Technical/ models query

[Big Bear]

Does anyone think it is possible, or can anyone actually help produce a hindcast model, where data from previous years, leading up to classic cold spells can be fed into current systems to produce a 'mock' ensembles.

It would be facinating to see how the ensembles would behave with forecasted GFS data (for example) for the early part of Jan 87, 1963 or 1947.

The aim would be to test how accurate the current models are by testing old data as new input, then analysing how far out a classic cold spell is potentially spotted.

Thoughts & discussion please.

[Paul]

There is hindast data available but not that far back unfortuantely. The simple reason for this is that the models injest a huge amount of data which isn't available for earlier dates, so any comparison wouldn't mean a thing unfortunately.