Integration of contributed data with HEC-RAS hydrodynamic model for flood inundation and damage assessment: 2015 Dallas Texas Case Study

Transportation infrastructure networks in urban areas are highly sensitive to natural disasters, yet are a very critical source for the success of rescue, recovery, and renovation operations. Therefore, prompt restoration of such networks is of high importance for disaster relief services. Satellite and aerial images provide data with high spatial and temporal resolution and are a powerful tool for monitoring the environment and mapping the spatio-temporal variability of the Earth's surface. They provide a synoptic overview and give useful environmental information for a wide range of scales, from entire continents to urban areas, with spatial pixel resolutions ranging from kilometers to centimeters. However, sensor limitations are often a serious drawback since no single sensor offers the optimal spectral, spatial, and temporal resolution at the same time. Specific data may not be collected in the time and space most urgently required and/or may it contain gaps as a result of the satellite revisit time, atmospheric opacity, or other obstructions. In this study, the feasibility of integrating multiple sources of contributed data including remotely sensed datasets and open-source geospatial datasets, into hydrodynamic models for flood inundation simulations is assessed. The 2015 Dallas floods that caused up to $61 million dollars in damage was selected for this study. A Hydraulic Engineering Center - River Analysis System (HEC-RAS) model was developed for the study area, using reservoir surcharge releases and geometry provided by the U.S. Army Corps of Engineers Fort Worth District. The simulated flood inundation is compared with the "contributed data" for the location (such as Civil Air Patrol data and WorldView 3 dataset) which indicated the model's lack of representing lateral inflows near the upstream section. An Artificial Neural Network (ANN) model is developed that used local precipitation and discharge values in the vicinity to estimate the lateral flows. This addition of estimated lateral inflows is expected to improve the model performance to match with the observed flows. Future work will focus on extending this preliminary work to assess the model performance after integrating these additional data sources.