Journal of Water Resources Planning and Management

The typical approach to assess climate change impacts on water resources systems is based on a vertical integration/coupling of models. Since the range of uncertainty that can be explored with GCM is limited, researchers rely on ensembles to enlarge the spread, making the modeling approach even more demanding in terms of computation time and resource. When a particular water system must be analyzed, the question is to know whether this computationally intensive vertical approach is necessary in the first place or if we could rely on projections available in neighboring systems? The proposed study addresses this question by comparing the performance of a water resource system under future climate conditions using the vertical and the horizontal approaches. The methodology is illustrated with the hydropower system of the Gatineau River Basin in Quebec, Canada. The analysis of the simulation results reveals that the annual and weekly energy productions of the system derived from both the vertical and the horizontal approaches are similar.

Optimal design of reliable looped water distribution systems (WDSs) is challenging because these problems comprise nonlinear relationships between discharge and energy in the pipes and junctions, and nonconvex objective functions that relate cost and reliability to discrete choices of pipe diameters. Efficient algorithms for identifying optimal WDS designs also must be practical for realistic systems. Differential evolution (DE) algorithms, which harness the mutation operator to identify populations containing very different alternatives, have been shown to be more efficient than other heuristics for solving single objective versions of these problems, e.g., minimizing the cost of WDSs. By examining the entire population and evaluating those members that dominate in terms of two or more objectives, nondominated sorting algorithms efficiently identify the Pareto optimal front in multiobjective problems, and one such algorithm, the improved nondominated sorting genetic algorithm (NSGA-II), has been applied successfully to evaluate the cost-reliability tradeoff for WDSs. The nondominated sorting differential evolution (NSDE) algorithm, which takes advantage of the mutation operations in DE and nondominated sorting, and its variation NSDE with ranking-based mutation (NSDE-RMO), have been demonstrated as efficient for solving multiobjective problems. In this paper, NSDE and NSDE-RMO were applied to discrete WDS optimization for the first time, and their high performance was demonstrated compared with NSGA-II and the multialgorithm, genetically adaptive multiobjective (AMALGAM) method, a widely applied hybrid metaheuristic multiobjective algorithm. For three benchmark networks, NSDE and its variation performed similarly to or better than NSGA-II and AMALGAM except at high cost levels. For hypothetical randomly generated networks ranging from 100 to 400 nodes, and 100 to 800 pipes, the Pareto optimal front of the NSDE algorithms dominated all other algorithms, exhibiting more, and more varied, Pareto optimal solutions, and they converged sooner.

A planning framework is developed to optimize runoff control schemes at scales relevant for regional planning at an early stage. The framework employs less sophisticated modeling approaches to allow a practical application in developing regions with limited data sources and computing capability. The methodology contains three interrelated modules: (1) the geographic information system (GIS)–based hydrological module, which aims at assessing local hydrological constraints and potential for runoff control according to regional land-use descriptions; (2) the grading module, which is built upon the method of fuzzy comprehensive evaluation. It is used to establish a priority ranking system to assist the allocation of runoff control targets at the subdivision level; and (3) the genetic algorithm-based optimization module, which is included to derive Pareto-based optimal solutions for mesolevel allocation with multiple competing objectives. The optimization approach describes the trade-off between different allocation plans and simultaneously ensures that all allocation schemes satisfy the minimum requirement on runoff control. Our results highlight the importance of considering the mesolevel allocation strategy in addition to measures at macrolevels and microlevels in urban runoff management.

Although a number of adaptation strategies for coping with the ever-growing problem of inflow nonstationarity have been proposed in the reservoir operation literature, the robustness of reservoir operating rules to inflow nonstationarity has largely been ignored. This present study fills this gap. Fuzzy inference system based reservoir operating rules (FIS-ORs) are developed by optimization-simulation and compared with linear, nonlinear, and artificial neural network rules under simulated stationary and nonstationary inflow conditions. Results are obtained for two case studies assuming, in each case, a single reservoir with water supply, flood control, and environmental flow allocation functions. The FIS-ORs are found to be most robust to inflow nonstationarity. Applying the FIS-ORs to 30 years of projected future inflows show their advantage to be enhanced when they are recalibrated every 10 years or so, as compared to when there is no recalibration. It is also observed the advantage of the FIS-ORs to be more significant when the weighting of the problem objectives is such that the overall objective function is more difficult to satisfy.

