Influence of spatial heterogeneity of rainfall on drainage network performance

Urban flooding, driven by increasingly intense and localized rainfall events, is a critical challenge to stormwater infrastructure in densely built environments. This thesis investigates the hydraulic response of an urban drainage network located in a part of the city of Turin, Italy under spatially variable rainfall conditions using a numerical modeling approach with the Storm Water Management Model (SWMM). The primary objective is to assess how the spatial structure of rainfall, specifically the location and shape of precipitation, affects the performance and criticality of the drainage system. An exponential distribution is applied to simulate rainfall fields of varying return periods (2, 5, and 10 years), durations (10 minutes and 1 hour), and spatial configurations (two distinct storm centers). The total rainfall volume is changing across scenarios and the urban drainage model, composed of over 300 subcatchments and 800 nodes, is used to simulate runoff generation and propagation under each scenario. Results show that both the location of the maximum rainfall intensity and the shape of the rainfall significantly affect network performance. Upstream-centered storms that are far from the main outlet consisting of smaller conduit networks lead to higher flooded volumes, more pressurized conduits, and greater spatial heterogeneity in flooding. Short-duration, high-intensity events revealed the most severe system stresses among the scenarios. There is a need to integrate spatial rainfall variability into urban flood risk assessments. For this purpose, the study highlights how conventional methods that rely on uniform or temporally distributed rainfall may underestimate system vulnerabilities.