Stormwater Green Infrastructure (SGI) systems such as rain gardens, permeable pavement and bioswales are commonly used in municipalities to reduce urban flooding and water pollution. In conjunction with these direct benefits, SGI systems provide additional social and environmental "co-benefits". Our goal was to investigate the co-benefits of commonly used SGI systems in five cities in the United States, including Baltimore, Denver, New York City, Philadelphia, and Portland. The i-Tree Eco model was used to predict carbon storage and sequestration, air pollution removal, UV reduction, and cooling effects of trees for individual tree species and estimated SGI tree inventories across the five study cities based on observed tree characteristic data. Aspects of SGI design, environmental factors, and model inputs were assessed to understand what parameters impacted SGI co-benefits predicted by the model. We evaluated the most highly influential parameters using a global sensitivity analysis method. As expected, the type of SGI design, and the overall number of trees utilized within those designs, played a large role in determining the overall amount of co-benefits predicted by the model. However, climate also influenced estimation of benefits produced, with similar responses predicted for cities in the same climate zone (e.g. Baltimore, Philadelphia, and New York City). In particular, the global sensitivity analysis showed that variables influencing environmental conditions and tree growth also impacted final co-benefit predictions produced by i-Tree Eco. study revealed how various assumptions and prevailing equations within the i-Tree Eco model can play a major role in the final outcomes predicted by the model. Studies that use i-Tree Eco to analyze potential co-benefits of SGI projects, especially when the goal is to compare projects across climate zones, should consider what aspects of the results are simply a function of the model itself. Overall, the model predicts that more co-benefits are provided in certain climate zones, an assumption currently supported in the literature.