Climate change may critically impair the performance of reservoirs in meeting operational objectives, but reservoirs may also aid in adapting to climate change. To understand how the reliabilities, sensitivities, and vulnerabilities of reservoir operations vary across hydrogeologic settings, a bottom-up approach was applied to investigate the reliability of two water resources systems in the future. To represent the uncertainty associated with future streamflow, global climate model projections were integrated with a formal Bayesian uncertainty analysis and groundwater–surface water hydrologic modeling. Finally, the effectiveness of variable rule curves for mitigating the effects of climate change was evaluated. Increasing air temperature appeared to reduce the reliability of meeting summer environmental flow targets in the future by 42 and 12% for the groundwater basin and surface water basin, respectively, but had negligible impacts on reservoir refilling and flood regulation. Variable rule curves mitigated the impact of climate change on summer flow target reliability without compromising flood risk reduction. Differences in subbasin sensitivity to changing climate were evident across the two hydrogeologic settings, and uncertainty associated with modeling groundwater resources and decision thresholds were identified, with implications for reliability assessments in other basins.

In the event that a contaminant is introduced to a water distribution system, utility managers must respond quickly to protect public health. Mitigation strategies specify response actions, such as warning consumers to reduce water activities using the news media and flushing contaminated water at hydrants. The performance of alternative response actions may be influenced by sociotechnical dynamics, as consumer reactions to contaminant exposure and information about the event can change water demands, hydraulics, propagation of a contaminant plume, and public health consequences. This research develops a modeling framework to test and evaluate mitigation decisions that a utility manager may take to protect public health over a wide range of contamination events. An agent-based modeling framework is developed to integrate social behaviors with technical artifacts in a sociotechnical model to evaluate the public health consequences of a water event. Social actors, including consumers and utility managers, are represented as agents and are coupled with a water distribution network model and a news media model to evaluate the performance of response strategies. Strategies for flushing hydrants are encoded as decision trees that specify the location and timing of hydrant flushing, based on the activation of water-quality sensors. The agent-based model is described using the Overview, Design, and Details protocol and is demonstrated for a virtual midsized municipality, Mesopolis. Results compare the effectiveness of flushing hydrants using cautious and adaptive response strategies and the use of the news media to disseminate warning messages. The framework can be applied for cities to evaluate alternative management strategies.

Hydraulic models that include background leakage estimates can be used to support effective decision-making aimed at reducing leakage levels and improve the understanding of the leakage hydraulics of a system. However, this opportunity is dependent on the ability to estimate the background leakage parameters. The calibration procedure is often simplified by assuming that the leakage exponent is known ($\alpha =1.2$) and that all pipes have the same leakage coefficient (parameter stratification). There is limited experience about the accuracy of background leakage calibration results. In this paper a Bayesian method for background leakage parameter estimation is developed, compared with a point calibration method, and the effects of assumptions about the leakage coefficient and the parameter stratification level have on the parameter estimation results are investigated. The results show that the assumption that $\alpha$ is known is not always appropriate, and that overstratification can reduce the accuracy of the estimates. Accurate background leakage modeling will typically require calibration of multiple leakage coefficient and exponent pairs, which will necessitate better data than what is usually available in a water distribution system (WDS) today; the paper concludes by suggesting a data collection strategy that can be applied to achieve this goal.

This article presents an innovative streamflow process model for use in reservoir operational rule design in stochastic dual dynamic programming (SDDP). Model features, which can be applied independently, are (1) a multiplicative process model for the forward phase and its linearized version for the backward phase; and (2) a nonuniform time-step length that is inversely proportional to seasonal variability. The advantages are (1) guaranteeing positive streamflow values still maintaining problem linearity, and (2) making system adaptation faster during high variable periods. Model identification is straightforward, as with the additive periodic autoregressive model generally used in SDDP. The proposed model is applied on the Senegal River system for the optimal operation of Manantali Reservoir and evaluated against the streamflow process model currently used in the water management literature.