Predictability and Dynamics of a Vertically-Sheared Tropical Storm

            The past decade has brought significant improvements in tropical cyclone (TC) forecast track improvements while intensity forecasts have lagged far behind.  Our ability to accurately predict cyclogenesis, rapid intensification, and decay are severely limited.  Though there have been significant improvements in numerical weather prediction (NWP) models and data assimilation (DA) techniques, the initialization of today's forecast models still needs improvement.  More specifically, while the assimilation of synthetic vortices to initialize forecast models has skill in predicting well organized TCs, it has inherent limitations with weak and vertically sheared storms.  Such storms may exhibit asymmetries that are not well captured into a symmetric and vertically stacked synthetic vortex.  Of greater concern are the effects of initial asymmetries and vortex tilting on TC predictability.

            This study utilizes a unique case study of Tropical Storm Erika (2009) to assess the predictability and dynamics of vertically-sheared TCs.  Erika's intensity forecasts were very poor with respect to climatological forecast errors and presents the weaknesses of current forecast models to accurately predict the life-cycle of a weak, disorganized tropical system.  Two, 60-member ensemble forecasts from the Weather Research and Forecasting (WRF) model compose the dataset.  An Ensemble-Kalman Filter (EnKF) is utilized and small perturbations are introduced into initial model fields (Sippel and Zhang 2009), which produce large ensemble spreads consistent with mesoscale convective vortex studies by Hawblitzel et al. (2007).  It will be shown that there are significant limitations in the predictability of weak, vertically sheared tropical cyclones and that the vertical alignment of the initial model vortex may play a significant role in such processes.  This study also displays the inherent need for operational use of ensemble-based forecasting systems that can accurately convey the uncertainty for systems in which moist convective processes significantly reduce predictability.