Parameter sensitivity analysis of storm water management model for densely built-up urban areas

To achieve efficient calibration of the storm water management model (SWMM) for Wuzhong city, modified Morris screening method and mutual information method are used to analyze the local and global sensitivity of simulated peak flow and runoff coefficient to 7 parameters of the model． Both methods identified IMP-N and IMP-DS as the two major sensitive parameters of SWMM． Peak flow is the most sensitive to IMP-N and IMP-DS, runoff coefficient is th most sensitive to IMP-DS． The sensitivity of the parameters increases first and then decreases with increasing rainfall intensity． Sensitivity of peak flow to IMP-N and IMP-DS reaches a maximum at about 10 and 3 a, respectively． Sensitivity of runoff coefficient to IMP-N and IMP-DS reaches maximum at about 3 and 2 a, respectively． Synergistic effects between parameters decrease with increasing rain intensity． These data suggest that surface roughness be increased, water storage depth be larger in areas prone to flooding in Wuzhong city． This work will provide some reference for peak flow and drainage reduction in densely built-up urban areas characterized with high impermeability．