Don't overlook that this data is compiled by Hansen. Given the standards of Hadley CRU why would you think Hansen is any different?