Search

Groundwater dependent mega-cities exert enormous pressure on the watersheds in which they are situated, creating a demand for complex hydrogeologic modeling and analysis that provides decision-makers with the information they need. However, uncertainties about the characterization of the subsurface, hydrologic fluxes, and other model input, along with assumptions made during the modeling process can create difficulties in developing the models and interpreting the results in a way that ensures sustainable and equitable regional aquifer management. This dissertation uses a multiobjective analysis approach to determine the performance of spatially distributed aquifer management alternatives for the case study of the Valley of Mexico, where the Mexico City Metropolitan Area is situated. First, a three-dimensional finite-difference groundwater model is developed alongside an initial set of managed aquifer recharge alternatives and planning objectives in Chapter 2. The management alternatives tested show spatially distinct reductions in drawdown over the historical period for equivalent changes in storage and that combining multiple alternatives results in a more than an additive risk of groundwater flooding. These results point to the importance of including planning objectives calculated over diverse spatial extents in capturing impacts to groundwater security in urban basins. To address the potential for endogenous uncertainties in models to influence decision-making, Chapter 3 carries out a global sensitivity analysis across model parameter values and subsets of well observations using the model and alternatives developed in Chapter 2. Error metrics generally used to calibrate groundwater models are found to be sensitive to distinct parameters from those to which management objectives are sensitive and the coupled effects of the endogenous uncertainties have amplifying effects on the ranking of management alternatives. Both of these findings highlight the importance of performing sensitivity analyses that are carried through to the decision-making stage and not just for calibration purposes. Finally, Chapter 4 builds on the findings from Chapters 2 and 3 first by testing improvements on the best performing alternative from the previous chapters and second by examining the limitations of regional aggregation of management objectives in providing equitable groundwater supply planning. A method is developed to alleviate scale issues that are found to arise when attempting to use existing subregional spatial units to evaluate socioeconomic indicators and groundwater management objectives simultaneously. Findings suggest that the most marginalized spatial units experience deviations from the more advantaged spatial units in performance of groundwater pumping policies, and can diverge from both the regional preference regime and the preference regime of the full set of spatial units depending on the scale of the spatial unit definition. Overall, this dissertation develops methods to represent anthropogenic impacts, assess aquifer recharge policies, and evaluate the effects of model uncertainties on decisions at the regional scale to improve the tools available needed to approach the challenges inherent in long-term urban groundwater supply planning.

Increased population and recent droughts in 1996 and 2009 for the Barton Springs segment of the Edwards Aquifer have focused attention on groundwater resources and sustainability of spring flow. These springs serve as a local iconic cultural center as well as the natural habitat for the endangered Barton Springs salamander. In response to the potential compromise of these vulnerable groundwater resources, a two-dimensional, numerical groundwater-flow model was developed for the Barton Springs / Edwards Aquifer Conservation District and other governmental entities to aid in aquifer management. The objective of this study is to develop new methods of quantifying and distributing recharge for this model. The motivation for conducting this study includes the following: recent availability of more extensive data sets, new conceptual models of the aquifer system, and the desire to incorporate estimates of urban recharge. Estimates of recharge quantities and distributions for natural and artificial sources were implemented within this model to simulate discharge at Barton Springs and water-level elevations from January, 1999 to December, 2009. Results indicate that the new methods employed generated good agreement amongst simulated and observed discharge and water-level elevations (Root mean square error of 0.5 m3 sec-1 and 10.5 m, respectively). Additionally, these recharge calculations are decoupled from Barton Springs discharge which eliminates the circular logic inherent with the previous methodology. Anthropogenic, or artificial, recharge accounts for 4% of the total recharge between January, 1999 and December, 2009. Using observed data to quantify contributions from leaky utility lines and irrigation return flows, recharge estimates were completed with spatial and temporal resolution. Analyses revealed that on a month by month basis, anthropogenic contributions can vary from

Increased population and recent droughts in 1996 and 2009 for the Barton Springs segment of the Edwards Aquifer have focused attention on groundwater resources and sustainability of spring flow. These springs serve as a local iconic cultural center as well as the natural habitat for the endangered Barton Springs salamander. In response to the potential compromise of these vulnerable groundwater resources, a two-dimensional, numerical groundwater-flow model was developed for the Barton Springs / Edwards Aquifer Conservation District and other governmental entities to aid in aquifer management. The objective of this study is to develop new methods of quantifying and distributing recharge for this model. The motivation for conducting this study includes the following: recent availability of more extensive data sets, new conceptual models of the aquifer system, and the desire to incorporate estimates of urban recharge. Estimates of recharge quantities and distributions for natural and artificial sources were implemented within this model to simulate discharge at Barton Springs and water-level elevations from January, 1999 to December, 2009. Results indicate that the new methods employed generated good agreement amongst simulated and observed discharge and water-level elevations (Root mean square error of 0.5 m3 sec-1 and 10.5 m, respectively). Additionally, these recharge calculations are decoupled from Barton Springs discharge which eliminates the circular logic inherent with the previous methodology. Anthropogenic, or artificial, recharge accounts for 4% of the total recharge between January, 1999 and December, 2009. Using observed data to quantify contributions from leaky utility lines and irrigation return flows, recharge estimates were completed with spatial and temporal resolution. Analyses revealed that on a month by month basis, anthropogenic contributions can vary from