A critique of the Freeze and Harlan blueprint for a distributed physically based hydrological model leads to the conclusion that it will be abandoned. An alternative blueprint as a modelling methodology is proposed that explicitly recognises the potential for equifinality in scale-dependent model representations. An inductive rather than deductive definition of physically-based is proposed that reflects the important role of observables in constraining the feasible models.
This is a pre-print of an article published in Hydrological Processes, 16 (2), 2002. (c) Wiley.