey intructions, show us that you can be fair.

  200. Eric McFarland
    Posted Feb 25, 2008 at 7:12 PM | Permalink

    204: what I meant to say was your questionable take. I know where the graph came from. The bottom line is that you are insinuating things into the graph … without backing them up precisely.

  201. Eric McFarland
    Posted Feb 25, 2008 at 7:20 PM | Permalink

    So, this is all just a big mistake:
    Obstructions/
    Exposure: [ BLDG 105/18/9 ]
    [ TOWER 255/160/36 ]
    [ HILL 280-310/90-90/26-26 ]
    [ BLDG 230/180/12 ]
    [ BLDG 180/130/10 ]

  202. Anthony Watts
    Posted Feb 25, 2008 at 7:26 PM | Permalink

    RE207 Ok this is circular argumentum ad infinitum

    For Steve McIntyre, I suggest simply closing comments on this thread. I’ve said and done all I can possibly do. There’s nothing new to discuss here anymore.

  203. Eric McFarland
    Posted Feb 25, 2008 at 7:31 PM | Permalink

    208: So, questions you’ve never answered lead to a cry for closing time. As I said before, you want your cake … and you want to eat it to. In any case, let’s leave things like gentlemen: we just disagree.

    BTW — I did not know that it was such a sin to point out glaring and obvious discrepancies. I geuss people really don’t like being audited … do they.

  204. Severian
    Posted Feb 25, 2008 at 7:38 PM | Permalink

    And now finally we see the point, Eric’s convinced that all of Steve’s and others efforts boil down to nothing more than ignorant nitpicking, and being the good AGW troll he is feels that it’s his duty to pay everyone back for daring to doubt the Hockey Team’s results.

    I had hoped it would be something more original and mature, but am not surprised this is what it turned out to be.

  205. Anthony Watts
    Posted Feb 25, 2008 at 7:42 PM | Permalink

    RE 209 The questions were answered. You didn’t like the answers, so you keep changing the questions, while making baseless accusations.

    I do not think that your purpose here is or ever was understanding of the material, but simply to waste time. So, the only solution is end game.

  206. kim
    Posted Feb 25, 2008 at 8:01 PM | Permalink

    Audit in bitterness, understand by accident.
    ============================

  207. kim
    Posted Feb 25, 2008 at 8:02 PM | Permalink

    Audit in curiosity, understand by design.
    ========================

  208. eric mcfarland
    Posted Feb 25, 2008 at 8:11 PM | Permalink

    211: Mr. Watts:
    How about this … I apologize for calling your work snake oil. My words were harsh … not fair … and said for sting … and not for substance. However, I do believe that I raised some valid points regarding some pretty clear discrepancies. There may be some things in your work that you could do differently to address those issues … and we could discuss those things here or elsewhere … if you have any interest. All the best,
    Eric McFarland

  209. Kenneth Fritsch
    Posted Feb 25, 2008 at 8:18 PM | Permalink

    Eric McFarland’s apparent annoyances here are from a rather transparent and failed attempt to let us all at CA know how it feels to be audited.

    I suspect that nobody enjoys an audit that is motivated to teach someone a lesson and not meant to be helpful. In another life, I have been at the receiving end of audit processes by real professionals, and while for some of my colleagues it stated out as a daunting process, in the end the auditers and their processes provided a learning experience and without any personal bitternesses or bickering.

    Unfortunately in my view Eric has not learned from Steve M (or A Watts, for that matter) that auditing comes from an instinct for puzzling solving and bird dog adherence to proper analyses. Eric, it is hard (but fun) work and is certainly more than throwing stuff at the wall to see what sticks. Most surprising to me is the time and attention you have received here.

  210. eric mcfarland
    Posted Feb 25, 2008 at 8:48 PM | Permalink

    215: No. I just noticed major discrepancies in the data … such as that the long. and lat. put out by the surveyor did not match the official data. I had no intentions. My comment above was merely an observation. People around here usually chortle and act smug when auditing efforts raise neck hairs. So, I simply made an obvious statement … with reference to one of CA’s own.

  211. Anthony Watts
    Posted Feb 25, 2008 at 8:58 PM | Permalink

    RE214, Thank you.

    Ok, I’ll offer the benefit of the doubt. At the risk of creating yet another circle, here is the best way I see to continue a dialog, If you can agree to this, then I see a possibility of moving forward.

    1. Make a list of specific questions about what you don’t understand about this station survey. Please be specific about the questions, and limit them to about 2-3 questions total so it is manageable.

    2. I and/or others will attempt to answer the questions. If we don’t understand your questions, we’ll explain why and ask for a rephrase.

    3. If the answers don’t suit you, please don’t respond with new unrelated questions, but with requests for clarification such as “I don’t understand how this is possible in #2 because of X”

    4. Once the first group of questions are resolved, we move on.

    5. Respectful dialog shall rule on both sides, “please, thank you, sorry, and I misunderstood” are encouraged.

    Agreed?

    My interest is in learning if the presentation methodology of the ideas on station site problems is flawed or perhaps incomphrehensible outside of this forum.

  212. Steve McIntyre
    Posted Feb 25, 2008 at 9:45 PM | Permalink

    #210. Business people don’t “like” being audited, but, for the most part, they’re used to it and understand the role of due diligence.

    I put source code online for my published papers specifically so that people could audit my results.

    I do try to be careful, but I don’t claim to be infallible. If some one finds something wrong with anything that I write, I try to correct things. I certainly wouldn’t interpret such an error as evidence that verifying work is misconceived – quite the contrary.

    I haven’t followed the ins and outs of this station; it’s quite possible that someone made an error in understanding the station history. Errors happen when you are dealing with station histories that are poorly documented. All the more reason for checking and cross-checking. Enough snark, folks.,

  213. eric mcfarland
    Posted Feb 25, 2008 at 10:19 PM | Permalink

    Agreed.

  214. Steven Mosher
    Posted Feb 25, 2008 at 10:57 PM | Permalink

    RE 181. Anthony Well, I knew the instructions were written well before you found out some of the anomalies in the locations. So it struck me as Odd that Eric would have required that you have ESP, and would require that you issue warnings about some of the Noaa data being wrong. Imagine how that would have looked at the very start of the project. I figure the best way to make peace with eric is to torment him with kindness and light, for which I am well known. Just as TCO.

  215. eric mcfarland
    Posted Feb 26, 2008 at 12:21 AM | Permalink

    211

    3 questions:

    If NCDC data says that station X was moved from point A to point B, what research do you do on point A as a matter practice to assess the impact of the move on X’s temp. readings?

    If you know little to nothing about point A (other than, say, what old maps may indicate), what if anything can you say about the impact of the move on X’s temp. readings, including without limitation on X’s historical temp. trend?

    If NCDC data says that station X was moved 12 years ago, but the curator claims that station X has been in the same location for the past 20 years … who do you choose for the survey … NCDC data? … or curator testimony? … why?

  216. Steve McIntyre
    Posted Feb 26, 2008 at 12:52 AM | Permalink

    #215. Eric,

    1) the obligation to assess the impact of station moves rests with the constructors of the final results – NOAA, NASA. Anthony didn’t auhorize the moves – indeed, it’s hard sometimes to get correct information.

    2) if you don’t know the impact of station moves, how do you define th “historical trend”? Is it based on raw data? Or adjusted data? And how do you adjust for moves?

    3) One of them is wrong. Either of them could have made a mistake. Seems to me that you need more information.

    But why are you asking Anthony these questions? Shouldn’t you be asking NOAA or NASA?

  217. welikerocks
    Posted Feb 26, 2008 at 6:01 AM | Permalink

    SteveMc,
    you are mentioned in this article from American Thinker about Dr. Hansen. (and a link to CA is given) dated today 2/26/08
    link

    “NASA’s Hansen Reaches Escape Velocity”

    Prof. Hansen and his colleagues argue that rapidly melting ice caps in Antarctica and Greenland could cause oceans to swell several metres by 2100 – or maybe even as much as 25 metres, which is how much higher the oceans sat about three million years ago.

    ” He wants all the 2008 candidates to sign a Declaration of Stewardship for the Earth and all Creation.”

  218. henry
    Posted Feb 26, 2008 at 7:38 AM | Permalink

    eric mcfarland says:

    “3 questions:

    If NCDC data says that station X was moved from point A to point B, what research do you do on point A as a matter practice to assess the impact of the move on X’s temp. readings?”

    So far, we’ve seen that the NCDC data can be inaccurate as to CURRENT location. There’s no guarantee that the locations listed for either point A or B is correct. Thus the need for current metadata, and as much history we can find. If NASA/GISS wants to use this updated data, it will be available. If they have different metadata, then they’re not posting it.

    “If you know little to nothing about point A (other than, say, what old maps may indicate), what if anything can you say about the impact of the move on X’s temp. readings, including without limitation on X’s historical temp. trend?”

    Once again, that’s the point of the thread. Ask the users of the data (and creators of the graphs) the same questions. NASA and GISS are supposed to be the experts on all this.

    “If NCDC data says that station X was moved 12 years ago, but the curator claims that station X has been in the same location for the past 20 years … who do you choose for the survey … NCDC data? … or curator testimony? … why?”

