How not to measure temperature, part 59

I’ve mentioned this before, but it is worth mentioning again. In the 30 years I’ve been involved in meteorology, I had no idea that water and sewage treatment plants were locations used for climate monitoring until I started the surfacestations.org project. Given the environment at these places, the idea seems simply absurd to me.

Yet thanks to volunteer Michael Caplinger, here we are again with another gem of a station that is a USHCN “high quality” climate station of record located at a water treatment plant. This is COOP ID# 461220 in Buckhannon, West Virginia:

Looking ESE – click for a larger image

In the above photo, note the placement of the NOAA MMTS temperature sensor. As we tally potential measurement biases we can see the ubiquitous air conditioner, brick masonry, concrete, a metal walkway just inches from the sensor, and of course, the big grey elephant in the room, the “tank”. Then there’s the “big valley” of wind shelter.

Summed up: variable shade, wind shelter, heatsinks, heat sources. Certainly not an easy environment to untangle the actual air temperature measurement from.

As I’ve mentioned before, water treatment plants pump massive amounts of water through, with a tank holding tens of thousands of gallons of water this close to the temperature sensor, it acts as a big heat sink. A sewage plant adds waste heat and humidity to the air, which can affect Tmins.

Here’s another view of the facility from the opposite end:

Looking WNW – click for a larger image

Let us look at the temperature record. Here is the unadjusted USHCN data for Buckhannon, as plotted by NASA GISS:

Click image for the original source plot

The trend looks flat. I did a curve fit to the station data and came up with about a .25°C trend over 100 years. With the temperature being measured so close to a nearly steady state heat sink (tens of thousands of gallons of water) is it any wonder there isn’t much trend at this location? The NCDC records for this location only go back to the mid 1940’s, but it appears temperature has been measured at this location since then.

One of the caveats listed in NOAA’s guide to station placement in addition to the 100 foot rule is that stations should “not be installed near bodies of water, unless that is representative of the area”. Good advice to avoid heat sink effects, I’d say.

Fortunately, the new Climate Reference Network, which is carefully designed and sites selected with care, will avoid such problems.

Advertisements

7 thoughts on “How not to measure temperature, part 59

  1. A sewage plant adds waste heat and

    Please complete the sentence. :)

    REPLY: ….humidity to the air, which can affect Tmins.

  2. “A sewage plant adds waste heat and…” And what? You left me hanging there man!

    That the heat sink tends to moderate the temperatures would not surprise me.

    John M Reynolds

    REPLY: ….humidity to the air, which can affect Tmins.

  3. The Climate Reference Network will still need to be watched carefully. I don’t doubt that it will start out with well chosen sites and good coverage but entropy will get you if you don’t watch out.

  4. Pingback: Dead Cant Rant » Blog Archive » False Temperature Readings

Comments are closed.