Initial Condition Sensitivity and Predictability of Tropical Cyclone Development and Intensification

It has been suggested in a number of previous studies that the development of tropical cyclones and intensification may be sensitive to aspects of large-scale forcing, as well as internal mesoscale dynamics. In this study, we explore the hypothesis that the development and intensification of tropical cyclones are sensitive to small perturbations to the basic properties of the background state through organized mesoscale convection and synoptic-scale forcing. The recently developed adjoint and tangent linear models for the atmospheric portion of the nonhydrostatic Coupled Atmosphere/Ocean Mesoscale Prediction System (COAMPS) are used to explore the mesoscale sensitivity of tropical cyclone development and subsequent intensification (or lack of) to the initial state. A unique aspect of this system is that an exact adjoint to the explicit microphysics has been developed. The forward, adjoint and tangent linear models are applied at horizontal resolutions ranging from 10-40 km and are used to explore predictability issues for several tropical cyclones and non-developing storms. We will focus on the Western Pacific basin during the THORPEX Pacific Asian Regional Campaign (T-PARC) and the ONR Tropical Cyclone Structure-08 (TCS08) experiments (fall 2008) and the Impact of Typhoons on the Ocean in the Pacific (ITOP) (fall 2010). Preliminary results from the W. Atlantic basin will be presented based on the first field phase of the NASA Hurricane and Severe Storm Sentinel (HS3), planned for the fall 2012.

The adjoint results indicate that the short term (6-24-h) forecasts of tropical cyclone intensity (e.g., kinetic energy) are very sensitive to the initial state. The adjoint-based sensitivity fields indicate highly structured patterns in the wind, thermal, moisture, and microphysical fields that project on to the model simulated deep convection, which ultimately influences the intensification rate. The highest-resolution adjoint simulations (~10 km grid increment) indicate that the most efficient intensification is through low- and mid-level moistening and heating in banded regions that are coincident with vorticity maxima in the initial state. Growth originates at small scales and projects on to the scale of the vortex, a manifestation of perturbations that project onto organized convection embedded in regions of cyclonic vorticity.