    The notation should include both statements. Remember, when the site was moved, a form had to be filled out and sent in. We can’t see those forms. Maybe better questions can be asked. It would help if we could see all the metadata that exists on the sites.

  219. MarkW
    Posted Feb 26, 2008 at 8:48 AM | Permalink

    It’s up to those who want to use the data to show that either the move had no impact on temperature, or that the effect has been correctly removed.

  220. bender
    Posted Feb 26, 2008 at 9:42 AM | Permalink

    As has been stated many times, these weather station networks were never intended to assay long-term climatic fluctuations. They were designed for real-time weather monitoring, especially air safety. A station move that might not affect real-time meteolorology applications could have a huge impact on climatic inferences.

    To bitch about the lack of records of station moves is misguided. I find it amazing that they would bother at all to record any station moves in any form. #219 has it right. It’s the person who decides that this network is good enough to be used in an application for which it was not designed – climate monitoring – that needs to prove the network has the stated accuracy. This has NOT been done.

    You lurkers – and you know who you are – get off the LA-Z-BOY and just do it!

  221. Eric McFarland
    Posted Feb 26, 2008 at 12:51 PM | Permalink

    216:
    Steve:

    The questions were addressed to AW, but all are free to chime in on the matter at hand.

    The problem that I have with your response, I guess, is that people are still opining on the sites’ historical temp. trends, etc., without addressing (clearly and transparently) the obvious historical issues/questions/limitations that I have raised in this thread. In fact, this very thread began with a thesis that one could say something material (spikes in the data, etc.) about historical temp. trends … not only vis-a-vis the existing conditions of the site … but also vis-a-vis the fact that the site may have moved from point A to point B in the past. No? And, so far, I’ve seen no analysis concerning the conditions at A, or pre. A for that matter. Perhaps conditions were far worst … but, who could say … from the current data presented in the survey … a survey that purports, ultimately, to say something about the quality of the observed historical temp. trend (spikes, etc.)?

    In any case, simply saying that it’s all NASA’s problem (or whoever you want to toss under the bus) seems to duck the big issue/question surrounding the surveys and the originating thesis of this thread. AW is clearly attempting to say something material about the sites’ historical trends … but, perhaps, without doing proper due diligence re the historical conditions affecting those sites … ? The point that I am making here is NOT that surveys should NOT (sorry for the 2x neg.) be conducted (they should) … but that, perhaps, they should have more to say about their historical/data limitations. Moreover, perhaps less definitive statements should be made regarding the sites’ historical trends … where the history is so subject to question …? The originating post included…?

    You may respond by simply saying that NASA, et. al., should do the same. However, that is not the issue right now. I’m sure (or would assume) that those parties would have very different responses and explanations re the quality of ncdc’s historical data … the historical trends of the sites in question … and that would generate a whole different discussion. The question at hand is how does AW deal with the very real and obvious historical data limitations (that I have raised) in connection with his endeavors to make definitive statements re the historical temp. trends in question?

    Sorry for all the words … I did not have enough time to write a shorter note.

    All the best,
    E.

  222. bender
    Posted Feb 26, 2008 at 12:53 PM | Permalink

    people are still opining on the sites’ historical temp. trends, etc., without addressing (clearly and transparently) the obvious historical issues/questions/limitations that I have raised in this thread

    Yes. People like Jones, Hansen, Parker, … who else?

  223. Eric McFarland
    Posted Feb 26, 2008 at 1:04 PM | Permalink

    222: Don’t dodge:
    “You may respond by simply saying that NASA, et. al., should do the same. However, that is not the issue right now. I’m sure (or would assume) that those parties would have very different responses and explanations re the quality of ncdc’s historical data … the historical trends of the sites in question … and that would generate a whole different discussion. The question at hand is how does AW deal with the very real and obvious historical data limitations (that I have raised) in connection with his endeavors to make definitive statements re the historical temp. trends in question?”

  224. MarkW
    Posted Feb 26, 2008 at 1:10 PM | Permalink

    Do you believe that it’s OK for Hansen to use the data collected, without bothering to check if the data is any good?

  225. bender
    Posted Feb 26, 2008 at 1:21 PM | Permalink

    #224 No.

  226. bender
    Posted Feb 26, 2008 at 1:23 PM | Permalink

    #223 Hear that? Jim? Reto? Don’t dodge.

  227. Eric McFarland
    Posted Feb 26, 2008 at 1:46 PM | Permalink

    224:
    Since Hansen is not here … and has not raised a host of responses re the quality of the data in connection with the discrepancies that I noted throughout and above… and since AW asked for my questions … I’d like to stay on topic, if possible.

    Having said that, if the historical data is proven to be bad (which we are questioning here) … No. But I don’t know what Hansen, et al., have said or done vis-a-vis the issue … so, I can’t have that discussion. In contrast, I do know what AW has said here … so, we can have that discussion.

    Will anybody actually respond on point?

  228. bender
    Posted Feb 26, 2008 at 1:48 PM | Permalink

    #227 Yes he is.

  229. henry
    Posted Feb 26, 2008 at 1:50 PM | Permalink

    Eric:

    I thought maybe I missed something (after 224 replies), so I went back and re-read the current thread.

    Anthony took existing GISS data, and compared RAW with HOMOGENIZED GISS DATA. He then made another chart, in which he overlaid the GISS data on the RAW.

    Comment was:

    “Notice that after the GISS homogeneity adjustment, the past temperatures go down, with the present acting as a hinge point, thus making the slope of the temperature trend rise. The new slope is purely artificial, and appears to be an artifact of data adjustment by NASA GISS on this rural station.”

    He’s only using the data supplied by NASA/GISS.

    Sam Urbinto said it best: (February 19th, 2008 at 11:50 am)

    “How can the station be properly adjusted for by GISS (or anyone for that matter) if nobody knows exactly where it is and what’s around it and when and how and why the station was moved? This is unacceptable; how can anyone even claim the data is accurate much less that it has meaning if something so basic is questionable at best? This is more an unknown than anything else.”

    That seems to sum up your arguement. You think Anthony’s responsible. We think NASA/GISS/Hansen et al is responsible.

    Whose data do you think is correct, and why?

  230. Eric McFarland
    Posted Feb 26, 2008 at 1:53 PM | Permalink

    230: You’ve excluded things. I’m not going to re-read the post for you.

  231. Mike Davis
    Posted Feb 26, 2008 at 1:57 PM | Permalink

    I will jump in. Eric when you have put as much time as Anthony and crew in on a project such as this or when you have learned to read you will have a right to question the process. I have seen various sites that have been posted and most of them have a map showing the historical locations and a discussion of those sites. The general meaning of this is how not to measure temp. If you want ot see the good sites go to the web site you have already been reffered to.

  232. Anthony Watts
    Posted Feb 26, 2008 at 2:02 PM | Permalink

    RE 223

    Ok Now I see what your issue is. The problem can be cleared up quickly. I mentioned this in post 157, but I think you missed it.

    From 157: I’m normally not allowed to look at these B44 forms (NCDC does not put them online, nor will they provide them on request, citing “privacy concerns”) but I do have a set for California given to me by the former state climatologist that are about 15-20 years old. I’ve posted the ones online that are for public places like the fire station, in their entirety, but not for any private observers who run stations at residences.

    I’ll then answer your first question:
    It would be a simple matter to track and document all historical changes if I had access to the B44 forms, but I do not have that access. It is hidden from public view. A B44 form is filled out each time a station is moved, and if you will look at this sample I provided: http://gallery.surfacestations.org/main.php?g2_view=core.DownloadItem&g2_itemId=31477

    Here is another, from Pasadena, CA http://gallery.surfacestations.org/main.php?g2_view=core.DownloadItem&g2_itemId=31888

    As I mentioned I have these only for California (and they are 15-20 years old) only because I was given them by the former state climatologist I work with.

    You’ll notice a sketch on the back of the one from Pasadena. Each form has a site sketch (or is supposed to, some don’t get done) which is equivalent in some ways to the site surveys we are doing now.

    So armed with such info, it appears it would be an easy matter to create a historic timeline to go with our surveys, and I would be happy to do so. As a government researcher, anyone at GISS would have such access. But they use a different method, city nightlights, and don’t appear to have done any examination of the measurement environment at all, instead relying upon a questionable proxy to determine site quality. I as a private researcher am denied access to B44 forms and therein lies the crux of the problem.

    NCDC says they will not release this info on B44 forms in the “interest of observer privacy”, but since they already publish names and photos of observers on their NWS websites, it seems a moot point. NCDC does put the monthly dtata report B91 forms online along with transcribed and tabulated data, but full metadata is still behind the firewall.

    Does that answer your first question?

  233. Eric McFarland
    Posted Feb 26, 2008 at 2:11 PM | Permalink

    232:
    Is it fair to say that we have little to no historical data? Or simply put, how would you state it?

  234. henry
    Posted Feb 26, 2008 at 2:16 PM | Permalink

    Sounded like a simple question: Do you believe that NASA/GISS/Hansen made proper adjustments to the temperature record, based on the information posted?

    You’re the one questioning Anthony’s work.

    And as for your statement:

    “Since Hansen is not here … and has not raised a host of responses re the quality of the data in connection with the discrepancies that I noted throughout and above… and since AW asked for my questions … I’d like to stay on topic, if possible.

