How not to measure temperature, part 8

The picture below is from Radio station KQEN in Roseburg, OR. It is the official climate observing site for Roseburg. This picture is courtesy of Oregon State University Climatologist, George Taylor, whom is collaborating with me on my project


Note that Mr. Taylor is not responsible for the thermometer placement, the National Weather Service in Medford is in charge of the observer program. and sets up the equipment. Only a couple of problems here, like the roof itself being hotter than any surrounding area on any given day, but lets just throw in a nearby rooftop air conditioning unit for good measure. A/C units exhaust quite a bit of hot air.

Here’s the temperature plot from NASA’s GISS database, which is used in global climate modeling and climate forecasting. Yes, this data is part of the USHCN “high quality” dataset used to make climate change predictions.


I’m not making this up, check it out for yourself at NASA’s GISTEMP database


10 thoughts on “How not to measure temperature, part 8

  1. It makes me want to scream. Thank goodness for the even tempered, reasoned, and analytical effort Anthony. If reasonable, capable, and dedicated, folks like you keep this up you will raise the bar in climate science.

  2. Reminds me of the old George Carlin joke about why do the weathermen give the temperature at the airport, when nobody lives there.

    At my house, which has the normal amount of concrete, gravel, and landscaping, I usually get a reading of about 5 degrees higher in summer than the official city temperature.

    So what is the definition of the city temperature? Is it the temp out on some grassy field? or the urban temp where most of us live?

  3. Measuring tempos ON A ROOF????

    Geeze. Down here in the south a roof can get 140 degrees.


  4. first of all sensors read mean temperature as stated. gain and loss are measured on a time line basis with the baseline simply being set by the unit itself, after calibration. with that said the only data retrieved from these units is mean fluctuations over a given time frame( hi low) a/c or no a/c these fluctuations are not going to change the overall measurement. When unit is calibrated at install all existing circumstances are “existing”.

  5. karl,

    That is a silly argument. You were correct about measurement of gain or loss, but not so much about the a/c having no effect. These sensors measure night and day, and throughout seasons. If you consider the diurnal and seasonal changes in a/c use, as a direct response to temperature changes, you’ll realize that the hotter things actually are up there on the roof, the more the a/c vent output will contribute to the heat.

    the real question is how relevant is the heat energy contribution to the temperature measurement, one tenth of a percent, perhaps? In fact, on a very hot day, is it possible that simple movement of air has a slight dampening effect on escalating temps at the sensor? I don’t know, but I would guess that a meteorologist with direct experience in installing, calibrating, and maintaining banks of sensors would have some idea. I’m certain it doesn’t account for the last decade of increase, but then again neither does CO2

  6. ♠karl – The roof warming at a greater rate than the surroundings, the a/c turning on and off as the case may be, these are what you’re referring to as “fluctuations”? The temperature change on a hot roof will be greater throughout the day than a less exposed location, and one with less albedo effect. If the background absorbed more heat than it reflected, fluctuations would be less. This is why ocean and lake temperatures are more stable throughout the year than ground temperatures. Following your reasoning, it’d be a as valid to put the sensors in the shade on the north side of something obstructing the sun, ignoring the fact that the range of temperatures is lower (less fluctuation) than a site directly exposed to the sun.

    “Global warming” is a hoax, pseudoscience at best. Actual data indicate the earth has been cooling for a decade, and this in past year the greatest annual drop in global mean temperatures even was recorded.

  7. As someone who at one point made a living doing meteorological research and has also come across such poor installations I can say it’s probably the radio station responsible here, often a data site is given a set of guidelines on how to install the sensors and asked to comply. Frequently the local NWS will ask ‘Are you compliant?’ and they’ll accept whatever answer they get. Proper installation takes a lot more than some places are willing to provide.
    This installation also violates the electrical code, note how the temperature sensor is attached to the power service riser.

  8. Pingback: US climate change plan 'disastrous and Neanderthal' - The Political Hotwire

  9. Pingback: How not to measure temperatures… « Check Your Premises

Comments are closed.