Keywords
cluster computing, parallel simulation, python, pypar, openmpi, openmp, climate change, impact assessment
Start Date
1-7-2012 12:00 AM
Abstract
The paper introduces different approaches to parallelization based on Open- MPI and OpenMP applied to the Universal Soil Loss Equation (USLE). The USLE was used as a proxy for similar models from the “impact assessment toolbox”. The simulation of impact assessment takes into account climate change and changes in management. Even such a simple model as the USLE can lead to a time-consuming simulation when applied to a large region and when including stochastic data. The paper discusses the pros and cons of the implemented parallelization techniques. The key technology is to divide the simulation into two parts: a binary part implemented in C++ and an interpreter part which controls the parallel simulation written in Python. Python and its modules were also used to pre- and post-process the simulation.
Parallel implementation to support large spatial simulations
The paper introduces different approaches to parallelization based on Open- MPI and OpenMP applied to the Universal Soil Loss Equation (USLE). The USLE was used as a proxy for similar models from the “impact assessment toolbox”. The simulation of impact assessment takes into account climate change and changes in management. Even such a simple model as the USLE can lead to a time-consuming simulation when applied to a large region and when including stochastic data. The paper discusses the pros and cons of the implemented parallelization techniques. The key technology is to divide the simulation into two parts: a binary part implemented in C++ and an interpreter part which controls the parallel simulation written in Python. Python and its modules were also used to pre- and post-process the simulation.
