There is overwhelming evidence that climate change can alter the precipitation patterns, which in turn will affect the quantity and timing of freshwater availability and also the reservoir operations. Moreover, the exisiting approaches for reservoir operation correspond to historical conditions, and do not account for future climate non-stationarity. In this thesis, we explored strategies to combat the effects of climate variablility and extremes on reservoir operation and freshwater supply.

First, we examine an approach to mitigate the effects of reservoir supply uncertainty due to climate variability. We develop a joint water supply model by combining the operation of a reservoir with wastewater recycling plants and seawater desalination plants. In comparison to stand-alone...[ Read more ]

There is overwhelming evidence that climate change can alter the precipitation patterns, which in turn will affect the quantity and timing of freshwater availability and also the reservoir operations. Moreover, the exisiting approaches for reservoir operation correspond to historical conditions, and do not account for future climate non-stationarity. In this thesis, we explored strategies to combat the effects of climate variablility and extremes on reservoir operation and freshwater supply.

First, we examine an approach to mitigate the effects of reservoir supply uncertainty due to climate variability. We develop a joint water supply model by combining the operation of a reservoir with wastewater recycling plants and seawater desalination plants. In comparison to stand-alone reservoir operation, the proposed joint system is less vulnerable to climate uncertainties, and is also cost effective. The joint model is operated on a daily time step following a cost-based priority based rule to govern the water release or production from the facilities. The joint system is modeled as a multi-objective optimization problem with the double objectives of minimizing risk and vulnerability, subject to a minimum limit on resilience and water quality, capacity and budget constraints. The system is optimized using a genetic algorithm. The joint model is applied to two hypothetical but realistic cases based on India and California, considering various scenarios of costs and budget. The results show the joint operation to improve water supply security. Even though the joint operation requires an additional cost over that of stand-alone reservoir operation, this cost is significantly smaller than the case when the three sources are operated separately in a non-integrated fashion. This study contributes to a greater quantitative understanding of desalination as a redundancy measure for adapting water supply infrastructures for a future of greater scarcity and uncertainty.

Next, we examine the problem of reservoir operation under future climate uncertainty. To this end, we develop an adaptive, forward-looking reservoir operation model to account for the changing hydroclimatic conditions of future. The forward-looking optimization model with sequential decision-making ability is developed as a linear program, and solves given an ensemble forecast of future streamflows. The model executes on double time steps (daily and weekly) to simultaneously capture the effects of both the occurrence of extremes and long-term seasonality of flows. We run the model iteratively with rolling forecasts to obtain the optimal system behavior over the operating horizon.

Finally, using the forward-looking reservoir operation model developed in the above approach, we compare the performance of three strategies for adapting the operations of a theoretical Indian water supply and flood control reservoir under future climate extremes. We investigate the benefits of (i) expanding the reservoir capacity, (ii) introducing wastewater reclamation and seawater desalination to augment supplies, and (iii) improving real-time streamflow forecasts for more optimal decision-making. Our results across past and future climate scenarios suggest that of the three adaptation options, improved streamflow forecasts to be the most effective in mitigating both excess spill and water shortage, and supply augmentation by wastewater reclamation and seawater desalination to be least effective. However, we still find the roles of supply augmentation measures and reservoir capacity expansion to be significant in reducing system failure, especially in future with increases in flow extremes. This work may have significance for policy makers interested in climate change adaptation of water resources.