[cogitator]

Outlier is a term that should be reserved for obviously wrong data

That's fine, but raw data can have outliers. Something has to be done about them.

I rechecked St. Louis for February, 2008 and it was warmer than 2007 by 1C, but not 3C.

But St. Louis doesn't account for the whole dot.

[Old Professer]

“That’s fine, but raw data can have outliers. Something has to be done about them.”

You finally agree with something I say.

Now, let’s get the instrument shelters back in the middle of open, well ventilated areas, maintain them and calibrate the instruments inside and add more of them instead of letting them die and fall off the grid.