Standard Article

134 Downward Approach to Hydrological Model Development

Part 11. Rainfall-Runoff Modeling

  1. Murugesu Sivapalan1,
  2. Peter C Young2,3

Published Online: 15 APR 2006

DOI: 10.1002/0470848944.hsa141

Encyclopedia of Hydrological Sciences

Encyclopedia of Hydrological Sciences

How to Cite

Sivapalan, M. and Young, P. C. 2006. Downward Approach to Hydrological Model Development. Encyclopedia of Hydrological Sciences. 11:134.

Author Information

  1. 1

    The University of Western Australia, Centre for Water Research, Crawley, Australia

  2. 2

    Lancaster University, Centre for Research on Environmental Systems and Statistics, Lancaster, UK

  3. 3

    Australian National University, Centre for Resource and Environmental Studies, Canberra, Australia

Publication History

  1. Published Online: 15 APR 2006


This article presents the top-down or downward approach to hydrological model development as an alternative to the bottom-up, reductionist approach, which is the current, dominant paradigm in the hydrological sciences. It discusses the philosophical underpinnings as well as the pros and cons of each approach, and illustrates the application of the downward approach through several examples. These examples variously emphasize three key elements in the application of the downward approach. First, the analysis of identifiable “signatures” or “features” in the available data including the definition of a generic model form that can accommodate these features; second, model development including model structure identification and parameter estimation; and last, model refinement or “fingering down” (Klemes, 1983; Jarvis, 1993), in which the model is improved by the use of additional information or by adding more causal factors. Two main approaches to downward modeling are identified and discussed: the classical stochastic systems approach and the deterministic, conceptual modeling approach. The advantages of the downward approach are highlighted, particularly as they relate to the development of models that have sufficient causal basis to make predictions in basins other than those on which the model was developed.


  • aggregated dead zone (ADZ) models;
  • data-based mechanistic (DBM) modelling;
  • dominant mode analysis;
  • downward approach;
  • model development;
  • model identification;
  • model refinement;
  • model simplification;
  • parameter estimation;
  • rainfall-runoff models;
  • signature analysis;
  • transfer function models