I recently started a PC entry to highlight a new study evaluating the quality of the land surface data used by the IPCC in determining average global temperatures. http://priuschat.com/forums/showthread.php?t=40422 Related to this, I thought it might be interesting to share the work of Anthony Watts at surfacestations.org. He is currently auditing the entire U.S. surface station dataset by looking at each collection station and classifying its degree of temperature bias. So far, he and project volunteers have evaluated 460 out of 1221 stations. Here is what they have found: Class 1 & 2 Stations (bias of <1 degree C): 13% of all stations Class 3 (bias of >=1C): 18% of all stations Class 4 (bias of >=2C): 55% of all stations Class 5 (bias of >=5C): 14% of all stations See the presentation here for more info (note the presentation is a bit dated and reflects only 331 audited stations - the current count is 460). http://gallery.surfacestations.org/UCAR-slides/index.html#thumb92.jpg For a related map, see here: http://icecap.us/images/uploads/USHCNMAP120807.jpg
