A Tale of Two Models: How Agent Based Modeling Filled in the Human Variables of Hydraulic Modeling for Low Impact Development in Somerville, MA

Hydraulic and hydrologic models (e.g. EPA SWMM (Rossman, 2009)) have historically been used by engineers to investigate the physical impact of proposed or potential stormwater management techniques, but the social constraints and incentives to implementing such techniques remain elusive to this type of modeling. Agent Based Modeling (ABM) can model human effects, such as positive or negative word of mouth information dissemination or the impact of educational programs, and how those components effect implementation of Low Impact Development (LID) stormwater management techniques. Used together, the models can provide a more complete picture of the possibilities for stormwater management in a given urban watershed, and the best way to achieve stormwater management in a changing climate, which will likely include both structural/physical and non-structural/social approaches (Parkinson, 2003). The models complement each other, providing the ability to incorporate more of the important factors in a real system, and improving the power of both models by capturing interactions between the social and physical components of the system.Under new climate regimes, precipitation events are predicted to be more extreme (Katz and Brown, 1991; Emori and Brown, 2005). Heavily urbanized areas, such as Somerville, MA, are likely to experience flooding and other impacts from increased runoff. The northeastern United States and the metropolitan Boston area specifically may be at high risk for flooding and property damage under changing climate scenarios (Kirshen et al, 2008). Although low impact development (LID) has the potential to decentralize stormwater management, these techniques have been slow to be implemented at the local level (Shuster et al, 2008). Using Somerville as a test case, this study used SWMM to model the physical possibilities for stormwater management. We then investigated stakeholder opinions around stormwater management and responsibilities, and modeled the social possibilities for management using ABM.