    Having said that, if the historical data is proven to be bad (which we are questioning here) … No. But I don’t know what Hansen, et al., have said or done vis-a-vis the issue … so, I can’t have that discussion. In contrast, I do know what AW has said here … so, we can have that discussion.”

    I disagree – Hansen’s fingerprints are all over the data. We DO know what Hansen has done vis-a-vis the issue: he made adjustments to the past records, with NO explanation why.

    Steve: THis last sentence is not correct. Hansen has reported adjustments, and while his reporting may not meet best practices standards, it is better than Phil Jones and CRU, where so little is made available that it’s very difficult even to get started with analysis. Hansen’s made enough available that analysis is possible; it’s more work than it should be if the methods were described better, but still your statement is unfair.

  235. Anthony Watts
    Posted Feb 26, 2008 at 2:21 PM | Permalink

    RE233

    Eric you didn’t answer my question in 232.

    I’m not going to respond then, per our agreement until you provide that answer. One of the most irritating habits you have and the reason why you aren’t getting much respect is that you seldom acknowledge any answer given, but simply ask another question.

  236. Earle Williams
    Posted Feb 26, 2008 at 2:44 PM | Permalink

    I admire everyone’s efforts to strive to maintain a civil, rational discussion with the poster calling itself Eric McFarland (or eric mcfarland). This patience for overt trolling exceeds what I can maintain. I posit that this person (or these persons) is playing this thread for whatever muck it can raise regarding a virtual pot-pourri of issues regarding surface station observations.

    It has in this thread alone created several strawmen and attempted to place all blame for its trumpeted mistakes upon Anthony Watts.

    Clear and simple information in the GISS tremperature record and the NOAA metadata is ignored and various incongruities are thrown against the wall in an attempt to make them stick. Disparagement is indeed the motivation despite all its protestations to the contrary.

    And it still ignores the very critical question as to why the homogenized temperature is adjusted down. Perhaps this station moved, perhaps it didn’t. What in the data suggests that it is appropriate to adjust the historical temeratures down by up to 0.5 C? I’m still waiting for eric or Eric to answer that question, but I have no faith or expectation that an answer will be forthcoming. Again, my hat is off to those of you those of you still holding out hope.

  237. Craig Loehle
    Posted Feb 26, 2008 at 3:05 PM | Permalink

    Eric: For a given site, the raw data is adjusted somehow by GISS. When you compare them, the past data have been made cooler, which increases the warming trend. The best effort by Watts observers looking at the station history in most cases shows the stations to have been moved to a worse (warmer) location at some point (like next to a building) and/or to have been engulfed by a town (like the Tucson site–compare the historical photo). In either case it looks like the proper adjustment would go the other way, where temps in recent years should be reduced to fix the problems with siting. Qualitatively the GISS adjustment seems backwards. Whether you can tell exactly when the station moved is not at all the point, unless you are GISS claiming you have properly adjusted the data. Anthony is NOT altering any data or adjusting it.

  238. MarkW
    Posted Feb 26, 2008 at 3:29 PM | Permalink

    If the historical record proves bad.

    That’s no longer in question. At least in terms of being useable for climate studies.

  239. Eric McFarland
    Posted Feb 26, 2008 at 3:30 PM | Permalink

    I think it does, but forgive me for summing it up. You said:

    “I’ll then answer your first question:
    It would be a simple matter to track and document all historical changes if I had access to the B44 forms, but I do not have that access …”

    In sum, for reasons not of your own making, you do not do detailed historical research/analysis re the sites (e.g. re the conditions at A before the site moved to B).

    Fair?

  240. Eric McFarland
    Posted Feb 26, 2008 at 3:31 PM | Permalink

    My last was direct at AW at 235.

  241. Eric McFarland
    Posted Feb 26, 2008 at 3:36 PM | Permalink

    i’m checking out for a bit. Check in later.

  242. henry
    Posted Feb 26, 2008 at 3:44 PM | Permalink

    Steve, Re: 234

    My bad. I guess I just got caught up in the trap. Eric is continuing to try to pin the mistakes on Anthony, without ever answering the question as to whether he thinks Hansen’s methods are wrong.

    He as much as said so in the referenced post:

    “Since Hansen is not here … and has not raised a host of responses re the quality of the data in connection with the discrepancies that I noted throughout and above… and since AW asked for my questions … I’d like to stay on topic, if possible.”

    In short, since Hansen hasn’t commented publically in this forum, his methods are right, and Anthony’s statements are suspect. Strange.

    Snip if necessary. Just had to explain.

  243. Posted Feb 26, 2008 at 4:04 PM | Permalink

    RE239

    As has been demonstrated on this thread repeatedly, I and the volunteers do as much examination of the historical record for each station as is possible, from all publicly available data provided by NOAA/NCDC which does include a record of site moves. This includes examination of the previous site whenever possible. In many cases, we are able to locate the previous site and to document it. I have several examples of that in the surfacetstaions.org database. The NCDC MMS database has locations for previous sites, and they can be located from those descriptions.

    Lovelock, NV is an example, where the previous site was at the old post office. Sites “A” and “B” are both in the database.

    So no, your summary saying that “you do not do detailed historical research/analysis re the sites” would not be fair nor accurate.

    Your focus appears to be different now, you are now creating specific statements with an attempt to get me to agree to them. As another commenter posted “trying to pin mistakes”.

    Now I have a question for you:

    Are you directly connected or indirectly by association/proxy connected with the Sierra Club? The reason I ask is that an almost identical statement has been posed by a known Sierra Club member to me previously.

  244. Eric McFarland
    Posted Feb 26, 2008 at 4:21 PM | Permalink

    243: No.

    How about the next question …

  245. Posted Feb 26, 2008 at 4:23 PM | Permalink

    244 No what?

  246. Severian
    Posted Feb 26, 2008 at 4:24 PM | Permalink

    How about you treat the first one properly…

  247. Craig Loehle
    Posted Feb 26, 2008 at 4:26 PM | Permalink

    I like how eric completely ignored my summary of what SurfaceStations is about which would clear up where he is off track.
    How about responding to my comment, eh eric?

  248. Posted Feb 26, 2008 at 4:36 PM | Permalink

    Stll there, Eric? We all want to hear more about what you meant in 244, since it is not clear.

  249. Eric McFarland
    Posted Feb 26, 2008 at 4:42 PM | Permalink

    Anthony:
    I’m not a member of any group or political party … pretty pathetic if you ask me. I’m just a private citizen … who also happens to post under his real name … unlike some around here. There’s no game here. If you want to shut this discussion off … feel free to let me know.

  250. Posted Feb 26, 2008 at 5:13 PM | Permalink

    RE249 Ok, Thank you. You seem to not like having questions posed of you. Since everything I do with this project is published and open for inspection, I think I’m entitled to ask.

    Second question:

    “If you know little to nothing about point A (other than, say, what old maps may indicate), what if anything can you say about the impact of the move on X’s temp. readings, including without limitation on X’s historical temp. trend?”

    McIntyre already answered this, but I’ll offer additional info on this by saying that you can explore the station history based on known station move dates (which are documented).
    Surfacestations volunteer John Goetz did a detailed analysis of this type which you can read here:

    http://www.climateaudit.org/?p=2745

    Even if there is little to go on from the metadata other than a date and lat/lon from the previous station, tools such as Google Earth can be and have been used to show that previous locations may be/are/are not significantly different than the current location. Examples of this are in the surfacestations.org database.

    And for question 3
    “If NCDC data says that station X was moved 12 years ago, but the curator claims that station X has been in the same location for the past 20 years … who do you choose for the survey … NCDC data? … or curator testimony? … why?”

    I choose neither because the data conflicts. In such a case I ask the surveyor to follow up with the curator to verify the memory. I also investigate the official record by calling the COOP manager for the site to get that info. Once the discrepancy is resolved, the site survey will be amended/corrected.

    Final Question:

    Eric McFarland #66 I have submitted a request to NOAA on this particular device/location. Let’s see what I get back in response … May take a couple of weeks … in all events.

    In this case, since you volunteered early on in #66 to contact the COOP manager, that checking now falls under your responsibility. Though I’m not sure why it should take a “couple of weeks”. How is that going? Have you contacted the COOP Manager there in Phoenix, Keith Kincaid, and gotten a response to that question? Please respond with more than yes/no.

  251. Dave Dardinger
    Posted Feb 26, 2008 at 5:33 PM | Permalink

    BTW, though it might not be the best place, I’m posting this here because it does concern Anthony.

    At present The Drudge Report has a link to a story about global cooling with reference to an analysis of it by Anthony Watts. It will be interesting to see if it sparks interest here and elsewhere in the less-than-warmer world.

  252. Sam Urbinto
    Posted Feb 26, 2008 at 5:50 PM | Permalink

    I’ll just comment simply based upon 3. If one bit of data tells me it has been it its current location for 10 years, and another for 20 years, if one is data from a person, then it’s easy enough to ask: “How long have you worked here, or in other words do you personally know it has been here for 10 years {or 20 years}, or are you reporting what a form says, or what you were told?”

    There are not just two choices; they could both be wrong.

    Thanks Henry. I’ll say it again; if it’s this difficult for us to find out what the facts are, and we’re trying to, what does that say about the people that are adjusting for current or historical readings if nobody knows what they are?

