Modeling and Dissecting Hurricane Sandy's Storm Surge and Overland Inundation

Hurricane Sandy caused flooding in Northern New Jersey and New York City (NYC) that exceeded all the measurements of water level in the historical record and also exceeded all estimated flood levels for hurricanes that made landfall at NYC over the two prior centuries. Numerical model storm surge forecasts at NYC were generally lower than the observed elevations. Here we report on model experiments run to examine several potential sources of the low bias. Omitting remote forcing in the forecast model runs leads to a large decrease in peak water elevation. The accuracy of the meteorological forecasts used in the analysis varied substantially with Sandy, and this is shown to have also had a major impact on storm surge forecast accuracy. A simulation made with the best available wind forecast provides the most accurate water level forecast with the lowest RMS error average across five stations (0.16 m) and the best forecast skill (0.99). Nested model grids have been developed and validated for over-land flood simulations into neighborhoods, demonstrating flood pathways and inundation depths during the event.