How to divide a catchment to conquer its parallel processing. An efficient algorithm for the partitioning of water catchments

Authors:
M. Grübsch;O. David
Affiliations:
Friedrich-Schiller-Universität Jena, Institute for Geographic Sciences Department of Geoinformatics, Hydrology and Modelling Löbdergraben 32, 07743 Jena, Germany;Friedrich-Schiller-Universität Jena, Institute for Geographic Sciences Department of Geoinformatics, Hydrology and Modelling Löbdergraben 32, 07743 Jena, Germany
Venue:
Mathematical and Computer Modelling: An International Journal
Year:
2001

Citing 2
Cited 1

An improved spectral graph partitioning algorithm for mapping parallel computations

SIAM Journal on Scientific Computing
Effective graph clustering for path queries in digital map databases

CIKM '96 Proceedings of the fifth international conference on Information and knowledge management

Maximum speedup ratio curve (MSC) in parallel computing of the binary-tree-based drainage network

Computers & Geosciences

Quantified Score

Hi-index	0.98

Visualization

Abstract

For the sake of distributed modelling, water catchments are subdivided into a set of subareas whose number is theoretically unbounded. To cope with such a computational challenge-100 to 100,000 subareas are usual-the subareas which are to be calculated separately could be distributed among several networked (multiprocessors) computers. By the means of graph theory and the analysis of 'typical' routing systems, the present paper develops an heuristic algorithm to partition the set of subareas into subsets, each of them assigned to a particular processor such that the computational load of the processors is balanced, and the need for interprocess-communication is minimal.