    Eric: Answer the question directly; if Site A is claimed to be at a location by the people that runt the network, and it’s not at the location stated, are not those reporting the location to blame, not the people looking for it?

  253. Eric McFarland
    Posted Feb 26, 2008 at 7:01 PM | Permalink

    251:
    Yea, I’ll follow up. Thanks for the contact in AZ. I went the slow boat route back through Asheville. May not be right away as I just back on my feet from the worst flu that I have ever had in my life … and am behind on other things that relate to matters of bread and butter.

  254. Anthony Watts
    Posted Feb 26, 2008 at 8:11 PM | Permalink

    RE254 Eric,

    This should help you get the answer quickly.

    The phone number there for the COOP Manager at NWS Phoenix, Keith Kincaid, is: (602) 275-0073

    His email is: keith.kincaid.hays[ at ]noaa.gov

    We look forward to your report.

  255. Eric McFarland
    Posted Feb 27, 2008 at 11:28 AM | Permalink

    Just adding this in for the record to date:

    Miami
    Additional Station History Data from MMS
    Date Began Date Ended Lat/Lon Elevation
    meters/feet COOP
    ID WBAN Call
    Sign WMO
    ID Type
    MIAMI
    26 Jun 2007 Present 33°24’N / 110°52’W 1085.1m / 3560′ 025512 . . . COOP-A COOP AB
    21 Feb 1996 26 Jun 2007 33°24’N / 110°52’W 1085.1m / 3560′ 025512 . . . AB COOP
    15 Dec 1986 21 Feb 1996 33°24’N / 110°52’W 1085.1m / 3560′ 025512 . . . COOP AB
    01 Nov 1965 15 Dec 1986 33°24’N / 110°52’W 1085.1m / 3560′ 025512 . . . COOP AB
    01 Jul 1948 01 Nov 1965 33°24’N / 110°53’W 1097.9m / 3602′ 025512 . . . COOP

    Top of Page

    ——————————————————————————–

    NCDC / Climate-Radar Data Inventories / Locate Station / Search

    This page dynamically generated 27 Feb 2008 from:

    http://www.ncdc.noaa.gov/oa/climate/stationlocator.html

    Please send questions or comments about this system to ncdc.webcliserv@noaa.gov
    Please see the NCDC Contact Page if you have questions or comments.

  256. Steven Mosher
    Posted Feb 27, 2008 at 11:38 AM | Permalink

    re 251 I like the legal language being employed by eric. It’s like a deposition. Of course,
    he dont like questions. I’d say this Anthony. When eric coughs up his personal data,
    name, place of residence, place of business, then I would engage him. Otherwise, he can #sand

  257. Eric McFarland
    Posted Feb 27, 2008 at 11:50 AM | Permalink

    Anthony:
    I wrote a brief intro email to Keith this AM. I did not give him any details … just inquired if he would have any information on the site’s location and how its location (through time) is represented in the metadata. I’ll let you know just as soon as I know.
    Cheers,
    E.

  258. EW
    Posted Feb 27, 2008 at 12:32 PM | Permalink

    I did some reading about station homogenization data in Czech – I stumbled upon a PhD Thesis of P. Stepanek dealing with a homogenization of data in Czechia. The homogenization is done on the raw data (even daily) when these behave strangely or are missing. The correction is done based on the closest stations whose data have a high correlation with the one with problematic data point/s. Also, the cause for the erratic points should be searched, if possible. He also mentioned problems with metadata about the stations, especially the historical ones.
    Interestingly, in winter, the correlations between stations decrease with their distance slowly, but are high (approaching 1), whereas in summer, the correlations are smaller and their decrease with distance quicker.
    He also mentioned problems related to introduction of automated measurements and opined, that running parallel manual and automated measurements on few stations only would not solve the homogenization problem. There was no talk about “adjusting” the data without the analysis of the station situation.

  259. Anthony Watts
    Posted Feb 27, 2008 at 12:44 PM | Permalink

    RE257, well what do you say to that Eric?

  260. Anthony Watts
    Posted Feb 27, 2008 at 1:05 PM | Permalink

    And also, Eric, how about sharing that email to the COOP manager?

  261. Eric McFarland
    Posted Feb 27, 2008 at 2:26 PM | Permalink

    261:
    Give me your email … and I will cc you on all communications … including the one sent today.

    260:
    Sounds paranoid to me … what can I say? Is there some specific concern or nerve that I am rubbing aside from being skeptical? If so, be direct … and I will address the issue to the best of my ability.

  262. Stan Palmer
    Posted Feb 27, 2008 at 2:37 PM | Permalink

    http://weather.gladstonefamily.net/site/SLCO

    Does the location issue discussed for the site detailed at the URL above provide insight for this discussion of metadata?

  263. henry
    Posted Feb 27, 2008 at 2:44 PM | Permalink

    Anthony – from the site referenced above, it appears like a European version of surfacestations, including photos of weather sites:

    http://weather.gladstonefamily.net/cgi-bin/wxphoto.pl

    Questions about the surface stations is world wide…

  264. Severian
    Posted Feb 27, 2008 at 2:46 PM | Permalink

    Since you’re choosing this forum to make your allegations and such on Eric, why not post the requested email here?

    And being skeptical isn’t the issue, I suspect you know that already. It’s the intellectually disingenuousness of your tactics and posts that seem to go beyond simple ignorance and poor thought processes into deliberate obsfucation. If you want to be taken seriously, you need to be as forthcoming and transparent as the people who you are criticizing

  265. Eric McFarland
    Posted Feb 27, 2008 at 4:04 PM | Permalink

    Anthony:
    I just had a brief chat with Keith on the telephone. He’s agreed to review an email with detailed questions regarding the site in Miami, AZ. I’d like to send a draft of the email to you for approval before it goes to Keith. Let me know how I can get a draft of the email to you.
    E.

  266. Eric McFarland
    Posted Feb 27, 2008 at 8:12 PM | Permalink

    Here’s a list of proposed/draft questions for Mr. Kincaid:

    What are the GPS coordinates for the MMTS instrument in question? Are they N 33.401983, W 110.869483?

    Do you have any information that would lead you to conclude that the NOAA/NCDC location coordinates (i.e., longitude and latitude coordinates published in the metadata) are erroneous for this particular MMTS instrument?

    How often in your experience are the NOAA/NCDC reported location coordinates for MMTS instruments found to be erroneous?

    Are there any buildings or other structures next to, adjacent to, or near by the MMTS instrument in question? How close is the nearest building or structure to the MMTS instrument in question?

    How many times has the MMTS instrument in question been moved in its operational history? What are the dates of those moves, if any?

    Have there been any other changes made to the MMTS instrument in question (in its current or pre-MMTS form)? What are those changes and when were they made?

    Does NOAA or NCDC have or maintain any pictures of the MMTS instrument in question?

    Please feel free to add any other comments or information that you think would help the public better understand the present and historical location(s) of the MMTS instrument in question.

    I would send them to Mr. Kincaid along with nothing more than some brief introductory remarks and simply ask him to respond. Any comments, objections, suggestions? Please feel free to chime in …

  267. jws
    Posted Feb 27, 2008 at 9:14 PM | Permalink

    How often in your experience are the NOAA/NCDC reported location coordinates for MMTS instruments found to be erroneous?

    Unless this Kincaid fellow is very high up in the food chain, I’d say it’s unlikely he would be the appropriate person to ask this question. I would try to keep your questions focused on the location that he is responsible to track.

  268. Alan S. Blue
    Posted Feb 27, 2008 at 9:32 PM | Permalink

    Number 1 tied up 150 comments or so because you have to state how you’re figuring out the lat/long. Is it from surveying the site, GPS, Google Earth, plotting on a NIST map, etc.

    Number 2 is (nearly) always going to be “no” for anyone who hasn’t hauled out a GPS or plotted it themselves. The question isn’t “Are they _wrong_?” But “How accurate is this measurement really?” Your data from #255 just plots to the nearest _minute_, not even the nearest second. (Unless you’re claiming ’00’ is the correct number?) Heck, we’ve been arguing with you over this – yet I bet a fair number of people _here_ would say “no” to this question as it is worded.

    Number 3 Keith is a COOP guy AFAIK. He isn’t going to have a study about how often the database is in error.

    My additions:
    When and where was the Stevenson Screen when it was operational? Decommissioning date?

    http://maps.google.com/maps/ms?hl=en&ie=UTF8&msa=0&msid=108574942756538540027.000446d76b3f9a55d8778&t=h&z=18

    I have problems posting the above URL as a link (too long or something), but it’s a satellite picture of the area. Which pin most-closely describes the location of the MMTS that is actually in use?

  269. Anthony Watts
    Posted Feb 27, 2008 at 10:28 PM | Permalink

    Eric,

    I agree with jws 267 the question is not really appropriate. Stations existed before GPS, and COOP managers use driving directions off the B44 form (which we don’t have) to locate the stations. They don’t follow a GPS lat/lon to find them for maintenance/inspection AFAIK.

    A COOP manager initially surveys the station, then sets the GPS coordinates for it in the database. The question you ask comes from the reverse angle of thinking, as if the COOP manager were a surveyor trying to find it.

    They don’t expect the station to move, so they aren’t likely to resurvey it with GPS on each visit.

    So please drop that question.

