Peterborough Integrated Flood Model: A Comprehensive Approach to Managing Flood Risk

This presentation will highlight the development of an Integrated Flood Model (IFM) for the City of Peterborough, Ontario. The City-wide IFM was built in MIKE+ and divided into 5 models which include 350 km of storm sewers, 41 km of 1D riverine system, 40 storm water management ponds, and approximately 55 km2 of two-dimensional mesh to represent the existing terrain. The fully dynamic 1D/2D IFM model provides the City with an effective tool to assess, forecast, plan, and mitigate against the risk of urban and riverine flooding events. This tool will act as a data-driven and evidence-based tool to support the City's decision-making, planning and prioritizing of future flood mitigation projects, and to inform asset management efforts for both existing and future conditions. The impetus for this project was a severe urban storm in the City of Peterborough in 2004, which resulted in over $100M in damages, and the likelihood of hybrid (urban-riverine) flooding due to the rivers traversing the city and increasing urbanization. To address the City's resilience to flooding events, a cross-collaborative effort was undertaken to develop performance scenarios including the City's planned growth and urbanization, climate change forecasts, and evaluate the impact (positive or negative) of adding the City's planned flood reduction and Low Impact Development (LID) projects. This process involved the complete model build-out using GIS information provided by the City and discretizing of the model's catchments to a catch basin level. Additionally, the project used master plan information to identify locations of planned City expansion and densification to develop a high-level schematization to evaluate potential impacts to the existing collection system and watercourses. The model also included a high level representation of LIDs to assess its impact in terms of both flood mitigation through attenuation of overland flows, and potential water quality benefits through increased infiltration potential. The approach used to develop the IFM was selected to provide a balance between providing a high level of detail and minimizing simulation run times. For example, rather than providing a fully 'rain-on-grid' schematization, a hybrid approach was selected by using delineated subcatchments to perform the hydrologic simulation which are coupled to the two-dimensional mesh via one dimensional elements. This approach allows for entirely 1D simulations to be performed for smaller storm events where it is safe to assume the collection system will capture a majority of runoff. Additionally, since rain on grid simulation by nature will activate the entire two dimensional mesh during simulations, the level of refinement that is required to avoid localized areas causing a model instability can be significant. Particularly in the case of the IFM where schedule was a major driver for the approach for the model build, using a hybrid approach can greatly reduce troubleshooting while performing simulation of larger storm events. A full suite of storm events was used to evaluate each performance scenario to help identify flood risk associated with both volume- (historical) and intensity- (design storms) based events. Intensity, Duration and Frequency (IDF) Curves were adjusted using the tool developed by the Institute of Catastrophic Loss Reduction (ICLR) and Western University to calculate projected rainfall intensity which incorporates the effects of climate change. An IFM model of this scale generates multitudes of data that must be post-processed to create the tools and products to support the City's decision making. To effectively communicate and illustrate results and analyses, an innovative approach was required to compile GIS data into a single geodatabase and automate the processing of model results. The IFM results were used to facilitate a comprehensive GIS analysis for each scenario to evaluate a variety of different metrics related to the extent and severity of flooding for the entire city, which included the number of impacted 'critical' sites and buildings, deficiencies in the storm sewer system, and locations of excessive overland flooding. Dynamic flood maps and dashboards were also developed using ArcGIS Online to give the City the ability to review, compare and/or share results with stakeholders for each performance scenario effectively, considering the large scale of the project. Training sessions with City staff on how to operate, modify and maintain the completed models as well as the master GIS database were also completed to ensure that the integrated flood model will be a useful tool for this City in the years to come.