Ensemble Sensitivity of Atlantic Basin Tropical Cyclone Intensity to Storm Structure

Prediction of tropical cyclones' (TCs) intensity is a major forecasting challenge. Substantial improvements in the prediction and representation of the large-scale ambient environment have not led to reduced intensity forecast errors in the same steady manner as they have for TC track forecast errors, which suggests that the storm structure plays an important role in the intensity evolution of TCs. Rapid intensification (RI) is particularly difficult to forecast and statistical-dynamical models, such as those of DeMaria and Kaplan, as well as high-resolution numerical modeling studies and satellite observations point to the importance of TC structure in predicting RI.

We have simulated the rapid intensification of several TCs from the 2009, 2010 and 2011 Atlantic basin hurricane seasons using a 96-member WRF model ensemble.The statistics of the ensemble are used to investigate the sensitivity of rapid intensification to TC structure and the validity of the results are tested using perturbed initial condition experiments. Preliminary ensemble sensitivity analysis on Igor (2010) and Earl (2010) show that their intensity is sensitive to kinematic and thermodynamic variables in the inner core of the storm and to rainbands in a region 100 to 200 km from the center of the storm. Perturbed initial condition experiments based on the sensitivity patterns show a linear response even to large perturbations, which produce additional strengthening of about 20 hPa before non-linear effects dominate. Further results from the RI periods of Bill (2009), Julia (2010), and Ophelia (2011) are shown as well as results for Katia (2011) as a false alarm case.