An Operational Implementation Of The EPA-SWMM Model With Radar QPE Fields And Precipitation Nowcast For Urban Hazards Assessment

Resilience and risk management in large metropolitan areas strongly depend on the state, maintenance, and control of critical urban infrastructure. In particular, the behavior of conventional engineered infrastructure (grey infrastructure), including but not limited to streets, bridges, pipelines, canals, sewer and urban drainage systems during extreme weather events often determine the associated risk level for the population and the operations efficiency of a city as a whole. Most of the critical engineered infrastructure is reliant on the full functionality of one or more other critical systems to ensure the fulfillment of essential services to the public. In Medellin, Colombia, often the urban drainage system in some specific areas acts as the weak link in risk management, collapsing during extreme storm events, affecting the road network and traffic, flooding surrounding buildings, and even generating road flash road in steep areas threatening human lives. Early alert of an urban drainage system collapse could lead to better traffic management, a faster response of maintenance personnel, and protection of human life. In this work, we present an operational methodology to provide early warning for vulnerable sections of the urban drainage system using the EPA-SWMM model, antecedent and real-time rainfall estimates based on a C-band radar quantitative precipitation methodology, and precipitation nowcasts using a Lagrangian extrapolation method. The time-step of the model is 5 minute and provides the current and future states of all the nodes in a sub-drainage system. Due to the lack of level and streamflow sensors in the drainage system, the methodology is validated using three collapse episodes that occurred between 2016 and 2017. Current work includes the deployment of low-cost real-time level sensors and video cameras in specific areas.