H52F-02 A novel approach to scalable data driven pluvial flood modeling utilizing a 1&2D hydrodynamic flood model tool for urban flood management

During the summer of 2023 the Cleveland experienced extreme convective storms that resulted in significant localized flooding and damage. The nature of intense rainfall was such that the limited capacity of catch basins to transfer this intense runoff to the subsurface drainage system resulting in major surface flooding over the urban landscape.

Ex-post simulation of storm events using a catchment flood model of the drainage system showed no flooding and that the drainage system was adequate. Figure 1 shows flooding at a rail underpass not identified by the catchment model.

The failure of catchment models to capture surface flooding has driven this research in developing a novel approach to scalable data driven pluvial flood modeling utilizing 2-D Rain-on-Mesh hydrodynamic flood model, ICM, combined with detailed geo-specificity of the urban textures. Modeling of the Cleveland event using, ICM, that explicitly models surface runoff, produced model output identifying flooding at the precise points that flooding occurred.

To address the shortcomings of catchment-based models, the Northeast Ohio Regional Sewer District partnered with MIT to start the development of a regional 1&2-D rain-on-Mesh ICM model. Using a new tool for policy making and investment decision making requires that the model be validated.

A major element of employing high resolution 2-D flood models is validation. There have been recent advancements in sensors, monitoring as well as real time crowd-sourcing data to support fidelity of flood impact. However, even where they exist, they are rarely used for spatial-temporal flood model result validations.

Northeast Ohio Regional Sewer District has in place a comprehensive flood monitoring and reporting system for Cleveland and the City of Cambridge has detail data for recent flash flood.

This presentation reports on the novel approach of validating the ICM model using the unique and comprehensive processes and tools that the combines a wide network of remote sensing, monitoring and crowd sourcing data for high fidelity of validation for spatial-temporal ICM results. This approach was applied to two city scale models for the case studies of Cleveland OH and Cambridge MA, with collaboration with the municipalities.