Open AccessModifications to the Lagrangian particle dipsersion model
Author(s) -
R. Buckley,
B.L. O`Steen
Publication year - 1997
Language(s) - English
Resource type - Reports
DOI - 10.2172/469150
Subject(s) - turbulence , code (set theory) , particle (ecology) , lagrangian , reduction (mathematics) , statistical physics , dispersion (optics) , vectorization (mathematics) , stochastic modelling , atmosphere (unit) , algorithm , computer science , mathematics , meteorology , physics , parallel computing , statistics , geometry , geology , optics , oceanography , set (abstract data type) , programming language
An advanced stochastic Lagrangian Particle Dispersion Model (LPDM) is used by the Atmospheric Technologies Group to simulate contaminant transport. The model uses time-dependent three-dimensional fields of wind and turbulence to determine the location of individual particles released into the atmosphere. Improvements to the LPDM from both operational and scientific standpoints are discussed by comparing results obtained from the original code with those from a revised version. Memory requirements are decreased by using dynamic memory allocation for many of the code`s larger arrays, while more efficient vectorization of portions of the code produces a two-fold reduction in computational time for large particle releases. In addition, the original version contained a routine for calculating random turbulent velocity components which produced anomalously large variations in particle distributions depending on the sequence in which the random numbers were selected. A revision to the technique which eliminates this sensitivity is discussed along with the improved results