Author: Berina Kilicarslan, Marouane Temimi – Stevens Institute of Technology
Title: High-resolution flood inundation mapping at the block-scale level in the New York City metropolitan area
Abstract: The goal of this study is to simulate flood inundation at a street scale in the Hackensack River Watershed, in New Jersey, US. The study domain, in the New York City Metropolitan area, is modeled with the operational Stevens Flood Advisory System (SFAS), a one-way coupled hydrodynamic-hydrologic system using the Hydrologic Engineering Center’s Hydrologic Modeling System (HEC-HMS) to simulate streamflow. This study investigates the introduction of the Weather Research and Forecasting-Hydrological modeling system (WRF-Hydro) and evaluates the performance of the block-scale (50-m) gridded WRF-Hydro model for a potential integration into the SFAS operational system. The calibration of the model was carried out using the Dynamically Dimensioned Search (DDS) algorithm using streamflow observations. The model performance was assessed using i) streamflow observations, ii) USGS High Water Marks (HWMs), and iii) crowdsourced data on street inundation. The model was validated for Hurricane Irene which made landfall in southern New Jersey on August 28, 2011 and caused significant pluvial flood inundation. Results show that the calibrated version of WRF-Hydro outperforms the default version and the HEC-HMS model with KGE values of 0.82, 0.71, and 0.36, respectively. However, the model exhibited inaccuracies arising from the synthetic rating curves and the modeling structure errors, leading to over and underestimation in flood inundation. Nonetheless, an agreement was noticed between WRF-Hydro and crowdsourced data on flood extent. The future research will concentrate on the integration of the developed version of WRF-Hydro in the National Water Model (NWM) using Basic Model Interface (BMI) according to the NextGen framework.