The project also will put its calculations on the Internet in a "transparent" way, Muller said. Other scientists, he said, "put homogenized data online. They don't put up the [software] tools that get you from the raw data to the homogenized data. How do they pick the [weather station] sites? That involves human judgment."
Peter Thorne, a leading expert on temperature data at the National Oceanic and Atmospheric Administration's National Climatic Data Center in Asheville, N.C., said the three main data sets by the NOAA, NASA and Britain's Hadley Centre adjust for the heat island effect, as well as for measurements at different times of day. Muller's use of 39,000 weather stations, he said, "will make next to no difference" in the final result.
Thorne said the data and computer code for the NASA analysis have been publicly online for five years, while NOAA's data and code have been online for three years. Most of the British center's data are online, except for information shared on a confidential basis by commercial groups, and the code is available, he said.
Thorne said he was unsurprised by the Berkeley project's focus on temperature data. "For those who wish to discredit the science, this record is the holy grail," he said. "They figure if they can discredit this, then society would have significant doubts about all of climate science."