To explore the physical processes responsible for SEF, a cycling, ensemble Kalman filter (EnKF) approach is combined with the full physics NCAR Advanced Hurricane Weather Research and Forecasting Model to generate ensemble forecasts of Hurricane Igor (2010), which underwent a well documented SEF and ERC. The data assimilation system generated a 96 member analysis ensemble every 6 h for a basin-scale 36 km domain, and 12 km nested domain that follows the TC, which are then used to initialize high resolution (1.33 km) forecasts out to 120 h. Analysis of 48 members of the ensemble shows significant variation in the evolution of Igor. Approximately one quarter of the members exhibited at least one ERC with the secondary eyewall initiated at varying times and radial distances from the center. Preliminary work suggests that ensemble members that undergo ERCs are more intense, evolve in weaker vertical wind shear, and exhibit a more uniform moisture distribution relative to the members that do not. Future work will explore the differences between ensemble members in the context of inertial stability, angular momentum, inner core vorticity structure, and the evolution of eyewall and inner rainband convection.>