    I would amend this one:

    What are the GPS coordinates for the MMTS instrument in question? Are they N 33.401983, W 110.869483?

    to read:

    What are the latitude/longitude coordinates for the MMTS instrument in question? Are they N 33.401983 W -110.869483 or close by? If close, what would be the approximate error in feet or meters be?

    Would this Google Earth Map in the following link be the location of your MMTS instrument as you know it to be?

    http://maps.google.com/maps?hl=en&ie=UTF8&ll=33.403457,-110.869657&spn=0.002539,0.003712&t=h&z=18

    And add:

    Is this the official Miami, AZ MMTS station, correctly located and identified as seen here near the south corner of the long warehouse building?

    http://gallery.surfacestations.org/main.php?g2_itemId=34343

    and also add:

    How many other MMTS instruments currently exist in Miami, AZ operated by your Weather Service Office? Can the MMTS unit be purchased and used by anyone not affiliated with the NWS COOP program or are they expressly used for it and it alone?

    Please have Mr. Kincaid cc: me directly when he replies to you, rather than having you forward a message.

    my email is info [ at ] surfacestations dot org

    Please send an initial test email to that address to ensure a clear path through the spam filter before sending your list of questions to Mr. Kincaid.

  270. Anthony Watts
    Posted Feb 27, 2008 at 10:37 PM | Permalink

    RE 268, Alan Blue, I just saw you post after composing mine. Nicely done.

    Your detailed map is better annotated than what I offered, but the same view.

    Eric you may want to use his better link. However, since Mr. Blue’s map matches the NCDC metadata pretty well, including identifying the tower mentioned in the metadata, you may realize that the station has indeed been correctly identified, and choose not to waste Mr. Kincaid’s valuable time with your questions.

    If you choose to go ahead, please send me test email per instructions above and then cc: your email to Mr. Kincaid with request that he cc: my email address as well.

  271. eric mcfarland
    Posted Feb 27, 2008 at 11:55 PM | Permalink

    Digesting. Will respond. Will send test mail to AW.

  272. Jim Edwards
    Posted Feb 28, 2008 at 12:11 AM | Permalink

    Eric, #266:

    You proposed these questions:

    How many times has the MMTS instrument in question been moved in its operational history? What are the dates of those moves, if any?

    Have there been any other changes made to the MMTS instrument in question (in its current or pre-MMTS form)? What are those changes and when were they made?

    You raised a question earlier to Anthony about what to do when the caretaker and metadata disagree. It’s a decent evidentiary question to pursue, especially considering that the caretaker could, in some cases, be relying on unreliable hearsay. [Not that the B44 forms Anthony refers to are necessarily more correct, but at least they would have been filled out roughly contemporaneously to any changes - we hope!]

    If Mr. Kincaid says the station moved in 19XX and 19YY it would be nice to know what the basis for that assertion is. Did Joe tell him ? Does he think he remembers ? Did he write it in a personal journal after he moved it himself ? Is he looking at an invoice from when he hired a contractor ? Is he simply looking at the metadata [and potentially misreading it...] ?

    It would be nice to add a request to explain whether he was relying upon specific records or personal recollection for the date and nature of changes in location or instrumentation.

    Good luck.

  273. Severian
    Posted Feb 28, 2008 at 7:20 AM | Permalink

    With respect to this question Anthony Watts suggested:

    What are the latitude/longitude coordinates for the MMTS instrument in question? Are they N 33.401983 W -110.869483 or close by? If close, what would be the approximate error in feet or meters be?

    I’d rephrase it to say:

    “If close, what would the approximate OFFSET in feet or meter be?”

    Often people will take the word “error” as negative, even a negative comment about themselves. When you’re begging for someone’s time like this, choosing a less confrontational word that will get you the same info usually works a bit better.

    With respect to the presence of other MMTS sensors, I’d also ask specifically if the curator knows of any such sensors operated by other people/agencies in the area. I think this is slightly different than the way the similar question above approaches it, but it may just be semantics.

  274. bender
    Posted Feb 28, 2008 at 9:21 AM | Permalink

    Don’t like the term “error”? Use “precision”. More neutral still: “±”. Means the same thing.

  275. Eric McFarland
    Posted Feb 28, 2008 at 11:28 AM | Permalink

    Thanks for the comments. As soon as I find the time, I will put together a revised draft for one more review … then send (hopefully). AW will have final say in all events. Thanks to all.

  276. Sam Urbinto
    Posted Feb 28, 2008 at 7:18 PM | Permalink

    “How close is the actual location to N 33.401983 W -110.869483?”

  277. Anthony Watts
    Posted Feb 28, 2008 at 7:58 PM | Permalink

    RE276 Sam

    I used the Google Earth Plus distance measurement tool, and it measures out at 485 feet from N 33.401983 W -110.869483 to the location of the MMTS at the edge of the building.

    Using the same tool, the distance to the “official” coordinates listed by the NWS of N 33.4044 W -110.87 is 428 feet.

    Simple errors between mapping systems, nothing more.

  278. Phil.
    Posted Feb 28, 2008 at 8:39 PM | Permalink

    Re #277

    I did something similar with the Puerto Maldonado map, the official location of the VOR (navigation beacon) is marked on Google Earth however the satellite image clearly shows the VOR a short distance away (200′ according to the measurement tool).

  279. George M
    Posted Feb 28, 2008 at 10:33 PM | Permalink

    The reference location of the airfield at NAS Rota, Spain is listed on the base maps we were issued to a fraction of a second. When GPS came along, we took one to the point designated on the map and found the numbers disagreed by several hundred feet. Common problem worldwide.

  280. Posted Feb 29, 2008 at 11:38 AM | Permalink

    I missed this thread when it was originally posted. I don’t have time to read the comments, so pardon me if this has already been said. It looks to me that the jump in temperatures does not correspond to any dates in the station history. Instead, it appears to be several months after one of the recorded dates (not an actual move), perhaps suggesting that the station moved but was not documented.

    For more details, see Miami, AZ: Undocumented Station Move?

  281. Eric McFarland
    Posted Feb 29, 2008 at 12:52 PM | Permalink

    280:
    Please let me know if you have any (fewer … better) questions for the COOP manager. I will probably have a revised list of draft questions for Mr. Watts this weekend. Thanks. E.

  282. Eric McFarland
    Posted Feb 29, 2008 at 1:07 PM | Permalink

    280:

    I’ll post my question here too:

    # Eric McFarlandon 29 Feb 2008 at 12:07 pm
    Is there a specific reason why you can rule the “jump” out as being natural (i.e., a real temp. jump) vs. a move induced or land use induced jump?

  283. Anthony Watts
    Posted Feb 29, 2008 at 1:45 PM | Permalink

    RE280, It could also simply be that they spread a fresh patch of that black gravel all around the vicinity. A previous commenter who worked there said it was a product of the mining process.

    Such an event could easily affect temperature given its proximity. And of course, NOAA/GISS/NCDC would have no record of it, and corrections would miss it wholesale.

    See the aerial view here, that sensor sits in the middle of an island of blackness compared to surrounding soil albedo. Sometimes simple explanations exist. I’m surprised Atmoz didn’t pick up on this. It is a darker albedo than even the Tucson U of A weather station in the parking lot.

    http://maps.google.com/maps?hl=en&ie=UTF8&ll=33.403457,-110.869657&spn=0.002539,0.003712&t=h&z=18

    It is also on the south (sunlit side) of a metal building with a concrete vault next to it, which makes a splendid night time heatsink.

    Microsite bias is the likely cause. I’d say for this sensor, it is surely man-made warming ;-)

  284. Sam Urbinto
    Posted Feb 29, 2008 at 2:09 PM | Permalink

    Re: 277 Anthony I was just saying in #276 that’s how I would ask the distance question.

    One thing everyone should keep in mind is that there’s two ways to do decimals; conversion from DMS and via GPS where some systems convert the seconds are converted to decimal minutes as a minute value. . You will get errors if you mix them.

    For example, 110.555555 converted from DMS would be 110.333333 in some GPS systems. That should be in addition to any issues from using the wrong datum.

    I think that’s like 14 KM if I have that correct. Anyway, there’s a difference.

  285. Eric McFarland
    Posted Feb 29, 2008 at 2:28 PM | Permalink

    The question for me still remains … where exactly is the damn instrument? The official map does not place it over the black gravel … gravel which can be black when newly spread … but which arguably can bleach pretty quickly thereafter … unless that’s its dry color or its mixed with something like tar. If you’ve every actually spread new gravel or dirt in your life … you’ll know exactly what I mean. In any case, simply postulating or assuming the cause isn’t very satisfying for me. That’s why I hope the COOP manager may will have some concrete answers for us. We’ll see … I’ll get back to you with those questions this weekend, Anthony.

  286. Posted Feb 29, 2008 at 2:29 PM | Permalink

    RE281-283:
    The jump occurs in June 1998. Over 2 years after the last “station move”.

    Regarding why I think the jump is not natural, off the top of my head I can’t think of a better response than we don’t see this behavior in geophysical variable very often, especially temperature. The temperature for a typical natural system at time T is going to be highly dependent upon the temperature at time T-1 (for time steps ~

  287. Posted Feb 29, 2008 at 2:32 PM | Permalink

    RE281-283:
    The jump occurs in June 1998. Over 2 years after the last “station move”.

