| Many municipalities within the United States anticipate rising demand for water as populations grow. Traditionally, rising demand has often been addressed via infrastructure projects, such as reservoirs. However, a variety of factors has combined to make such projects less attractive, such as increased development costs, stricter environmental regulation, and greater public opposition.;By contrast, transfers of water from existing sources can be used to more efficiently manage risk posed by rising demand, allowing water to be acquired on more of an as-needed basis. When developing transfer agreements, however, questions of timing, quantity, and type of transfers must be settled if transfers are to be effectively employed. Regional differences in water law, the nature of the available resources and the degree of hydrologic variability further determine how transfers might be applied.;This research contributes to knowledge in three specific areas: (i) This work examines the manner in which different types of market-based transfers can be combined with firm capacity to form minimum expected cost "portfolios" of different transfer types (e.g., permanent rights, leases, options) that meet defined reliability and/or cost variability constraints. In doing so, a Monte Carlo simulation is paired with the "implicit filtering" optimization routine, designed to optimize portfolios despite the sampling error, or "noise", inherent in searching for an optimal expected value. (ii) The second phase of research applies a modified technique (control variate) to reduce the level of noise inherent in the simulation, thereby improving the efficiency and accuracy of the optimization approach. This method is applied to the study region as the simulation is expanded from a one-year to a 10-year model, and results in a significant reduction in computational burden (as much as 50%). (iii) A technique is developed to generate synthetic streamflow time series in a manner that reproduces autocorrelation in the historic record. This method is used to develop streamflow records representative of future climate scenarios, which are then used as inputs for a model that assesses different risk-based transfer agreements within the Research Triangle region of North Carolina. Results demonstrate that even minor changes in expected streamflows can significantly impact transfer activity and costs. |
