Monday, April 20, 2009: 1:30 p.m.
Coronado I (Hilton Tucson El Conquistador Golf & Tennis Resort )
The improved speed of computers over recent years has resulted in the capability of solving larger and larger types of numerical models. Improvements in processor speeds are now less dramatic. As a result the computation process must focus on alternate means of optimizing model solutions such as parallel processing and automatic mesh refinement. Automatic mesh refinement holds the promise of optimizing calculations such that error is minimized and computation time is optimized. Mesh refinement is especially of interest in the areas of i) infiltration of water into dry soils, and ii) the movement of a contaminant front in a fate and transport analysis. Both of these scenarios will be examined in this paper and the ability of mesh refinement to improve these types of analysis will be reviewed. Improvement in the solution of difficult numerical analysis such as the classic “advection only” problem will also be examined. The implication of these computational improvements include potential gains in accuracy as well as reduced solution times. The specific application to models typical to groundwater as well as fate and transport analysis will be examined.