Abstract
Ensemble forecasting is well-established as an effective tool for understanding forecast uncertainty. Ensembles of Tropical Cyclone (TC) forecasts have been used to determine the sensitivity of TC track forecasts to synoptic scale features such as subtropical ridges and longwave troughs. This study extends the ensemble sensitivity analysis technique by exploring the sensitivity of TC intensity forecasts to mesoscale features such as inner-core moisture and mid-tropospheric radial ventilation. The feasibility of such a technique is discussed using the HAFS (Hurricane Analysis and Forecast System) Ensemble Running in the Cloud, or HERC, as a test bed after establishing HERC’s reliability as a skillful high-resolution ensemble via storm track and intensity verification metrics. Using the 2025 cases of Hurricanes Erin, Gabrielle, and Melissa, this technique demonstrates utility in identifying key regions of sensitivity to radial ventilation and environmental relative humidity without necessitating analysis of each individual ensemble member. Such a technique has applications in both research and operational forecasting; it can quickly help forecasters identify important regions of developing storms, and it can serve as an analysis tool when studying the mechanisms responsible for TC development and nondevelopment. While this analysis shows promise, further work is required to establish a consistent methodology, and statistical significance remains a challenge to establish in the small ensemble of 21 members.