Assessing the impact of incomplete stormwater network data on uncertainty in simulation results

Insufficient or erroneous stormwater network data can lead to significant uncertainties in model simulation results, which may adversely affect decision-making processes. Therefore, it is crucial to develop effective methodologies for evaluating and mitigating these uncertainties. This study assesses the impact of rainfall characteristics and spatial connectivity on the effects of missing parameters at both the system level and the local manhole level. The results indicate that: 1) At the system level, the absence of slope has a relatively minor impact, accounting for only 11 % to 13 % of the effects caused by the absence of diameter. 2) An increase in rainfall intensity correlates with a reduction in the relative error of simulation results, while a delayed rainfall peak mitigates the impact of parameter missingness. 3) As the spatial connectivity of pipes with missing parameters increases, the magnitude of the error also rises; specifically, in a rainfall event with a 1-year return period, the maximum errors for 2, 3, 5, and 10 connected pipes are 66, 79.8, 130.8, and 297 m³, respectively. 4) The impact of missing slope on manhole waterlogging is greater than that of missing pipe diameter. To keep the relative error of the manholes within 20 %, the percentage of missing slope should be kept below 10 %, while the percentage of missing diameter should be limited to below 30 %. These findings enhance our understanding of the influence of missing parameters on SWMM simulations and provide essential guidance for the processes of parameter simplification, calibration, and optimization.