    Regarding why I think the jump is not natural, off the top of my head I can’t think of a better response than we don’t see this behavior in geophysical variable very often, especially temperature. The temperature for a typical natural system at time T is going to be highly dependent upon the temperature at time T-1 (for time steps ~< monthly); it has a high lag-1 autocorrelation. The step does exhibit this behavior, so the natural conclusion is that it’s not natural.

    [Silly < sign. Feel free to delete that first attempt.]

  288. Eric McFarland
    Posted Feb 29, 2008 at 2:37 PM | Permalink

    286:
    Pardon my dullness … but that would appear to imply that it’s not the alleged move per se that caused the jump … but some “x” factor … perhaps in conjunction with the move? And, would it be possible to separate out the affects … say move caused “a”% of the jump while “x” caused “b”%? Or … are we off on a mystery here?

  289. Severian
    Posted Feb 29, 2008 at 2:43 PM | Permalink

    Nature tends to abhor discontinuities almost as much as she does a vacuum.

  290. Sam Urbinto
    Posted Feb 29, 2008 at 3:12 PM | Permalink

    Severian, you’ll get in trouble calling nature abhor.

  291. Anthony Watts
    Posted Feb 29, 2008 at 3:13 PM | Permalink

    Eric you keep missing this key point. It’s right in front of you. It’s always been right in front of you.

    A lat/lon (of an unknown geographic mapping system) placed on Google Earth may or may not line up with the picture.

    But you can look at the picture provided by this link, and see it.

    http://maps.google.com/maps/ms?hl=en&ie=UTF8&msa=0&msid=108574942756538540027.000446d76b3f9a55d8778&t=h&z=18

    Now look at the 255 azimuth line, which is the approximate angle angle this photo was taken from

    http://maps.google.com/maps/ms?hl=en&ie=UTF8&msa=0&msid=108574942756538540027.000446d76b3f9a55d8778&t=h&z=18

    then look look at this picture taken by the site surveyor:

    http://gallery.surfacestations.org/main.php?g2_itemId=34343

    and this one

    http://gallery.surfacestations.org/main.php?g2_itemId=34364

    and this one

    http://gallery.surfacestations.org/main.php?g2_itemId=34370

    Now compare the buildings (and the tower) in the photos to satellite photo from Google Earth, and notice carefully the building structures. You can easily tell where the ground photos were taken from and where the sensor is.

    If you can’t tell where “the damn station” is from this, you are either wholly incapable of simple photo interpretation or you are being disingenuous for some reason.

    It’s quite plain where the station is to everybody else here in the discussion so far. No other person in this discussion seems to think the station has not been correctly located or identified.

  292. eric mcfarland
    Posted Feb 29, 2008 at 3:32 PM | Permalink

    291: I see that actually. What if it turns out that it has always been there?

    At any rate, I am reposing a question here:

    eric mcfarlandon 29 Feb 2008 at 2:33 pm
    Another dull question from me. Will the jump continue to stand out no matter what happens over the next 20 years — say even if temps drop? I guess what I am getting at … is there any data squish or extra effect the choice of graph itself is having on the data?

  293. Anthony Watts
    Posted Feb 29, 2008 at 3:37 PM | Permalink

    RE 292, Eric. “I see that actually”

    So you agree now that the correct station has been located, identified, and photographed?

  294. Sam Urbinto
    Posted Feb 29, 2008 at 3:43 PM | Permalink

    I still don’t see it, I mean, you photoshopped that in there.

    You and your waste of time photos. You haven’t disproven global warming!!!! Denier!!!! :D

    Anyone want to hazard a guess who wrote this? I highlighted the dig on the surface stations efforts, but you can rather see the tone by the rest, I think.

    If science is to be audited, it should be audited by those who are adequately qualified to do so and who have no politicial motivations with respect to the outcome. The purpose of an audit is to confirm that the processes and procedures and regulations were all followed properly and that the conclusions are correct. What has auditing done so far? Called into question the use of a statistical method to analyse paleoclimate proxy data, called into question the use of a particular set of proxy data, pointed out errors in the global instrumental record that ultimately had no bearing on the conclusions, and posted pictures of weather stations in Bumsrush, Nowhere. Has it undermined the science and trashed the conclusions? No. None of the science bodies have retracted their statements in light of the auditor’s claims. Few if any of the science papers being published in peer-reviewed journals, where science debates rightfully occur, have decided that global warming is not happening, is not largely the result of human activity, or is not a threat. It has not convinced anyone but people predisposed to be against the notion based on a political persuasion.

    Break over, lets get back to shaking the foundations of climate science and discrediting the hypothesis of AGW. ExxonMobil needs your best during these trying times. What the heck do I pay you people for, anyway? :)

    Full post

  295. Kenneth Fritsch
    Posted Feb 29, 2008 at 4:08 PM | Permalink

    I think that Susann and Eric both are rather transparent in their views and disdain for what is being done at CA or what they perceive is being done here. Susann is a more disingenuous when she continues to claim she is agnostic on the subject. It would not matter their dispositions if one thought they could learn from their discourse, but in my view that is not the case. What does surprise me is not their comments and views, but the time that CA regulars spend (waste) attempting to engage them.

  296. Eric McFarland
    Posted Feb 29, 2008 at 4:26 PM | Permalink

    I really hate it when the mob turns out with torches and stones … and sour looks on their faces. I sincerely have questions … have raised some legitimate issues re discrepancies in the data (particularly with respect to the date off the move, if any, as it appears now) … and have never as much as uttered the word photo shop. So, please take your high horses out for some green grass and cool water.

    Meantime, I have 2 other questions that I’d like to propose … which may have to go further up the food chain:

    1. Who are the end users of the data collected by the MMTS in question?

    2. For what purposes do they use the data?

    Any thoughts, objections, etc?

  297. K
    Posted Feb 29, 2008 at 4:29 PM | Permalink

    Eric #291. Why introduce the next two decades into the discussion? Which is not to say I understand your ‘data aquish’ question anyway. Anyway, graphing is about analysis, not the survey. Why not proceed with finishing the survey to your satisfaction first?

    To quote, probably futilely, from my #135.

    “Know what? Exact coordinates don’t really matter. What matters is whether the real station was surveyed. i.e. did that station produce the data we attribute to it?”

    To that I now add. Knowledge is limited. We may never know when new gravel was put at the site. Or how fast it weathered changing the albedo. Perhaps a former employee parked a big black SUV next to the station for a few years. We won’t know.

    Nothing has moved past the comments from #45-#49. There seem to be two questions. Why did the curator say 20 years? And, for those who give a hoot: exactly where, within a metre or so, is the station? Neither will be settled from data we have at the moment.

    You questionaire may yield a fact. If it gets sent and answered. Miami isn’t going to be dynamite that shatters the windows at NOAA, or NASA, or Anthony Watts. At best it hones the skills of those examining the matter. They learn more about what works and what doesn’t; what seems weak and what reliable.

  298. Posted Feb 29, 2008 at 4:33 PM | Permalink

    RE 297, 293

    Eric you still have not answered my question.

    Do you agree now that the correct station has been located, identified, and photographed?

  299. Eric McFarland
    Posted Feb 29, 2008 at 4:33 PM | Permalink

    298:
    Amen.

  300. Eric McFarland
    Posted Feb 29, 2008 at 4:39 PM | Permalink

    299:
    Anthony:
    I am officially reserving my judgment until we get some answers back. Just a policy position … no more … no less.

  301. Posted Feb 29, 2008 at 4:39 PM | Permalink

    Sorry, I am a bit late to the party. My name is Warren Meyer and my son Nicholas and I did this particular survey. Mr. McFarland and other seem to be questioning the accuracy of the survey. I must admit I don’t totally follow his concerns, since it seems to be shifting all of the place. So, I thought I would give some of the background on this site survey.

    Before I give the process we went through, I will try to address the fact I am not an “expert” or am somehow “biased.” To a large extent, this is hugely irrelevant. The main aspects of the survey are 1. taking photos and 2. Using a handheld GPS. I am fully qualified in both. One might argue that our observations, such as “it is within 8 feet of a metal sided building” might be subject to some bias or lack of expertise, but I think the photos serve pretty well to let third parties confirm these observations. I will add that much of the coop network has for 100 years been run by amateurs and volunteers, so to impugn the survey for using amateur volunteers is to impugn the surface temperature record for the very same reason.

    Here is what we did:

    1. We first had to find the dang thing. For anyone who has not tried this, good luck. Part of this is technology. We are so used to being able to use a handheld GPS to get lat/long to 5 decimal places that we forget they did not always have this. So there are always errors in the NOAA data base, including a lot of spurious data points that don’t really point to different sites but to different GPS readings of the same site. You can look at Gunnison, CO, and you will see old locations in a field where the instrument has NEVER been – it is just spurious GPS readings in the data base.

    2. Seeing that the site was on private property, I called the landlord (BHP copper) and got in touch with an extremely nice lady who gave us permission to visit, got us a gate pass, as showed us to the location.

    3. I also got in touch with a Mr. Kinkaid who is the government coop manager for this area. I asked him questions about some other sites I was surveying and told him I was going to survey Miami. He did not seem that interested, but he was courteous and helpful.

