Expansion of Ensemble-based Sensitivity to TC Hazard Forecasts

Tropical cyclones form and travel through oceanic regions, which can be characterized by relatively few in situ observations relative to continental locations. The relative difficulty and high cost of taking observations over the ocean means it is important to identify optimal locations where assimilating observations into numerical forecast models could lead to the largest change in some measure of a forecast outcome (i.e., a forecast metric). Ensemble-based sensitivity methods are an attractive way of estimating this quantity because it is computationally inexpensive and provides flexibility in terms of forecast outcomes and fields. Since 2019, the National Hurricane Center (NHC) has used guidance from the ensemble-based sensitivity technique to identify locations where the TC track forecast is sensitive to the steering flow, which in turn has been used to develop tracks for the NOAA Gulfstream IV aircraft tracks and for supplemental rawinsonde observations. Unfortunately, the TC track does not fully describe the location and magnitude of potential hazards, such as the wind speed and precipitation; therefore a revision of this technique could provide and even more useful information. This talk describes the extension of this sensitivity technique to a set of new metrics that quantify variability in the TC-related 2D wind and precipitation forecast using output from the European Centre for Medium Range Weather Forecasts (ECMWF) and Global Ensemble Forecasting System (GEFS) ensemble systems. Following a description of the automated algorithms that have been developed for each hazard-based metric, representative cases will be shown from running the system in real time during the 2023 season. For many cases, the wind field and precipitation forecasts appear to be sensitive to similar locations as the track forecast.