Abstract
Timely generation of accurate and reliable forecasts of flash flood events is of paramount importance for flood early warning systems in urban areas. Although physically based models are able to provide realistic reproductions of fast-developing inundation maps in high resolutions, the high computational demand of such hydraulic models makes them difficult to be implemented as part of real-time forecasting systems. This paper evaluates the use of a hybrid machine learning approach as a surrogate of a quasi-2D urban flood inundation model developed in PCSWMM for an urban catchment located in Toronto (Ontario, Canada). The capability to replicate the behavior of the hydraulic model was evaluated through multiple performance metrics considering error, bias, correlation, and contingency table analysis. Results indicate that the surrogate system can provide useful forecasts for decision makers by rapidly generating future flood inundation maps comparable to the simulations of physically based models. The experimental tool developed can issue reliable alerts of upcoming inundation depths on traffic locations within one to two hours of lead time, which is sufficient for the adoption of important preventive actions. These promising outcomes were achieved in a deterministic setup and use only past records of precipitation and discharge as input during runtime.
Highlights
The World Meteorological Organization estimates that flash floods are responsible for more than 5000 human deaths every year, making them the most frequent and with highest morality rate type of flooding [1]
Our study aims to fill this research gap by evaluating the applicability of a hybrid NARX-Self-Organizing Maps (SOM) structure to surrogate a coupled hydrologic/hydraulic model in the prediction of inundation maps (IMs) in a timely and accurate manner compatible with operational purposes for the Don River Basin in Toronto, Canada
Works This work presents evidence that hybrid NARX-SOM models can be effective in predicting inundation maps for flash floods caused by river overflow of catchments with short response time
Summary
The World Meteorological Organization estimates that flash floods are responsible for more than 5000 human deaths every year, making them the most frequent and with highest morality rate type of flooding [1]. Such a strategy has as drawbacks: (1) the need to train a potentially extensive number of subnetworks, which can and up being a constraint in operational environments when the hydraulic/hydrological model is continuously updated ( demanding continuous retraining of the surrogate model); and (2) the fact that the spatial correlation of nearby grids trained by independent predictors is lost, potentially resulting in unrealistic discontinuities in the estimated IM In this context, the use of a Self-Organizing Maps (SOM) model was proposed as a dimensionality reduction method by Chang et al [23] for regional flood inundation forecasting triggered by river overflow. Our study aims to fill this research gap by evaluating the applicability of a hybrid NARX-SOM structure to surrogate a coupled hydrologic/hydraulic model in the prediction of IMs in a timely and accurate manner compatible with operational purposes for the Don River Basin in Toronto, Canada
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