    4. We then drove up, went through the gate, and was shown the site by the nice lady who helped us. We asked her the history of the site as she knew it, which I put in my survey. My son and I took notes of the area, we took pictures from all angles, and we took a GPS reading (lat/long/altitude) with a handheld GPS (Garmin GPSmap 60CSx) holding the GPS right next to the instrument. We did not rely on google or any other data base for the lat/long, we read it right there.

    5. We thanked the nice lady and drove home. All tolled it took us about 5 hours including driving.

    As to the implication that this is somehow the wrong station, I just don’t see it. The NOAA lists the site as being on BHP Copper Mine property. The BHP copper folks confirmed that this is the only station they know of on their property. There is no reason anyone else would have a MMTS station here (I have found a few private weather stations near USHCN stations. This was the case at Sacaton, where I also did the survey, and in that case I called the COOP station manager Mr. Kinkaid to make sure I had the right station. I can say that private weather stations are almost always more complete with this one, with wind gauges and such.)

    Having done a number of these surveys, I found nothing unusual about this one. The location was not exactly as in the data base, but that is quite normal. The MMTS station is very close to a building, I have seen that in every one of my surveys with an MMTS station (I am told that is because of cable length limitations on the MMTS). Really, this was kind of an average installation — better than some (like the Tucson station in the parking lot) but worse than others (given how close it was to a concrete and metal building).

    I am not sure what an “expert” would have done differently in this survey, and to be honest, the “experts” have had about 100 years to get a better database together on USHCN installations and they have not done so. My goal in each of these surveys is to gather facts – eg. altitude, lat/long, pictures that those with expertise can use to do analysis. In a sense, I am trying to save the experts some time, performing the grunt work of driving out to these sites so others can do the analysis. However, I do have a brain (and a mechanical engineering degree, by the way) and I feel myself fully able to make a statement like “having a metal building reflect heat on a thermometer during the day, and a concrete building radiate stored heat on a thermometer during the night, probably imposes some upwards bias on the instrument’s readings.

    By the way, I don’t think the point of this exercise is to put coop managers on the defensive. They tend to have limited resources and I am sure none of them made the decision to standardize on the MMTS design with too short of a cord. That being said, what we are doing here is something the COOP managers should have been asked to do. A couple of us have done 3/4 of the stations in Arizona as volunteers, so it is not that hard. I was told once by a woman, I do not have her name here with me, who is apparently the AZ state climatologist that they had already taken “tons” of pictures of their coop stations. Of course, she was not sure where they were and no, they are not available online.

    I will try to read through the comments and see if there is anything else about the survey I can answer, but I certainly resent the implication that I have somehow fabricated this data.

  302. Posted Feb 29, 2008 at 4:43 PM | Permalink

    RE302, 301

    Well Eric there’s your “answer back”

    Same question: Do you agree now that the correct station has been located, identified, and photographed?

  303. Posted Feb 29, 2008 at 4:46 PM | Permalink

    Eric, 292:
    For the purposes here, I am going to assume there are 3 spatial scales that matter: global, regional, and local. Of course, in reality there is a continuum of spatial scales from really small to really big.

    What I am looking at in the plots is how the local temperature differs from the regional temperature. I defined the regional temperature as the average of the 10 closest stations It’s actually slightly different than this, but it’s insignificant for this discussion. If the global temperature goes up/down, both the regional average and the local temperature go up/down. If the regional temperature goes down/up then the regional average and the local temperature go down/up. In both cases, the difference between the regional temperature anomaly and the local temperature anomaly is zero.

    The only situation where the difference is non-zero is if there is a difference between the regional average and the local temperature. This can occur because of deviations in the regional average or because of deviations in the local temperature. Because of the way I defined the regional average, local effects at the regional stations will have only a small effect on the average. Thus, changes seen on the plot are most likely the result of local temperature changes not seen at the regional and global scales.

    The jump will always appear in the plot of the raw data. The jump will disappear in the filnet (supposedly fixed) data when/if it actually gets corrected for this type of bias.

  304. Posted Feb 29, 2008 at 4:54 PM | Permalink

    OK, reading the comments, I missed something important. There is discussion in the comments as to when this instrument came to this site. The lady we talked to at BHP said she thought the instrument was moved to this location around 1986-1990. Sometimes these folks are correct, and sometimes they are not. This date, though, would seem to fit the change of equipment in the NOAA database to MMTS in 1986. I can write a pretty good hypothesis about the station being a stevenson screen before this, further out by itself, but being moved to this building due to MMTS installation restrictions. Of course that is a guess, because the NOAA data base doesn’t bother to tell us such things. Just to add one complication, I think that at some point in the last 20 years the mine was closed for a while, but I am not sure about that. A lot of copper mines did close for a while and reopen more recently.

    By the way, for all this discussion, does no one else find it interesting that, whenever the instrument moved to this site or whenever the building was built, the NOAA folks and COOP folks NEVER saw anything relevant enough in a metal and concrete building 8 feet away to mention it in the site description? The description, even today, reads “TOPO-SMALL LEVEL AREA IN MIDST OF MINING OPERATION. ROUGH BROKEN MESAS SURROUNDING.” That is how the experts described the site. You have seen how I, the amateur described the site. Who has the better description? Who includes more relevant information? And I have not even mentioned the airconditioning exhausts a bit further down the building or the wastewater ponds filled with warm water just down the hill.

  305. Kenneth Fritsch
    Posted Feb 29, 2008 at 4:54 PM | Permalink

    299:
    Anthony:
    I am officially reserving my judgment until we get some answers back. Just a policy position … no more … no less.

    And believe me folks when I tell you: this will never end.

  306. Sam Urbinto
    Posted Feb 29, 2008 at 5:09 PM | Permalink

    Um, Anthony, well, uh, maybe you’lll get an answer one of these days.

    Dude, Eric, what’s your problem?

    Good job, coyote. I don’t think Eric is saying you falsified anything, it just seems he just has some strange idea that the sensor should be where it’s listed and that satellite images should show it at that location, and that if either of those are incorrect, there’s 10 or 20 sensors in the area that could be mistakenly found. I think it’s fairly obvious he hasn’t read the trials and tribulations of finding multiple possible sensors and not being able to figure out which is the correct one; hardly actions made by people trying to pass off the wrong station as the correct one.

    Atmoz, I have to disagree with you on some of that.

    If the global mean temperature anomaly goes up/down, some regional averages will be up and some down, but overall in total up/down. If a regional average goes up/down, some local will be up and some down, but overall in total up/down.

    If a grid goes up, not every station will be up for sure, although of course, the fewer stations, the more of a chance they’ll all be doing the same thing, depending upon how the grid is made up and where the sensors are.

    On the other hand, I can have the same mean from very different min/max readings, and even in that case, we don’t always know if two places with the same min/max are comparable; two places with a mean of 50 due to 70 max and 30 min might have had most of the 24 hour periods closer to one or the other of those two readings. Depends on location.

    I have certainly walked outside one day where most of the day was overcast but it was warm and humid the entire morning, and a cold front came in late afternoon and dropped the temps over 50 degrees for the rest of the 7 or hours in the day. I don’t see how a lot of stations take that into account, and certainly in the past with mercury thermometers it was even less percise.

    ————————————

    Just say it three times before you go to bed; the anomaly is not temperature. :)

  307. Posted Feb 29, 2008 at 5:14 PM | Permalink

    RE306, Kenneth, yes I’m beginning to think that Eric’s goal is not “sincerely have questions” but is truly disingenuous time wasting at best.

    I think Atmoz would concede that the correct station has been found. Mr. McFarland, despite overwhelming evidence, wishes to “reserve” judgment and won’t answer the question.

    In 280 Eric says: “Please let me know if you have any (fewer … better) questions for the COOP manager.”

    Then in 297 he adds two more.

    1. Who are the end users of the data collected by the MMTS in question?

    2. For what purposes do they use the data?

    These have ZERO RELEVANCE to the question at hand.

    Eric you are all over the map, and you just keep adding to a questions that has been asked, answered, answered, answered, and answered again, by many people.

    You can waste Mr. Kincaid’s time if you wish, but the evidence is overwhelming and clear already. And you keep stalling and adding and tweaking, and dodging, on asking the most important question of Mr. Kincaid, you have time to post on this forum but not not finalize the list and send it to him? That’s truly disingenuous.

    Same question: Do you agree now that the correct station has been located, identified, and photographed?

    The people doing this work have done so in good faith, laid it all on the table, and gone to great lengths to accommodate you. However, I think that time is now at an end. You’ve had plenty of time to contact Mr. Kincaid and keep delaying.

    Answer the question.

  308. Eric McFarland
    Posted Feb 29, 2008 at 5:28 PM | Permalink

    Anthony:

    “So I though I’d take a look at the raw GISS temperature plot for Miami, AZ to see if the move would show a spike, it did:”

    Your lack of diligence … and over eagerness to prove a point started this. I have never questioned the honesty of the people who did the survey.

    I all events, I will be presenting questions to Mr. Kincaid (something, frankly, that you should have done on your own) … so cool your jets and allow the process to work itself out. It seems like your are now afraid of what Mr. Kincaid will say. Why … I have no idea.

  309. Sam Urbinto
    Posted Feb 29, 2008 at 5:49 PM | Permalink

    How did you divine the idea that Anthony was scared of what Mr. Kincaid would say?

