Multi-Scale Interaction and Predictability of the Tropical Cyclone Rapid Intensification

Atmospheric deep moist convection has emerged as a highly complex subject matter for numerical weather prediction, due to its chaotic development and the multi-scale physical interactions involved. The consequences of individual convective storm itself can be devastating to society, but convective systems play a crucial role in the formation of organized severe weather events, such as the rapid intensification of tropical cyclones. In this study, we examined the impact of moist convective activity on the rapid intensification of tropical cyclones through a series of sensitivity and ensemble forecast experiments.

Our recent investigation revealed that meso-α (2000-200 km) and meso-β (200-20 km) scale initial features can help to restrict the general location of convective systems, thus enhancing convective activity with a few hours of lead time. However, meso-γ (20-2 km) or even smaller scale features with less than 30-minute lead time were found to be crucial in determining the exact timing and location of individual convection. To evaluate the effects of such individual convection on the intensification process of tropical cyclones, we employed the ensemble Kalman filter analysis with the assimilation of GOES-16 all-sky radiances, which can accurately capture cloud distribution. Through sets of sensitivity experiments aimed at modifying specific convective activity, using Hurricane Harvey (2017) as a test case, we discovered that the intensification of tropical cyclones is particularly sensitive to the moisture fields at specific times and locations. This points to the existence of a "sweet spot" of moisture field in terms of time, location, and scale, which influences convective activity and subsequent TC intensification. These results will have implications for the development of future observation and data assimilation systems, to effectively constrain the predictions of rapidly intensifying tropical cyclones.