Parameter Uncertainty for Capture Zone Delineation in a Complex Hydrogeologic Environment

Monday, May 5, 2014: 4:20 p.m.
Confluence C (Westin Denver Downtown)
Martinus Brouwers, MASc , Matrix Solutions, Breslau, ON, Canada
Paul Martin , Matrix Solutions Inc., Breslau, ON, Canada

Groundwater flow models are useful tools for the management of groundwater resources. While a vast amount of data can be collected and implemented in the development of a numeric model, it remains only an approximation of the real-world system. As such, data gaps and uncertainties exist in the parameterization of a model. These uncertainties carry forward when using the model to make predictions such as capture zones. 

Capture zones are typically delineated by applying backward (advective) particle tracking to a calibrated model and projecting the encompassed footprint to ground surface. It is recognized that considerable uncertainty exists in the size and shape of these capture zones. To compensate, professional judgment is typically applied to provide a degree of conservatism. This is done to account for features that cannot be explicitly represented in a numeric model such as the degree of heterogeneity and variability of the natural system.

In an effort to address this uncertainty, a backward-in-time advective-dispersive transport approach has been applied as an alternative approach to delineating capture zones. This technique has the advantage of representing local scale heterogeneity through the inclusion of the dispersion term. The resulting capture zones can be delineated within the context of a probabilistic framework.

This paper presents an alternative approach to developing capture zones that encapsulates the uncertainty of the hydraulic conductivity distribution. One hundred statistically equivalent models are generated by randomly sampling log-normal hydraulic conductivity distributions that are centered about model calibrated values. Each model explicitly simulates different flow paths, thereby sampling the variability of the flow system. A composite capture zone is created by combining all these capture zones. The composite capture zone can be likened to a spatial probability distribution of groundwater contribution to a given well. A comparison of the capture zones created using these two approaches is presented.

Martinus Brouwers, MASc, Matrix Solutions, Breslau, ON, Canada
Martinus Brouwers is a practicing groundwater modeling specialist with seven years of experience in the industry. He graduated from the University of Waterloo with a MASc in Civil Engineering. Currently his interests include watershed scale modeling with a focus on groundwater resource management and source water protection initiatives, parameter and prediction uncertainty analysis, as well as data analysis using database and GIS tools.

Paul Martin, Matrix Solutions Inc., Breslau, ON, Canada
Paul Martin is a Principal Hydrogeologist and Professional Engineer with Matrix Solutions. He holds a bachelor’s degree in Civil Engineering and a master’s degree in Earth Science from the University of Waterloo. Over the last 20-years Martin has worked on consulting projects ranging from water resources characterization to contaminant impact evaluation and remediation.