Modeling Urban Hydrology and Green Infrastructure using the AGWA Urban Tool and the KINEROS2 Model

Abstract

Green Infrastructure (GI) is being widely used to mitigate the effects of increased stormwater runoff due to urbanization. There is a need for hydrologic models to represent the various elements of urban hydrology and GI, discretely, spatially, and accurately. The Kinematic Runoff and Erosion (KINEROS2) model contains an Urban element that can represent various flow-on/flow-off processes at the parcel scale, including GI practices such as retention basins, permeable pavements, and roof runoff harvesting. The Automated Geospatial Watershed Assessment (AGWA) tool provides a geographic information systems (GIS) based framework that uses spatial datasets to generate parameters, execute the KINEROS2 model, and import the results in GIS for visualization and analysis. Three studies were conducted using the modeling capabilities of the AGWA tool and KINEROS2. Study 1 models the La Terraza subdivision in Sierra Vista, AZ, by individually representing 66 parcels using the KINEROS2 Urban element. A roof runoff harvesting analysis estimated a 15% capture of rainfall volume by roofs for all parcels combined over a 10-year simulation period. Comparison of different GI implementation scenarios reveal that retention basins had the highest potential of peak flow and runoff volume reduction, compared to permeable pavements or roof runoff harvesting. Study 2 models 2177 parcels in the High School watershed in Tucson, AZ. Current level of GI implementation consisting of 175 on-street basins and 37 cisterns, was simulated to estimate up to 2% reduction in runoff volume and up to 1% increase in peak flows at the watershed outlet for three design storms (5-year, 25-year and 100-year, 1-hour duration). However, localized flow accumulation was reduced by more than 25% in areas where GI was installed. Results from a 20-year simulation period indicate that roof runoff capture, as well as infiltrated water in retention basins has the potential to meet the water needs of xeric vegetation. Finally, study 3 analyzes the influence of KINEROS2 parameters that represent GI practices on infiltration, peak flow and runoff volumes, both at the watershed scale as well as the parcel scale, for different precipitation events. The global sensitivity analysis reveals that parameters differ in sensitivity across various precipitation intensities, durations, and shapes, as well as between parcel and watershed scales.