Green Infrastructure Evaluation and Planning for Adaptive Stormwater Management
Embargo until
Date
2018-10-26
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Johns Hopkins University
Abstract
Stormwater has been a significant source of pollution in water bodies adjacent to urban area. In cities with combined sewer systems, stormwater also causes combined sewer overflows (CSOs), resulting in disturbance of water uses and threats to human and ecosystem health. Green infrastructure (GI), which utilizes natural hydrological processes to treat stormwater, is argued to be a more sustainable solution for stormwater pollution comparing to traditional engineering solutions. However, literature has indicated that decision makers are facing the uncertainty concerning GI efficacy in controlling stormwater pollution and costs. To facilitate GI planning to manage the uncertainty, this dissertation develops three innovative tools for adaptive investment planning, integrated evaluation and planning, and CSO control for GI, respectively.
Chapter 2 introduces a new method for adaptive investment planning based on the idea of Bayesian inference where the near-term investment may result in learning about GI’s cost-effectiveness, while constraining the risk of the undesired outcomes at a user specified level. Although this method is developed for GI planning, it is generalized for adaptive management problems. A hypothetical example is presented to demonstrate its ability to identify tradeoffs between alternative Pareto optimal investment strategies.
Chapter 3 introduces the integrated evaluation and planning framework for GI, which combines hydrological simulation for GI performance assessment and optimization of GI investment. The GI investment planning extends the adaptive investment planning method with the consideration of performance deterioration and the knowledge transfer between locations in a case study in Philadelphia, PA.
Chapter 4 presents a theoretical framework for the evaluation of GI performance in CSO control. The analysis focuses on the characterization of the interactions between GI, the watershed and the climate. Furthermore, boundary lines are derived from the theoretical framework and the critical flow for CSO that separate CSO generating storms and indicate which storms may be treated by GI. Simulations are performed with several decades of precipitation records from Philadelphia and Seattle.
In conclusion, this dissertation develops a new method for adaptive investment planning, an integrated framework for GI evaluation and planning, and a theoretical framework for CSO control.
Description
Keywords
Stormwater, Green Infrastructure, Stochastic Programming, Learning, Adaptive Management