Predicting debris-flow clusters under extreme rainstorms: a case study on Hong Kong Island

Abstract

Debris flows can cause severe loss of human lives and damage to property, especially on densely populated hilly terrains. In the changing climate, the frequency of debris flows is on a rising trend. Therefore, it is important to forecast possible scenarios of debris flows under extreme weather conditions. Previous numerical studies often deal with one individual debris flow in one analysis. Yet a large number of debris flows can occur in a large storm. This paper presents a physically based model to predict likely debris flow clusters on Hong Kong Island with an area of approximately 80 km2 considering the influence of the changing climate. Firstly, a slope stability analysis is conducted, and unstable cells and landslide deposition areas are predicted. Then clusters of debris flows initiating from these landslides are simulated considering hillslope erosion. The models are validated with historical debris flows triggered by a rainstorm in 2008. Finally, debris flow clusters under three reference extreme rainstorms (i.e. 44%, 65% and 85% of the 24-h probable maximum precipitation, PMP) are predicted. With the increase of rainstorm magnitude, numerous debris flows can occur simultaneously and merge, posing much greater threat to society. The consequences of debris flows grow dramatically when the magnitude reaches a certain extent, i.e. 65% of the 24-h PMP.