How not to measure temperature part 73, in the middle of nowhere

The idea with measuring climate accurately, is to get as far away as possible from human/urban influences so that those things don’t bias the readings of the thermometer. For example, on my way from Las Vegas to Reno this week, I passed through the near-ghost town of Mina, Nevada, which has a USHCN station. Mina is about as in the “middle of nowhere” as you can be. In fact, the view to the east of the Mina USHCN station is stunning for it’s remote beauty:

According to Wikipedia, Mina has quite a varied range of temperature:

Average July high temperatures range from 61° to 96 °F, with January averaging between 22° and 47 °F. The highest temperature ever recorded in Mina was 110 °F in 1933, with a low in 1990 of –23 °F. Mina receives very little rainfall, and in an average year gets about six inches, with no month getting more than one inch in a normal year. The Mina Airport is at the southeastern end of town.

The USHCN station is at the private residence of the airport operator, who also runs a KOA type trailer/RV park. The airport is a simple dirt strip, so no runway to generate extra heat. I’ve been all over the USA looking at the USHCN network. In almost every station I visit, there’s some sort of surprise. Mina was no exception, and I discovered what Stevenson Screens are really used for:

- as mounts for other weather stations.

In this case, an Oregon Scientifc WMR-968 wireless weather station, which is quite possibly the worst electronic weather station on the market. I once sold these at my online store weathershop.com, and stopped doing so when failure rates started approaching 30% out of the box.

Fortunately, the WMR-968 is not the “primary” instrument of the USHCN station, though it appears to be used as backup for the primary MMTS/Nimbus instrument. In this photo, you can see the wire from the small solar panel running inside to the temperature sensor.

What is most interesting about this station, is that while it truly is in the “middle of nowhere” and has that great “rural” view to the east, the primary MMTS sensor is just a few feet from where all the RV’s park while they register at the office:


Click for a large image

Unlike the Stevenson Screen, The MMTS is just a few feet from the office due to the famous cabling issue. It also has some nice sized rocks to act as heat sinks for those cold desert nights:


View looking North – click for a larger image

Besides the mixture of shade, rock heat sinks, road and building proximity, there’s also the requisite BBQ or two:


View looking south – click for a larger image

The rain gauge also has issues, due to the wind ducting that is likely created by these two trailers:

Click for a larger image

You can see the complete set of photos at the Mina gallery at surfacestations.org

As for the temperature trend:

Data from GISS – see original plot here

The trend is up, curiously, even though the town appears to be dying, so urbanization shouldn’t be the cause. According to NCDC’s MMS database, the station switched from using the Stevenson Screen to the MMTS electronic sensor in 1986. MMTS is well known for building proximity, That may account for some of the trend. There was also a station move to the present location in August 2007.

US 95 is about 100 yards away from the Mina USHCN temperature sensor, so perhaps there is an urbanization component in the form of more traffic.  I simply don’t know.  Interestingly, the station at Bishop, to the west, shows very little long term trend, while the station to the south, Tonopah, does show a trend. But Tonopah is growing, unlike Mina, which is dying and is now listed on a Ghost Town forum. Tonopah also had it’s weather station converted to ASOS, which when combined with other airport improvements, tends to add a positive trend.

So it’s a puzzle, and I welcome comments with ideas.

The thing about the Mina station though, is that without knowing a site history and history of the surrounding changes, we simply don’t know how much of the signal is real or from land use changes around the sensor. In it’s new position at an active RV park, it is now in a dynamic environment within feet of daily vehicular traffic. We simply should not have to figure such things out for a climatic reference station, even if it is in the “middle of nowhere”.

About these ads

26 thoughts on “How not to measure temperature part 73, in the middle of nowhere

  1. Neil,
    European countries have always known the power of propagander,in war, or otherwise. This is no different,usually, look to the money?- do the owners of the telergraph stand to make money somehow?

  2. The BBC have the same story online. From the comments on the telegraph blog, it strikes me that people are just not convinced. Wonder why…

  3. Slightly off topic as well, so apologies, but related to the audit activities.

    In the UK I noticed that there are weather stations sited on major motorways (M3 near Basingstoke and M4 near Membury Services). I’ve tried to do a bit of research to find out who actually uses the data with little success.

    According to the DfT there are 800 weather stations across the UK they are used by forecasters. Anybody in the UK know whether they are used by anyone else: Met Office Hadley etc?

    Of the two I am conscious of, one is on the hardshoulder next to the tarmac and the other has tree cover. Reading Anthony’s audit logs, I understand these may not be well sited.

  4. Taking a look at Google’s terrain map of the region encompassing all three locations, Bishop, CA seems to be in a long protected valley, while Mina & Tonopah seem more exposed. Could there be a long term change in prevailing winds for the two Nevada stations, with more downslope winds helping to warm these locations over the past 30 years? Something to think about.

  5. Just from the plot it looks like a meager upward trend in the 60s and 70s took off in the 80s when the MMTS came in. Has it been checked for a drift in accuracy? Are comparative temps still read from the min-max thermometers? Where was the station moved from? Lots of questions make this a hard puzzle.

    This kind of station begs for a network of data loggers and IR images to analyze the 3D structure of the microclimate.

  6. The “rocks as heat sinks” are probabably normal for the area, meaning there are probably lots of rocks like that scattered all over the landscape. So those are part of the normal terrain and should be ignored. All the other stuff, though… The landscape shot doesn’t show a single tree…

  7. It could be simply RV traffic patterns, given there are not a lot of places in the Nevada to park for the night, other than the desert itself. Having an idea of the sites registration records might be helpful.

