Are historical records sufficient to constrain ENSO simulations?

Are historical records sufficient to constrain ENSO simulations?