Why would one use atmos data from reanalysis (ex ERA-Interim) but use SSTs from an alternative dataset (ex HadISST) instead of from reanalysis itself?

This would seem to be physically inconsistent, yet I see it often, and I am trying to understand why (e.g. is there a reason that the reanalysis SSTs are not believable?). Thanks! Dan

10 September 2015 9,174 2 View