Development of a Parallel Computing Watershed Model for Flood Forecasts

Ping Cheng Liu, Dong-Sin Shih*, Chau Yi Chou, Cheng Hsin Chen, Yu Chi Wang

*Corresponding author for this work

Research output: Contribution to journalConference articlepeer-review

1 Scopus citations

Abstract

In this study, using the WASH123D (WAterSHed Systems of 1-D Stream-River Network, 2-D Overland Regime, and 3-D Subsurface Media) numerical model that was developed by the University of Florida professor Ye, its feature is the ability to combine rivers, surface and groundwater with simulation, and it can be used in variety flow. Currently WASH123D has been extended in various research projects, since WASH123D can calculate different kind of cases, the model that is necessary to set a large number of simulation parameters, so it cause a long time to compute. In the research, WASH123D model as the basis for the development of HERO (HypErcomputing wateRshed mOdel) model, HERO model decrease memory usage by reducing the computation of matrix, and add parallelizing calculations to make the subroutine that computation is huge calculate in different core. Then more cores will decrease the computation time in CPU, and then it add the infiltration equation of Green-Ampt Method in the model, reducing the subsequent calculation of the output value. The results, WASH123D mode need to modify the grid for different cases, but HERO mode can accept various cases, and the highest memory usage can be reduced 120bytes, then the method of parallel, OpenMP, effectively reduces the computation time nearly 50%.

Original languageEnglish
Pages (from-to)1043-1049
Number of pages7
JournalProcedia Engineering
Volume154
DOIs
StatePublished - 1 Jan 2016
Event12th International Conference on Hydroinformatics - Smart Water for the Future, HIC 2016 - Incheon, Korea, Republic of
Duration: 21 Aug 201626 Aug 2016

Keywords

  • OpenMP
  • WASH123D

Fingerprint Dive into the research topics of 'Development of a Parallel Computing Watershed Model for Flood Forecasts'. Together they form a unique fingerprint.

Cite this