Computational challenges in large-scale air pollution modelling

Authors:
Tzvetan Ostromsky;Wojciech Owczarz;Zahari Zlatev
Affiliations:
Central Laboratory for Parallel, Processing, Bulgarian Academy of Sciences, Acad. G. Bonchev Str. 25 A, Sofia, Bulgaria;Danish Computing Centre for Research and Education, Technical University of Denmark - Build. 404, DK-2800 Lyngby, Denmark;National Environmental, Research Institute, Frederiksborgvej 399, P. O. Box 358, DK-4000 Roskilde, Denmark
Venue:
ICS '01 Proceedings of the 15th international conference on Supercomputing
Year:
2001

Citing 14
Cited 6

Direct methods for sparse matrices

Direct methods for sparse matrices
One-step and extrapolation methods for differential- algebraic systems

Numerische Mathematik
Numerical methods for ordinary differential systems: the initial value problem

Numerical methods for ordinary differential systems: the initial value problem
An evaluation of explicit pseudo-steady-state approximation schemes for stiff ODE systems from chemical kinetics

Journal of Computational Physics
Using MPI: portable parallel programming with the message-passing interface

Using MPI: portable parallel programming with the message-passing interface
PVM: Parallel virtual machine: a users' guide and tutorial for networked parallel computing

PVM: Parallel virtual machine: a users' guide and tutorial for networked parallel computing
Explicit methods for stiff ODEs from atmospheric chemistry

NUMDIFF-7 Selected papers of the seventh conference on Numerical treatment of differential equations
Efficient implementation of fully implicit methods for atmospheric chemical kinetics

Journal of Computational Physics
Improved Quasi-Steady-State-Approximation Methods for Atmospheric Chemistry Integration

SIAM Journal on Scientific Computing
Large Scale Computations in Air Pollution Modelling

Large Scale Computations in Air Pollution Modelling
Modeling the Long-Range Transport of Air Pollutants

IEEE Computational Science & Engineering
Exploiting the Natural Partitioning in the Numerical Solution of ODE Systems Arising in Atmospheric Chemistry

WNAA '96 Proceedings of the First International Workshop on Numerical Analysis and Its Applications
Running Large-Scale Air Pollution Models on Message Passing Machines

Proceedings of the 4th European PVM/MPI Users' Group Meeting on Recent Advances in Parallel Virtual Machine and Message Passing Interface
Running an Advection-Chemistry Code on Message Passing Computers

Proceedings of the 5th European PVM/MPI Users' Group Meeting on Recent Advances in Parallel Virtual Machine and Message Passing Interface

Flexible Two-Level Parallel Implementations of a Large Air Pollution Model

NMA '02 Revised Papers from the 5th International Conference on Numerical Methods and Applications
Parallel Performance and Scalability Experiments with the Danish Eulerian Model on the EPCC Supercomputers

Numerical Analysis and Its Applications
Parallel and GRID implementation of a large scale air pollution model

NMA'06 Proceedings of the 6th international conference on Numerical methods and applications
Parallel implementation and one year experiments with the danish eulerian model

NAA'04 Proceedings of the Third international conference on Numerical Analysis and its Applications
Runs of UNI–DEM model on IBM blue gene/p computer and analysis of the model performance

LSSC'09 Proceedings of the 7th international conference on Large-Scale Scientific Computing
New parallel implementation of an air pollution computer model --- performance study on an IBM blue gene/p computer

LSSC'11 Proceedings of the 8th international conference on Large-Scale Scientific Computing

Quantified Score

Hi-index	0.00

Visualization

Abstract

Many difficulties must be overcome when large-scale air pollution models are treated numerically, because the physical and chemical processes in the atmosphere are very fast. This is why it is necessaryto use a large space domain in order to be able to study long-range transport, of pollutants in the atmosphere,to describe adequately all important processes when an air pollution model is developed andto use fine grids in the discretization of the model.If all these conditions are taken in to account, then the resulting computational tasks are huge and the computer treatment of the model is very difficult even when powerful modern computers are in use. Therefore, the following two major tasks must be solved:fast and sufficiently accurate numerical methods are to be selected andthe selected numerical methods should be efficiently implemented on the available high-speed computers.The solution of these two important tasks in connection with a particular large-scale air pollution model, the Danish Eulerian Model, will be discussed in this paper. The use of standard parallelization tools allows us to achieve good performance when the model is run on different parallel computers. This will be illustrated in the paper by many numerical examples.