The Probabilistic Hurricane Storm Surge model (P-Surge) determines the probability of storm surge due to an active tropical cyclone. This is done by sampling the error space around the National Hurricane Center’s (NHC) official forecaster track as defined by their 5-year average along track, cross track, intensity, and size errors. Each sample is used as input by a parametric wind model to provide winds to drive a dynamic storm surge model (SLOSH) capable of computing inundation based on storm surge and tide. This is done for as many computational grids as necessary to cover the continental US’s coastlines along the Atlantic and Gulf of Mexico. The results of the samples are then combined based on their corresponding likelihood to create both probabilities of exceeding a threshold (e.g., probability of > 5 feet) and heights that exceed a given risk threshold (e.g., 10% exceedance = heights that are exceeded by just 10% of the storms). Those products are then merged onto a uniform grid which is equivalent to the National Digital Forecast Database's (NDFD) CONUS grid, but at 625m (instead of 2.5 km). Additionally a 2.5km product is provided to NHC.

See P-Surge's developmental website here: https://slosh.nws.noaa.gov/psurge

Documentation: https://slosh.nws.noaa.gov/psurge/about.php