SGER: New Enabling Technology for the Study of Groundwater Flow and Transport Under Uncertainty

University Of Vermont & State Agricultural College, Burlington VT

Investigators

Abstract

0130652 Pinder The objective of this research is to investigate the use of a new approach to the modeling of groundwater flow and transport systems. A Monte-Carlo simulation approach is the only viable method for the solution of such problems involving large uncertainties in parameter values, such as hydraulic conductivity; but current methods require enormous computational effort and they require the use of rectangular numerical meshes. The goal of this research is to explore a new approach to generating random fields for both parameters and state variables that results in significantly less computational effort when simulating groundwater flow and transport and that can also be used on a triangular (or other irregular) finite-element mesh. The approach uses a variant on Latin-hypercube sampling to perform the necessary Monte-Carlo simulations. The overall goal is to increase both computational speed and mathematical flexibility when solving such problems described using parameters characterized by random fields. The computational efficiency will be achieved by reducing the number of realizations required to represent a random field to a specified degree of accuracy. The mathematical modeling flexibility will be enhanced by extending current random field representation technology to accommodate irregular finite-element simulation meshes, a technology that does not currently exist. This project could lead to significant gains in efficiency and effectiveness in solving difficult problems of groundwater contamination and flow, as well as a wide range of other problems in engineering and physics.