HR: 1340h
AN: H23F-1486    [Abstracts]
TI: Performance of a new algorithm for analytic-based modeling of groundwater flow on super-regional scales
AU: * Bandilla, K
EM: bandilla@eng.buffalo.edu
AF: University at Buffalo, Department of Civil, Structural & Environmental Engineering 207 Jarvis Hall, Buffalo, NY 14260 United States
AU: Jankovic, I
EM: ijanko@eng.buffalo.edu
AF: University at Buffalo, Department of Civil, Structural & Environmental Engineering 207 Jarvis Hall, Buffalo, NY 14260 United States
AB: The Analytic Element Method (AEM) is an alternative to Finite Element and Finite Difference Methods for solving subsurface flow and transport problems on large scales. Each hydrogeologic element (e.g. surface water feature, inhomogeneity in hydraulic conductivity, well, etc.) in the domain of interest is represented by an analytic function. In order to solve the flow problem, the coefficients of the analytic function for each element must be computed so as to satisfy the boundary conditions. For the iterative approach used in this presentation, most of the computational effort is spent on computing the interactions between elements. Although this iterative approach has proven very efficient for models containing many inhomogeneities, the approach has shown to be inefficient for models containing large numbers of surface water features on large geographic scales: the iterative approach takes many iterations to converge and often fails to converge when parallel processing is used. A recently developed iterative AEM algorithm alleviates the divergence problems in parallel processing and significantly reduces the number of iterations needed for convergence both in serial and in parallel. The new algorithm uses a block Jacobi method to compute element coefficients and controls oscillation that may lead to divergence by explicitly computing the total flux of each surface water feature at the beginning of each iterative step. This enables modeling of groundwater flow and transport on super-regional scales. The performance of the new algorithm is examined using a model of the Lake Ontario watershed. The reduction of the number of iterations and the parallel processing speedups are demonstrated. The implications of computationally efficient large-scale groundwater modeling on watershed detail, multi-watershed modeling, model calibration, and reactive transport modeling are also discussed.
DE: 1805 Computational hydrology
DE: 1829 Groundwater hydrology
DE: 1832 Groundwater transport
DE: 1847 Modeling
DE: 1879 Watershed
SC: Hydrology [H]
MN: Fall Meeting 2005