The uncertainty associated with multiple conceptual models of groundwater recharge and subsurface hydrostratigraphy and the impacts of this uncertainty on predictions made by a regional groundwater flow model is quantitatively evaluated. The composite prediction formally incorporates the uncertainty in the alternative input models using maximum likelihood Bayesian model averaging. The alternative models are weighted by model probability, which is the degree of belief that a model is more plausible given available prior information and site measurements. Flow predictions are found to be more sensitive to hydrostratigraphy than to recharge. Furthermore, posterior model uncertainty is dominated by inter-model variance as opposed to intra-model variance, indicating that conceptual model uncertainty has greater impact on the results than parametric uncertainty. Without consideration of conceptual model uncertainty, uncertainty in the flow predictions would be significantly underestimated.
[1]
Charles E. Russell,et al.
Reconnaissance Estimates of Recharge Based on an Elevation-dependent Chloride Mass-balance Approach
,
2002
.
[2]
Keith Beven,et al.
The future of distributed models: model calibration and uncertainty prediction.
,
1992
.
[3]
S. P. Neuman,et al.
Maximum likelihood Bayesian averaging of uncertain model predictions
,
2003
.
[4]
Ming Ye,et al.
Maximum likelihood Bayesian averaging of spatial variability models in unsaturated fractured tuff
,
2003
.
[5]
Adrian E. Raftery,et al.
Bayesian model averaging: a tutorial (with comments by M. Clyde, David Draper and E. I. George, and a rejoinder by the authors
,
1999
.