Groundwater decision support: linking causal narratives, numerical models, and combinatorial search techniques to determine available yield for an aquifer system

Access full-text files




Pierce, Suzanne Alise

Journal Title

Journal ISSN

Volume Title



Meeting future water demands requires sustainable management of aquifer resources. Over-abstraction of aquifer systems may occur if inequitable and inefficient methods, which do not consider science or people, are used to determine groundwater allocation and management strategies. The complexity of water resource issues is caused by the interconnectedness of hydrologic systems with other biophysical and social systems, such that competing requirements and non-linear behavior often overwhelm decision-makers and inhibit the creation of clear management strategies. A groundwater decision support system (GWDSS) is constructed to address the complexities associated with determining an acceptable groundwater allocation policy. Stakeholder concerns were gathered via elicitive interviews for inclusion in the GWDSS to link stakeholder input with a groundwater simulation-optimization model. GWDSS is a software tool that supports the implementation of a multi-disciplinary methodology consistent with the nature of water resource allocation problems for calculating the available yield of an aquifer A tightly coupled simulation-optimization system, GWDSS is based on a preexisting two-dimensional groundwater flow model for the Barton Springs segment of the Edwards aquifer. This test case is a karst aquifer undergoing rapid urbanization in central Texas. Active stakeholder groups in the region have vested interests in management strategies for the resource, and provided social inputs that serve as soft constraints for a tabu search optimization algorithm. Tabu is a metaheuristic algorithm that uses querybased search techniques to gather knowledge about an objective and identify a list of deterministic solutions for consideration by a decision-maker. The tabu algorithm is applied to the transient version of the Barton Springs groundwater model to rank management scenarios ordinally. This method determines a set of feasible solutions, within the bounds of physical system behavior and community defined constraints, for extraction policies. The GWDSS addresses the issue of aquifer sustainability by merging existing, spatially-variant, numerical groundwater models with stakeholder narratives into a relational database structure that is compatible with metaheuristic optimization searches. This approach allows for more generalized management decisions to be mediated by a vetted scientific groundwater model and permits the determination of an available yield that incorporates both social considerations and aquifer performance in an integrated manner.