A numerical study of the role of the vertical structure of vorticity during tropical cyclone genesis

An eight-level axisymmetric model with simple parameterizations for clouds and the atmospheric boundary layer was developed to examine the evolution of vortices that are precursors to tropical cyclones. The effect of vertical distributions of vorticity, especially that arising from a merger of mid-level vortices, was studied by us to provide support for a new vortex-merger theory of tropical cyclone genesis. The basic model was validated with the analytical results available for the spin-down of axisymmetric vortices. With the inclusion of the cloud and boundary layer parameterizations, the evolution of deep vortices into hurricanes and the subsequent decay are simulated quite well. The effects of several parameters such as the initial vortex strength, radius of maximum winds, sea-surface temperature and latitude (Coriolis parameter) on the evolution were examined. A new finding is the manner in which mid-level vortices of the same strength decay and how, on simulated merger of these mid-level vortices, the resulting vortex amplifies to hurricane strength in a realistic time frame. The importance of sea-surface temperature on the evolution of full vortices was studied and explained. Also it was found that the strength of the surface vortex determines the time taken by the deep vortex to amplify to hurricane strength.

A divide and conquer strategy for scaling weather simulations with multiple regions of interest

Accurate and timely prediction of weather phenomena, such as hurricanes and flash floods, require high-fidelity compute intensive simulations of multiple finer regions of interest within a coarse simulation domain. Current weather applications execute these nested simulations sequentially using all the available processors, which is sub-optimal due to their sub-linear scalability. In this work, we present a strategy for parallel execution of multiple nested domain simulations based on partitioning the 2-D processor grid into disjoint rectangular regions associated with each domain. We propose a novel combination of performance prediction, processor allocation methods and topology-aware mapping of the regions on torus interconnects. Experiments on IBM Blue Gene systems using WRF show that the proposed strategies result in performance improvement of up to 33% with topology-oblivious mapping and up to additional 7% with topology-aware mapping over the default sequential strategy.