    And in any case, why would Anthony ask him anything when Mr. Meyer kindly went out and gathered the information on the sensor; are you suggesting Anthony should do all the surveying, check up on all the surveyors, talk to all the people involved in the network or some such?

    Pretty interesting from somebody that’s not surveyed 1 site and asks a bunch of questions without ever answering any himself.

    Get back with us when you have some answers that make you happy, on a subject that nobody except you isn’t totally convinced has been taken care of. Other than that, you’re like the fish that jumped over the lawn with a stick in its mouth.

  310. Posted Feb 29, 2008 at 5:50 PM | Permalink

    RE310
    Ask if you wish, I do not fear the answer, in fact I’m COUNTING on it, but no more stalling. As far as I’m concerned you are being disingenuous with all your stalling and piling on of more and more questions without resolving the first ones.

    As for “cool my jets”, lets examine some of your language in this thread:

    “the damn station”, “mob turns out with torches and stones …”, “True believer … you apostles of Watts”, the moves and your “stupid station”, “I just don’t believe that Watts plays fair …”, “…snake oil”.

    So it seems that Eric has had quite a bit of hot language in this debate, yet gives nothing back, he can’t even finish the list, despite having ample time. Truly disingenuous.

    I respectfully ask, one last time: Do you agree now that the correct station has been located, identified, and photographed?

  311. Stan Palmer
    Posted Feb 29, 2008 at 5:55 PM | Permalink

    Your lack of diligence … and over eagerness to prove a point started this. I have never questioned the honesty of the people who did the survey.

    Your issue seems to be with the assertion in the original post that the move caused the “spike” in temperature. How do your questions to Mr. Kincaid address this issue?

  312. Eric McFarland
    Posted Feb 29, 2008 at 6:05 PM | Permalink

    313:
    Official data says no move.

    Curator says been in same place 20 years.

    AW argued just the other day that move was 12 years ago.

    Today … we learned that something may have happened in about 1998.

    Kincaid may have information to resolve.

    Good practice to follow up and ask … in all events. Something AW should incorporate as a practice … perhaps.

  313. Posted Feb 29, 2008 at 6:13 PM | Permalink

    Sam Urbinto, 307:
    The procedure I outlined in 304 was to find the effects of station moves, and other factors, on the temperature record. I was going to save this image for my post tomorrow, but I guess I’ll post it here.

    The red dashed lines indicate dates when the surface station at Saguache was moved. The dramatic temperature changes seen suggest that the procedure seems to work well.

  314. Posted Feb 29, 2008 at 6:21 PM | Permalink

    RE314 “Something AW should incorporate as a practice … perhaps”

    Well you were offered an opportunity to provide input to the survey methodology back in comment #181 days ago, so far nothing constructive from you Eric.

    You are dodging again Eric.

    You have no good reason for the delay in submitting the questions to Mr. Kincaid. You’ve spent a lot of time on this forum posting additional questions and comments, yet ignore that duty, you keep stalling. You have no excuse for why you have not completed that task.

    You are all take, but no give. I see your methods as disingenuous.

  315. M. Jeff
    Posted Feb 29, 2008 at 6:29 PM | Permalink

    Anthony Watts, I have to compliment you on your efforts to inform and on your remarkable patience with those who resist your efforts in that regard.

  316. Anthony Watts
    Posted Feb 29, 2008 at 7:42 PM | Permalink

    Since I’m traveling the next few days, and will not be able to keep up, I’m closing comments on this thread. If Eric sends me that email for Kincaid, I’ll open them again.

  317. Posted Mar 2, 2008 at 9:01 PM | Permalink

    7:06 PM PST Sunday OK comments open again since I have a stable internet connection. Eric lets see that list of questions for the Phoenix COOP manager please.

  318. Anthony Watts
    Posted Mar 3, 2008 at 6:55 AM | Permalink

    Well the weekend is over,

    Since “Eric” still hasn’t posed the questions to the COOP manager in Phoenix, but rather chose to use that time to pose questions and challenges over on unthreaded #31 under a new name aka “Mr. Stovepipe”, I decided to simply send them myself. Here is the letter I sent this morning to the COOP manager:

    Mr. Keith Kincaid
    COOP Manager
    NWS WSFO
    Phoenix, AZ

    3/3/08

    Dear Mr. Kincaid,

    As a result of a recent photo survey done by Warren Meyers and his son as part of his school science project, some questions have arisen regarding the COOP station in Miami, AZ. If you would be so kind, there are questions below regarding the site, ranked in order of importance, that we would greatly appreciate answers to. If you don’t have the time or inclination to answer all of the questions, the first three would be quick and the most valuable to our discussion.

    When you reply, would you do a “reply all” please so that both recipients get the message? Thank you for your consideration.

    Anthony Watts

    ——————————————————————————–

    1. Is this the official Miami, AZ MMTS station, correctly located and identified as seen here near the south corner of the long warehouse building?

    http://gallery.surfacestations.org/main.php?g2_itemId=34343

    2. Would this Google Earth Map in the following link be the location of your MMTS instrument as you know it to be?

    http://maps.google.com/maps?hl=en&ie=UTF8&ll=33.403457,-110.869657&spn=0.002539,0.003712&t=h&z=18

    3. How many other MMTS instruments currently exist in Miami, AZ operated by your Weather Service Office? Can the MMTS unit be purchased and used by anyone not affiliated with the NWS COOP program or are they expressly used for it and it alone?

    4. If this is the official station, do you have any information regarding when the black gravel seen in the photo may have been applied near the sensor?

    5. What are the latitude/longitude coordinates for the MMTS instrument in question? Are they N 33.401983 W -110.869483 or close by? If close, what would be the approximate offset in feet or meters be?

    6. If this is not the correct or very close latitude/longitude, do you have any information that would lead you to conclude that the NOAA/NCDC location coordinates (i.e., longitude and latitude coordinates published in the metadata) are erroneous for this particular MMTS instrument?

    7. How often in your experience are the NOAA/NCDC reported location coordinates for MMTS instruments found to be erroneous?

    Are there any buildings or other structures next to, adjacent to, or near by the MMTS instrument in question? How close is the nearest building or structure to the MMTS instrument in question?

    How many times has the MMTS instrument in question been moved in its operational history? What are the dates of those moves, if any?

    Have there been any other changes made to the MMTS instrument in question (in its current or pre-MMTS form)? What are those changes and when were they made?

    Does NOAA or NCDC have or maintain any pictures of the MMTS instrument in question?

    Please feel free to add any other comments or information that you think would help the public better understand the present and historical location(s) of the MMTS instrument in question.

    Steve: I moved Eric’s comments as Stovepipe over to Unthreaded as it was OT on another thread.

  319. Posted Mar 3, 2008 at 8:35 PM | Permalink

    VERIFICATION – its the real station (as if everyone except Eric McFarland didn’t already know that)

    awatts@xxxxx.com
    Date: Mon, Mar 3, 2008 at 4:59 AM
    Subject: Questions about COOP station
    To: keith.kincaid.xxxx@noaa.gov <mailto :keith.kincaid.xxxx@noaa.gov
    Cc: Eric McFarland <mcferic@xxxxxx.com

    Mr. Keith Kincaid
    COOP Manager
    NWS WSFO
    Phoenix, AZ

    3/3/08

    Dear Mr. Kincaid,

    As a result of a recent photo survey done by Warren Meyers and his son
    as part of his school science project, some questions have arisen
    regarding the COOP station in Miami, AZ. If you would be so kind,
    there are questions below regarding the site, ranked in order of
    importance, that we would greatly appreciate answers to. If you don't
    have the time or inclination to answer all of the questions, the first
    three would be quick and the most valuable to our discussion.

    Thank you for your consideration.

    Anthony Watts
    ————————————————————————

    1. Is this the official Miami, AZ MMTS station, correctly located and
    identified as seen here near the south corner of the long warehouse

    building? *yes*

    http://gallery.surfacestations.org/main.php?g2_itemId=34343

    2. Would this Google Earth Map in the following link be the location

    of your MMTS instrument as you know it to be? *it appears to be*

    http://maps.google.com/maps?hl=en&ie=UTF8&ll=33.403457,-110.869657&spn=0.002539,0.003712&t=h&z=18

    3. How many other MMTS instruments currently exist in Miami, AZ

    operated by your Weather Service Office? *none*

    Can the MMTS unit be purchased and used by anyone not affiliated with
    the NWS COOP program or are they expressly used for it and it alone?

    *I do not know*

    4. If this is the official station, do you have any information
    regarding when the black gravel seen in the photo may have been

    applied near the sensor? *No*

    5. What are the latitude/longitude coordinates for the MMTS instrument
    in question? Are they N 33.401983 W -110.869483 or close by? If close,

    what would be the approximate offset in feet or meters be? *lat and
    long are taken from the location of the rain gauge. the lat/long you
    have is very close.*

  320. Eric McFarland
    Posted Mar 3, 2008 at 10:18 PM | Permalink

    That still leaves the “move” unresolved … as well as the cause for the apparent jump. And yes, it is the real station by all accounts. Thank you for your patience and assistance in getting the questions submitted.

Follow

Get every new post delivered to your Inbox.

Join 3,306 other followers

%d bloggers like this: