Files
Abstract
Uncertainty in our knowledge of the behavior of environmental systems should be fully addressed and properly accounted for, if we are to have effective and reliable environmental policies. To that end, a systematic framework for watershed modeling and the analysis of uncertainty is proposed and presented in this dissertation. The framework houses several major components: a watershed simulation tool, itself composed of several parts; algorithms for the analysis of uncertainty and sensitivity; and assessment of the economics of water quality trading. Central to the analysis of uncertainty is a sampling-based approach based on what is familiarly known as a Regionalized Sensitivity Analysis. A case study is employed to illustrate the capabilities of the proposed framework. First, the problem of model identification from field data under uncertainty is discussed. Second, a more comprehensive watershed model is assembled in order to study two important aspects of water quality management in a significant portion of the Chattahoochee watershed, including the Metropolitan Atlanta area, again focusing on the nutrient phosphorus. Significantly, the watershed model treats the dynamic behavior of both nonpoint- and point-source discharges of phosphorus on a compatible basis. In the first part of this case study, suitable regulatory standards for water quality are used to develop a TMDL for phosphorus, which is allocated amongst the nonpoint and point sources in the watershed. The Chattahoochee case study is then completed with a more advanced analysis of nutrient trading, between point- and nonpoint-source dischargers. Feasibility and desirability of any such trading scheme are assessed according to the dual measures of economic surplus in the potential trading market and improvement in water quality. All the uncertainties involved, both scientific and economic, are evaluated by means of the sampling-based RSA algorithm. Since uncertainty is conventionally handled through a rather crude choice of a Margin of Safety in setting TMDL policies, and through the so-called trading ratio crucially for the design of effective water quality trading policies, the Chattahoochee case study realistically address two important contemporary problems. The dissertation argues that its proposed approach to the analysis of uncertainty in such matters is an important contribution to the design of reliable environment policies in the future.