“I believe Roger has made an important observation. The outlooks this
winter have been consistently poor and biased toward too warm and too dry
in the central U.S.

Roger is also correct, in my opinion, in noting the resistance many
climate scientists have toward verification (something we meteorologists
welcome as an important tool).

The climate models are not used for seasonal forecasting because of their
consistent lack of skill. It is often contended this does not affect
their validity for 50 year forecasts because “random” “weather” errors
cancel out. But is this really true? How is it known the errors are
random until a number of years worth of forecasts are created and
verified?

The clear dry bias of this winter’s weekly and seasonal forecasts serve as
a caution that we cannot assume “randomness.” “