How not to measure temperature, part 41

Lately we’ve been touring COOP weather stations in California to illustrate how they are prone to biases in their siting. This problem is documented in the recent paper LaDochy, S., R. Medina, and W. Patzert. Recent California climate variability: spatial and temporal patterns in temperature trends which is available for review in my post here.

The picture below is from the COOP station #04-5915  at Mount Diablo Junction, CA,  which is an “A” network station reporting climate to NCDC. Like many stations I’ve reviewed recently, this one is directly next to parking spot and a building. I wonder how many hot vehicle radiators this station has responded to? It has been in operation at the same location since 03/10/1952

The picture was taken by Canadian Ronald Franzmann while he was on vacation recently. he was kind enough to send it to me.


Here is another view of the same station courtesy of the San Francisco/Monterey National Weather Service Office. The view shows it to be clearly rural in nature, at an elevation of 2170 feet, but I think the local microsite biases are as great as what is seen in some city weather stations.

While I realize flat space on a  mountain comes at a premium, you’d think they could do a little better with placement of this station. Vehicles can park radiator inward within a few feet of the shelter, and the asphalt and building proximity certainly break the NOAA 100 foot siting rule. Since the station is just a few steps from the rear door, again it appears the placement was for the convenience of the observer, and not for the quality of the measurement.

For those interested, here is a PDF of the 30 year climate norms from this station.


4 thoughts on “How not to measure temperature, part 41

  1. If I recall correctly, there is the typical gravel / volcanic cinder laid down on the slope beyond the edge of the parking area in order to prevent erosion. If so, then it’s getting cooked from both sides.

  2. OT – looks like CA is down. Since this is not, and, I believe, on the same server, I am concerned about DDOS attack or other ne’er do wellism.

    REPLY: Yes not on the same server, but fixed nonetheless. It’s some PHP versus browser type weirdness, not DDOS attacks.

  3. On the subject of contaminated measurement data…

    Looks like Ross McKitrick has been busy:

    Seems to strongly suggest that the GISS adjustment process fails to correctly adjust the temperature data. Probably not controversial here, but likely to raise some blood pressure readings at RC… :)

Comments are closed.