Water utilities are increasingly looking to decentralized green infrastructure (GI) as a means of increasing the hydrologic functionality of urbanized watersheds. Though the funding for GI is principally to address wet weather management, GI systems also provide a wide range of other, stakeholder-valued, ecosystem services. Quantification of the value of these other services is very important in efforts to foster high levels of GI adoption across physically, socially, and institutionally heterogeneous urban landscapes. This talk presents recent efforts my lab has taken to quantify how water moves through different kinds of urban GI, emphasizing specifically where field observations deviate significantly from normative theories. Because local preferences often play a significant role in determining where, when, and how GI systems are built, the presentation will also discuss preliminary efforts to use agent-based models to simulate the emergence of GI across space and time as a derivative of individual and institutional decisions. The case is made for an interdisciplinary foundation for young engineers and design professionals, a task that can be engendered through problem-based learning and applied research.