  8. Anthony:

    I spent the better part of the last 20 years dealing with radio site issues for the government. Here’s how it goes: A remote site is selected, and a radio facility installed. Background noise is low, and signals are received well. Soon, the requisite beer joint appears nearby, then the convenience store, then residences for the staff of these, and so on. All as nearby as land availability permits. It takes a few years, but eventually every site worldwide which I visited has experienced the same degradation. It seems the US weather network has exactly the same problems. Must be human nature, especially the beer joints.

  9. This is off topic too but I think it may be Nutty Story #1 for “Global Cooling”. Everytime you see an article about some species struggling to survive there is somehow a tie in to AGW. This one actually sites “chilly temperatures” as the probable cause.

    Nasty fungus may be killing thousands of bats
    By Will Dunham Will Dunham – Thu Oct 30, 4:06 pm ET Reuters

    WASHINGTON (Reuters) – A previously unknown fungus that thrives in chilly temperatures may be the culprit behind the deaths of at least 100,000 bats hibernating in caves in the northeastern United States, scientists said on Thursday.

    http://news.yahoo.com/s/nm/20081030/sc_nm/us_bats_fungus_1

  10. Jeff Alberts: You’re right… So is the MMTS in the only group of trees in the area?

    And the cable issue made me go look it up… The NOAA CO-OP site has a photo of a two-conductor cable. Because the MMTS is described as a thermistor, either the MMTS display is measuring the thermistor directly (probably as a resistance through the cable), or both power and serial data are being sent through one pair of wires. However it works, “Sensor Cable Max Distance = 1/4 mile” suggests that trenching efforts are probably a significant cause of sensors being near buildings.

    Maybe the NOAA should supply trenching tips such as this collection (the E-Z Trencher link has moved from .htm to this .php URL).

  11. Jeff,
    I’m from a town 30 miles west of Mina (Hawthorne) and while there are some rocks in the valley, large ones like those situated near the MMTS aren’t prevalent unless you try to dig up something in your yard or are walking along a gully. If you look at the first picture facing east, you’ll see what the terrain is like in most any direction throughout the region. It’s sagebrush, desert grass and tumbleweeds. There are some widely scattered rocky areas, but mostly you find those in and along the mountains. In the summer, large rocks like those in the picture get extremely hot and would undoubtedly act as a heat sink.

  12. Anthony, Thanks for your work.

    Re: “The highest temperature ever recorded in Mina was 110 °F in 1933…”

    Was the peak temp in ’23?

  13. It is good to see we are still out surveying stations.

    I have 3 more to report.
    1) New Hamphsire on the Mass. border on US 93 at the “Welcome to NH”
    rest area. The station is a few feet from the largest, busiest road in
    the state. Large electronics box under the temperature station.

    2) University of Massachusetts, Lowell. Station is mounted on a pole
    on the roof at the junction of 2 roofs. Heated by 2 roofs, AC units, and
    a large wall, the temp. station is about 30′ off the ground.

    3) Brunswick Naval Air Station, Brunswick Maine. Good location in a
    field far away from buildings and pavement.

    All are visible from Google Earth; I have recent shots for 1 and 2.
    NASB will not let civilians on the flight line to take pictures of 3, so no
    photos of that one.

  14. Anthony…I’m available to go and log the weather stations in Middle TN…contact me if you’d like and I can give you more information or answer any questions or take marching orders from you… :) …believe I have the tools (want to, digital camera, pen and paper!)…

    admin at cookevilleweatherguy dot com

    Thanks,
    Michael

  15. Re: Bill P (08:56:25)

    Perhaps I’m confusing annual with monthly and daily mean. High annual mean, in the chart is (about) 1922. “Highest” (in 1933) you refer to must be a daily.

  16. On Google Earth the only vegetation is around the site, the RV park just the south and a residential area to the north.

    Those trees/shrubs will be irrigated to get them to grow and that’s probably your answer – increased humidity from irrigation and transpiration.

  17. BTW, the trend comes from increased water need and transpiration as the trees get larger. Those trees look 30 to 40 years old to me, which means they were planted around the time the upward temperature trend started.

  18. Patrick H, thanks for the correction. The closest I’ve been to that type of terrain was Ft Irwin, and driving through So Cal and Arizona a couple times (Bakersfield, Needles, Tehachapi area)

  19. You guys all keep thinking in “ones”. As if there will be only one cause of the temperature change over time… start looking at all of your bases… this is an Oregon station so it will be affected by changes in PDO (check for step changes related to PDO). It looks like it is also a reverse ENSO location (El Nino causes cooling here… look at 98)… then look at equipment, urbanization / microclimate / irrigation issues and etc.

  20. I would like to see a schematic of the MMTS system. The two wire connection is probably a 4-20ma current loop. That would take out wire resistance. Too old a design for digital data riding on top of power. Still, a thermistor is a lousy way to measure temperature. They drift with time and calibration is a pipe dream. OK for controlling temperature where absolute accuracy doesn’t matter (think fish tank heater) but not good at all for research.

    UHI effect has to depend on wind patterns. If you had a huge array of temperature sensors, you would see an elevated temp envelope around the heat collector/generator. With only a few sensors, not even a SWAG. You cannot separate random from systemic effects. Correction factors are blue sky.

    Seems like an ideal way to bet the future of the economy. (NOT)

Comments are closed.