You can correlate the measurements across distance and use statistical analysis to work out if a reading is anomalous.
...says RickNRoll
And there is your problem. Climate is not homogenous because weather is not homogenous. You cannot reduce all local weather records to a single point analysis. The weather experienced in Harrietville is ALWAYS totally different to Rutherglen even though they are both within 50nm of each other. You cannot take the temp record of Rutherglen and weight it's data using data from Bairnsdale in Gippsland. The weather experienced in Rutherglen is not the weather experienced in Cobar. Why adjust temperature records...hard data...to fit a hypothesis? Who decides what record is anomalous? ..and why it is anomalous?