How not to measure temperature, part 17

The picture below comes to me via the National Weather Service Forecast Office in Monterey from their website. You can see the unannotated version here:
http://www.wrh.noaa.gov/images/mtr/cpm/1967.jpg

It is the Cooperative Observer Station ID# 041967 for Concord, California in Contra Costa County. It, like many temperature monitoring stations nationwide, is located at the Waste Water Treatment Plant there. The MMTS display console is inside the guard hut where the daily high/low is recorded and then sent of on a B44 form once a month to the National Climatic Data Center in Asheville, NC

Fortunately, this station is not part of the USHCN climate data set, as it appears it may be compromised by micro-site effects. Note the air conditioning unit just about 6 feet away, and the nearby parking lot. Then there is the varying shade from the tree.

One could argue that all these effects could cancel each other out: Parking lot- positive, A/C unit exhaust - positive, Shade - negative and I’m sure that can be true at certain times, but at other times, they may not be. Since you can’t for example calculate when the a/c turns out and compare that to when the tree’s shade may not be falling on the sensor, or know where all that is in context to a partly cloudy day when sun beams through, heating up the parking lot and surrounding air.

All of these cyclic factors can combine together, like a rogue wave in the ocean, to make a “temperature peak” which would be higher than if these factors didn’t exist. Similarly, a trough or minimum can also be created at other times. The point is, without having monitored these things through time, its impossible to tell when or to know the magnitude be it positive or negative.

In the next couple of days, I’ll start showing some well sited stations for comparison.