Russ Steele
No we are not discussing real estate, we are discussing the locations of temperature sensors used in the US Historical Climatology Network. A network of 1221 temperature recording stations with long records, some from the early 1900s, which are used by NASA's Goddard Institute of Space Science (GISS) to calculate global warming in the United States.
Anthony Watts started a project to survey all 1221 USHCN stations for micro climate impacts, like co-located AC units, burn barrels, BBQs, acres of asphalt, and other anomalies not covered in the National Weather Service Standards Manual. I am one of Anthony’s early volunteers and now Ellen and I are surveying stations while on a vacation trip to Idaho and Utah. Our most interesting find so far is the Salmon Idaho Station, located at the local radio station KSRA. At least that is where it has been since 1967, I am still tracking down the other locations, since the temperature record extends from 1905 to the present.
This first picture is of the current automated sensor installed in 2000 on a flat grass covered area between the radio station and the Salmon river. But, it was placed only 3 to 5 feet from a huge fir tree which shades the sensor most of the day. In contrast the old manual station, shown in the second picture, is on graveled/blacktop slope of more than three degrees. The day I visited at 10:30 in the morning, the temperature difference between the two stations was 3 degrees. The automated sensor reading 66 F and the manual one in the shelter read 69 F. The curator noted that the temperature difference can range between 6-9 degrees on hot summer days, and 2-3 degrees in the winter. Yet, these two sensors are only 75 feet from each other.
..................
The big difference is the location, one is in the shade on damp grass, the other on a dry gravel/blacktop slope with less shade. Thus, it is clear that the sensor location has a major impact on the recorded temperature.
.
.
.
This temperature plot from 1905 to 2005 does not show a warming trend,
with 1934 the clear winner for the warmest year. I will be looking at
the monthly records to seen if I can spot a shift when the sensor was
automated. And, will spend some time looking for the other locations
while we are staying in Salmon. An update is on the horizon. More
pictures at http://www.surfacestations.org/
UPDATE: To confirm the temperature differences between the MMTS and the thermometers in the Stevenson shelter, we revisited the Radio station today and took the following photos, that show a four + difference.
Click or a larger image to read the temperatures 
This is the automated systems temperature, which is about 4 degrees lower, only 75 feet away. Which should be the reported temperature?
