Informing green infrastructure decisions using multi-objective optimization

Amy Nicole Piscopo
US Environmental Protection Agency, Office of Research and Development
 
Green infrastructure (GI) is gaining recognition for its ability to reduce pollutant loads in 
stormwater runoff. However, to have a meaningful impact on water quality, GI must be implemented at the 
watershed scale. Developing water management plans at this scale is challenging because endless 
combinations of GI type and location are possible, each with different effectiveness in minimizing 
pollutant loads, minimizing costs, and maximizing co-benefits such as reduction of runoff. In this 
study, we efficiently generate management options that balance the tradeoffs between these objectives 
using a multi-objective evolutionary algorithm.  Management options specify the amount and location of 
different types of GI, including bioretention basins, infiltration trenches, and riparian buffers. Our 
approach is demonstrated for an urban watershed in Taunton, Massachusetts.