Zeke Hausfather claims that my US temperature statistics are skewed by “changing station composition.” His argument is that changes over time in the mean latitude of the set of USHCN stations, affects the data. Zeke says the data has to be adjusted to compensate.
That is easy to test out, by comparing a stable set of stations with the set of all stations. The green line below is the set of all 1218 USHCN stations. The red line consists of the 695 stations which were active in both 1917 and 2016 – thus little change in latitude.
As you can see, changes in station composition have very little effect. The 1930’s were much hotter. Just one more excuse for climate alarmists to tamper with data, and make the hot 1930’s disappear.
