Strategies for large-scale, distributed hydrologic simulation

Lawrence E. Band, Eric F. Wood

Research output: Contribution to journalArticle

21 Scopus citations

Abstract

Large-scale simulation of watershed-runoff and soil-water dynamics requires the development of a modeling and parametrization strategy to explicitly incorporate the spatial structure and variability of system behavior. A dynamic, contributing-area model that concentrates on the effects of drainage-basin topography and soil hydraulic properties (TOPMODEL) is interfaced with geographic data processing and digital terrain analysis for this purpose. From digital elevation data, the methodology automatically partitions the landscape into a set of hydrologically connected subcatchment areas, then can separately parameterize TOPMODEL in each unit with information drawn from the digital elevation model (DEM) or any registered, ancillary data plane. Scale flexibility is built into the system by allowing free choice of the detail of the watershed partition, which is based on a framework for geomorphic analysis and manipulation of stream-network and drainage-basin structure developed over the past two decades. Runoff generation from each subcatchment may be simulated in parallel from the separate parameter frequency distributions within each unit. Operationally, the area over which the model can be run is limited only by the parametrization process, as the computational effort is O(n), where n is the number of subcatchments, and is not dependent on subcatchment size. The performance of the system is currently being assessed by comparison with a finite-element, hillslope hydrology model, run on topographically distinct subcatchments.

Original languageEnglish (US)
Pages (from-to)23-37
Number of pages15
JournalApplied Mathematics and Computation
Volume27
Issue number1
DOIs
StatePublished - Jul 1988

All Science Journal Classification (ASJC) codes

  • Computational Mathematics
  • Applied Mathematics

Fingerprint Dive into the research topics of 'Strategies for large-scale, distributed hydrologic simulation'. Together they form a unique fingerprint.

  • Cite this