USGS USGS


USGS
USGS (Water Resources of the United States)
USGS National Research Program (NRP)
USGS National Research Program, Central Region (NRP, CR)
Contact Info.
Ideas
Upcoming Events
Curriculum Vitae
Previous Return to Curriculum Vitae Next

RECENT RESEARCH ACCOMPLISHMENTS

4. How can scale issues be most advantageously addressed?

Of the many issues ground-water modelers face, one of the most perplexing involves scale issues. Scale problems can be caused, for example, by hydraulic gradients that become dramatically steeper with proximity to a pumping well, rivers with widths that are small and dynamics that are fast relative to the groundwater systems with which they interact, and small geologic features that are important to simulating areas with contaminants plumes but elsewhere are less important. As suggested by these examples, scale problems occur in representing features of a wide range of sizes.

Our inability to accommodate such scale problems is a major contributor to groundwater model errors. Contributing difficulties include inadequate data and limitations of the capabilities of commonly used model. This problem has been addressed most through stochastic methods because data, computer capabilities, and numerical methods have not allowed us to build models capable of representing these variations deterministically. But these limitations are changing.

New field data methods are providing unprecedented representations of subsurface characteristics and computers are getting faster. Are ground-water models ready to take advantage of these opportunities? A promising way to address the scale issue is to reduce the errors associated with scale problems through local grid and time-step refinement. A new, accurate shared-node method has been developed by Mehl and Hill) (2002a, 2002b, 2002c).

Figure: Errors and execution time required for four local grid refinement methods: variably spaced, Telescopic Mesh Refinement (TMR) using heads or fluxes from the parent model as boundary conditions on the child model, and the new iteratively coupled method with shared nodes developed in this work. (from data presented in Mehl and Hill, 2002a, table 3, columns 4 and 5).

Previous Return to Curriculum Vitae  


mchill@usgs.gov
Last Modified: January 1, 2003