Climate cuttings

Steve McIntyre notices some surprising features of the temperature record. Examining the figures for a couple of locations (gridcells) in California, the temperatures as calculated by NASA-GISS and UEA Hadley are in good agreement for most of the last century, but..

• they diverge by up to 0.5^oC after 2000 - a time when the measurements should be improving
• the high temperatures recorded in the 19th century have been excised from recent versions of the Hadley figures, increasing the apparent warming trend
  

How, he asks, can these scientists claim to know the temperature a millenium ago to an accuracy of a couple of tenths of a degree when they have errors of half a degree in 2005?

Anthony Watts has got another truly hilarious example of poor weather station siting - this one's at a sewage treatment plant with the temperature sensor surrounded by brick walls and windows, and in close proximity to the sewage treatment tanks (which give off heat) and also to an air-conditioning unit exhaust.