Effect of cloud processes on hurricane tracks: Idealized simulations and operational forecasts

Recent work has demonstrated that cloud processes can have a substantial influence on hurricane motion. In particular, Fovell and Su (2007) showed that considerable track spread could be realized from a single model and initial condition via a physics-based ensemble that varied cloud microphysics and cumulus parameterization schemes. Using high-resolution idealized simulations, Fovell, Corbosiero and Kuo (2009) discussed how and why microphysical assumptions underlying the evolution and spread of condensate particles directly influences storm characteristics that control motion, including storm size, depth and the degree to which the storm modifies its own environment. In particular, they showed that manipulating aspects such as hydrometeor fallspeeds and cloud-radiative interactions can influence, in some cases dramatically, forecast position even in relatively short (~48h) time frames.

We will present new findings regarding cloud processes and hurricane track. Additional idealized simulations have helped to further clarify the role of microphysics in cyclone motion. In addition, a more extensive and systematic effort at employing convection-based ensembles for track forecasting using relatively low resolution models that took place during the last two Atlantic hurricane seasons has revealed some surprises, including unexpectedly high skill for a subset of the ensemble members. The goal of this work is to demonstrate that relatively poorly understood and handled cloud processes modulate hurricane motion to a sufficiently large degree as to represent a significant contribution to forecast uncertainty, and thus form a reasonable basis for an ensemble forecasting system.