Hydrological systems present several difficulties for modellers. A data-based approach utilizing objective statistical analysis of output and/or input-output time series behaviour has much to offer, both for understanding of the response dynamics of systems and future hypothesis formulation for improved model structure, as well as good predictive capacity. In this paper, the essential elements and tools of a data-based approach, using an Unobserved Components (UC) model, are presented and illustrated with two examples.

You do not currently have access to this content.