An Analysis of Recent P-Surge Performance Relative to a Historical Average

The accuracy of a storm surge forecast is highly dependent upon whether the tropical cyclone (TC) forecasts of track, intensity, and size are skillful. Due to forecast uncertainty related to these TC parameters, it is necessary to use a probabilistic framework in which the storm surge risk is calculated from an ensemble of scenarios that vary the TC’s track, forward speed, intensity, and size. The Storm Surge Unit at the National Hurricane Center (NHC) relies on real-time guidance from the Probabilistic Tropical Storm Surge (P-Surge) model to forecast the storm surge risk associated with a landfalling TC. Previous studies have evaluated the performance of the P-Surge model using a set of 25 tropical cyclones from 2008 to 2020. This study will expand upon previous work by including more recent storms and will examine the performance of the latest P-Surge upgrade (v3.0), as well as other experimental probabilistic guidance products. Additionally, it will also explore the predictability of P-Surge forecasts by examining results from “Perfect Prognostic” simulations that use the NHC Best Track to define the storm parameters instead of the Official Forecast. The compilation of these verification results will highlight the skill of the current operational version of P-Surge as well as provide a benchmark to evaluate future upgrades.