But to elaborate, as you know, they use downtown Allentown data from 1922 to 1944 (maybe 1938?) then airport data, between which there is really not much difference considering the warmer current normals: see:
Geology of Lehigh and Northampton Counties, PA - Historical Works
a few years ago I saw them list pre-1922 temp data in a list of record warm months and they told me that it was from Lehigh University (also in the above link), but their means are much warmer than Allentown which is probably why they no longer use them in the combined dataset