The transferability of SWMM model parameters between green roofs with similar build-up

While extensive green roofs are popular measures for reducing and delaying stormwater runoff, design tools are needed to better predict roof performance based on material properties, geometry and climate. This paper investigates the EPA's Storm Water Management Model's (SWMM) green roof module for this purpose based on observed runoff from several roofs with different build-ups, geometry, and climates. First, the general model performance was investigated and secondly transferability of model parameters for similar roofs but different geometries and climates was tested. Individual models reproduced runoff hydrographs well (NSE 0.56-0.96), while the long-term modelling showed relatively large volume errors most likely due to insufficient representation of evapotranspiration in the model. Model parameters obtained at one site were only partly transferable to similar roof build-up at other sites. Transferability was better from models calibrated with wetter climates and higher intensity events to drier climates, than the opposite way. Multi-site calibration resulted in model parameters performing well for most sites, giving model parameters that could be used for the design of similar roof build-ups in comparable climates. However, large variability in obtained model parameters, large volume errors and the fact that the calibrated model parameters did not directly correspond to measured material properties, place concerns on the generality of the SWMM green roof module as a design tool.