thumbnail

Multiobjective sampling design for parameter estimation and model discrimination in groundwater solute transport

Water Resources Research

By:
,

Links

  • The Publications Warehouse does not have links to digital versions of this publication at this time
  • Download citation as: RIS

Abstract

Optimal design of a sampling network is a sequential process in which the next phase of sampling is designed on the basis of all available physical knowledge of the system. Three objectives are considered: model discrimination, parameter estimation, and cost minimization. For the first two objectives, physically based measures of the value of information obtained from a set of observations are specified. In model discrimination, value of information of an observation point is measured in terms of the difference in solute concentration predicted by hypothesized models of transport. Points of greatest difference in predictions can contribute the most information to the discriminatory power of a sampling design. Sensitivity of solute concentration to a change in a parameter contributes information on the relative variance of a parameter estimate. Inclusion of points in a sampling design with high sensitivities to parameters tends to reduce variance in parameter estimates. Cost minimization accounts for both the capital cost of well installation and the operating costs of collection and analysis of field samples. -from Authors

Additional Publication Details

Publication type:
Article
Publication Subtype:
Journal Article
Title:
Multiobjective sampling design for parameter estimation and model discrimination in groundwater solute transport
Series title:
Water Resources Research
Volume
25
Issue:
10
Year Published:
1989
Language:
English
Larger Work Type:
Article
Larger Work Subtype:
Journal Article
Larger Work Title:
Water Resources Research
First page:
2245
Last page:
2258
Number of Pages:
14