In 2007 Watts launched the "SurfaceStations.org" project, whose mission is to create a publicly available database of photographs of weather stations, along with their metadata, in response to what he described as "a massive failure of bureaucracy to perform something so simple as taking some photographs and making some measurements and notes of a few to a few dozen weather stations in each state".
On July 6, 2009 the National Oceanic and Atmospheric Administration issued a preliminary report that charted data from 70 stations that SurfaceStations.org identified as 'good' or 'best' against the rest of the dataset surveyed at that time, and concluded, "clearly there is no indication from this analysis that poor station exposure has imparted a bias in the U.S. temperature trends." Watts issued a rebuttal in which he asserted that the preliminary analysis excluded new data on quality of surface stations, and criticized the use of homogenized data from the stations, which in his view accounts for the creation of two nearly identical graphs. Since then NOAA has released a detailed peer reviewed study confirming both reliability of the surface stations reviewed. The results show that poor stations produce a slight cooling bias, in stark contrast to Watts claim, but also that after corrections both poor and highly rated stations align very well.