Catchment sediment budgets are increasingly being used to guide and justify substantial investments in remediation works aimed at reducing deleterious effects of sediment export to receiving waters, with recent efforts to prioritize work and assess improvements in the catchments draining to the Great Barrier Reef using D-Sednet and Source Catchments providing well known examples. Catchment sediment budgets are often built from local-scale site based models of erosion and deposition. For such models, no matter how sophisticated, to have any practical use in river management it is crucial that they can be scaled up to the catchment scale, from the site scale. An essential requirement of any such scaling up process is the determination of appropriate channel geometry metrics across the catchment. In this paper we argue for an iterative stratification of the catchment into zones having consistent relationships between channel geometry and some universally measurable attribute. We also show how collecting high resolution data at nested scales throughout the catchment provides the basis for both describing site scale processes and modelling catchment scale distributions of channel forms, from which process can be inferred. We introduce a workflow that describes the steps and data requirements necessary for characterizing the distribution of channel metrics in large catchments such as the Normanby