- Storm surges,
- North Atlantic oscillation -- Environmental aspects,
- North Atlantic Ocean -- Climate
Recent studies of flood risk at New York Harbor (NYH) have shown disparate results for the 100 year storm tide, providing an uncertain foundation for the flood mitigation response after Hurricane Sandy. Here we present a flood hazard assessment that improves confidence in our understanding of the region's present-day potential for flooding, by separately including the contribution of tropical cyclones (TCs) and extratropical cyclones (ETCs), and validating our modeling study at multiple stages against historical observations. The TC assessment is based on a climatology of 606 synthetic storms developed from a statistical-stochastic model of North Atlantic TCs. The ETC assessment is based on simulations of historical storms with many random tide scenarios. Synthetic TC landfall rates and the final TC and ETC flood exceedance curves are all shown to be consistent with curves computed using historical data, within 95% confidence ranges. Combining the ETC and TC results together, the 100 year return period storm tide at NYH is 2.70 m (2.51–2.92 at 95% confidence), and Hurricane Sandy's storm tide of 3.38 m was a 260 year (170–420) storm tide. Deeper analyses of historical flood reports from estimated Category-3 hurricanes in 1788 and 1821 lead to new estimates and reduced uncertainties for their floods and show that Sandy's storm tide was the largest at NYH back to at least 1700. The flood exceedance curves for ETCs and TCs have sharply different slopes due to their differing meteorology and frequency, warranting separate treatment in hazard assessments.