5:15 PM - 7:15 PM
[AAS04-P02] Large Eddy Simulation of Entire Tropical Cyclone: Meosvortices and Vertical Motions
Keywords:Tropical Cyclone, Mesovortex, Large Eddy Simulation, Rapid Intensification, Vertical Motion
Large Eddy Simulation (LES) enables detailed analyses of vertical motions and small-scale structures in tropical cyclones. Previous studies using LES have provided new insights of them during the mature stage. In this study, we analyzed results from an LES performed on the supercomputer Fugaku, which achieved a LES throughout the entire tropical cyclone and the development. Additionally, we conducted a low-resolution control experiment (DX2 km) to compare. In particular, we investigate how vertical motion in the boundary layer governs both the generation of mesovortices and the overall dynamics leading to rapid intensification (RI).
Although there was no substantial difference in the maximum intensity between the LES and DX2 km simulations, we observed a discrepancy in the timing of RI. The LES tended to show a delayed start of RI. Analysis of vertical vorticity distributions revealed multiple mesovortices, about 10 km in diameter, forming prior to the onset of RI. These mesovortices are likely generated by horizontal shear instability, and energy production of non-axisymmetric components due to horizontal shear. As the mesovortices advected toward the typhoon center, they mixed and eventually dissipated. In the LES, however, they formed later and persisted longer than in DX2 km. Objective identification showed that a greater number of mesovortices appeared in the LES, and they moved more slowly toward the typhoon center compared to DX2 km. Composite analyses of their vertical structure revealed that these mesovortices exhibit flow features that block the radial inflow toward the typhoon center within the boundary layer. This negative vorticity likely inhibited axisymmetrization, thereby interfering with the mixing of mesovortices. We also noted that these negative vorticity patches tended to be transported outward simultaneously with mesovortex mixing. Furthermore, an analysis of vertical motion indicated that DX2 km produced stronger convective cells, presumably because lower resolution weakens entrainment, thus relatively enhancing convection compared to the LES. This resulted in more vigorous secondary circulation and facilitated mesovortex mixing, contributing to earlier RI onset in DX2 km.
Although there was no substantial difference in the maximum intensity between the LES and DX2 km simulations, we observed a discrepancy in the timing of RI. The LES tended to show a delayed start of RI. Analysis of vertical vorticity distributions revealed multiple mesovortices, about 10 km in diameter, forming prior to the onset of RI. These mesovortices are likely generated by horizontal shear instability, and energy production of non-axisymmetric components due to horizontal shear. As the mesovortices advected toward the typhoon center, they mixed and eventually dissipated. In the LES, however, they formed later and persisted longer than in DX2 km. Objective identification showed that a greater number of mesovortices appeared in the LES, and they moved more slowly toward the typhoon center compared to DX2 km. Composite analyses of their vertical structure revealed that these mesovortices exhibit flow features that block the radial inflow toward the typhoon center within the boundary layer. This negative vorticity likely inhibited axisymmetrization, thereby interfering with the mixing of mesovortices. We also noted that these negative vorticity patches tended to be transported outward simultaneously with mesovortex mixing. Furthermore, an analysis of vertical motion indicated that DX2 km produced stronger convective cells, presumably because lower resolution weakens entrainment, thus relatively enhancing convection compared to the LES. This resulted in more vigorous secondary circulation and facilitated mesovortex mixing, contributing to earlier RI onset in DX2 km.