This research investigates the projected intensification of tropical cyclone (TC) precipitation and associated landslide risks in South China under global warming. Using deep learning to enhance cloud-resolving simulations, the study finds that TC rainfall intensification surpasses the Clausius-Clapeyron (CC) scaling, with precipitation accumulation over city-sized areas potentially surging by 18% per Kelvin of warming. This super-CC intensification is attributed to changes in TC structure, involving the expansion of deep convective cores and suppression of shallow convection. Analysis using Hong Kong's slope data suggests a potential 215% increase in landslide risks with a 4-K warming, highlighting the amplified threats from compound disasters under climate change.

Xiaoming Shi, Yang Liu, Jianan Chen, Haoming Chen, Yueya Wang, Zhongming Lu, Ruo-Qian Wang, Jimmy C.-H. Fung, Charles W. W. Ng