Probabilistic life-cycle landslide assessment subjected to nonstationary rainfall based on alternating stochastic renewal process
Zhengying He, Mitsuyoshi Akiyama, Abdul Kadir Alhamid, Dan M. Frangopol, Yu Huang
Abstract
The increasing intensity and frequency of rainfall due to climate change poses a significant risk of landslides in the future. Therefore, a methodology that accounts for the nonstationary effects of rainfall is needed to accurately assess future landslides. This study presents a novel framework for probabilistic life-cycle landslide assessment under nonstationary rainfall due to climate change based on an alternating stochastic renewal process. The alternating stochastic renewal process is developed to evaluate the distribution of maximum rainfall within the life-cycle of the slope. A slope fragility assessment is carried out by employing the uncertainties associated with the soil properties to the seepage-stability analysis based on Monte Carlo simulation. Finally, the life-cycle landslide probability under nonstationary rainfall due to climate change is estimated by convolving the rainfall hazard with the slope fragility curves based on the total probability theorem. The proposed framework is applied to two municipalities in Japan, namely Hiroshima and Kobe cities. The results emphasize that the probability of landslides increases significantly when nonstationary climate change effects are considered, highlighting the critical importance of incorporating climate change effects in landslide assessments.