Hostname: page-component-848d4c4894-xm8r8 Total loading time: 0 Render date: 2024-07-07T15:13:43.198Z Has data issue: false hasContentIssue false
  • English
  • Français

An Analysis of Bias in Groundwater Modelling Due to the Interpretation of Site Characterization Data

Published online by Cambridge University Press:  03 September 2012

K. J. Clark ,
T. Ikeda ,
M. D. Impey ,
T. McEwen  and
M. White
K. J. Clark
Affiliation:
QuantiSci Ltd, Chiltern House, 45 Station Rd, Henley-on-Thames, Oxon, RG9 1AT, UK, [email protected]
T. Ikeda
Affiliation:
JGC Corporation, Yokohama, Japan.
M. D. Impey
Affiliation:
QuantiSci Ltd, Chiltern House, 45 Station Rd, Henley-on-Thames, Oxon, RG9 1AT, UK, [email protected]
T. McEwen
Affiliation:
QuantiSci Ltd, Melton Mowbray, Leics, UK, [email protected]
M. White
Affiliation:
QuantiSci Ltd, Melton Mowbray, Leics, UK, [email protected]
Get access

Abstract

Bias is a difference between model and reality. Bias can be introduced at any stage of the modelling process during a site characterisation or performance assessment programme. It is desirable to understand such bias so as to be able to optimally design and interpret a site characterisation programme. The objective of this study was to examine the source and effect of bias due to the assumptions modellers have to make because reality cannot be fully characterised in the prediction of ground-water fluxes. A well-defined synthetic “reality” was therefore constructed for this study. A limited subset of these data were independently interpreted and used to compute groundwater fluxes across specified boundaries in a cross section. The modelling results were compared to the “true” solutions derived using the full dataset. This study clarified and identified the large number of assumptions and judgements which have to be made when modelling a limited site characterisation dataset. It is concluded that bias is introduced at each modelling stage, and that it is not necessarily detectable by the modellers even if multiple runs with varied parameter values are undertaken.

Type
Research Article
Information
MRS Online Proceedings Library (OPL) , Volume 465: Symposium II – Scienfific Basis for Nuclear Waste Management XX , 1996 , 1125
Copyright
Copyright © Materials Research Society 1997

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

References

REFERENCES

1. Cole, C R et al. (1985). Understanding, testing and development of stochastic approaches to hydrologie flow and transport through the use of the multigrid method and synthetic datasets. In Preprints of the Symposium on the Stcobastic Approach to Subsurface Flow Greco 35 Hydrogeologie Ecole des Mines de Paris, Fontainbleau, pp 364–378. Ecole des Mines de Paris, Fontainbleau.Google Scholar
2. Desbarats, A J and Srivastava, R M (1991) Geostatistical characterisation of groundwater flow parameters in a simulated aquifer. EOS 73 (43) pp200 Google Scholar
3. Sheibe, T D and Freyberg, D L (1995). Use of sedimentological information for geometric simulation of natural porous media structure. Water Resources Research 31 (12) pp 3259–3270 Google Scholar
4. Williams, M J, Impey, M D, Sellar, C. AZURE User Guide. Quanti Scireport IM4138–3 Version 1.0 May 1996.Google Scholar
5. Nirex, , The geology and hydrogeology of the Sellafield area. Volume 3: The Hydrogeology, Nirex Report 524, 1993.Google Scholar
6. Nagra, , Sediment Study. Intermediate report 1988. Disposal options for long lived radioactive waste in Swiss sedimentary formations: Executive Summary, Nagra Technical Report 88–2 5E, 1989.Google Scholar