A new operational model for seasonal high tide flooding prediction

Minor coastal flooding, also known as high tide flooding, is quickly becoming a major challenge for coastal communities across the United States. Many coastal locations throughout the U.S. already experience 10 days or more of flooding per year, a number that is likely to increase dramatically over the next several decades with continued relative sea level rise. Communities require tools to enable them to better prepare for coastal flooding impacts in the months to years ahead. Here we present a novel statistical model to predict daily high tide flooding likelihood with lead-times up to one year. The model relies on astronomical tide predictions, sea level trends, monthly sea level anomaly persistence and climatological non-tidal residuals to predict daily flooding likelihood at U.S. National Oceanic and Atmospheric Administration (NOAA) tide gauge locations. Published results show the method to be skillful at predicting flooding a year or more in advance at most tested U.S. NOAA tide gauges. At 18 gauges the model accurately predicted more than half of all floods occurring from 1997-2019 with less than a 10% False-Alarm-Rate. In this presentation we also show the newly operational NOAA visualization products, which deliver model predictions for end-user decision-support. Lastly, we show initial results of the application of the model to a 43 year, 500m resolution hydrodynamic model reanalysis of coastal water levels to enable predictions away from tide gauge locations.