Mapping susceptibility of rainfall-triggered shallow landslides using a probabilistic approach |
| |
| Authors: | Chia-Nan Liu Chia-Chen Wu |
| |
| Institution: | (1) Civil Engineering Department, National Chi Nan University, Nantou, 545, Taiwan |
| |
| Abstract: | To prepare a landslide susceptibility map is essential to identify hazardous regions, construct appropriate mitigation facilities,
and plan emergency measures for a region prone to landslides triggered by rainfall. The conventional mapping methods require
much information about past landslides records and contributing terrace and rainfall. They also rely heavily on the quantity
and quality of accessible information and subjectively of the map builder. This paper contributes to a systematic and quantitative
assessment of mapping landslide hazards over a region. Geographical Information System is implemented to retrieve relevant
parameters from data layers, including the spatial distribution of transient fluid pressures, which is estimated using the
TRIGRS program. The factor of safety of each pixel in the study region is calculated analytically. Monte Carlo simulation
of random variables is conducted to process the estimation of fluid pressure and factor of safety for multiple times. The
failure probability of each pixel is thus estimated. These procedures of mapping landslide potential are demonstrated in a
case history. The analysis results reveal a positive correlation between landslide probability and accumulated rainfall. This
approach gives simulation results compared to field records. The location and size of actual landslide are well predicted.
An explanation for some of the inconsistencies is also provided to emphasize the importance of site information on the accuracy
of mapping results. |
| |
| Keywords: | Landslide Rainfall Mapping Probability Fluid pressure |
| 本文献已被 SpringerLink 等数据库收录！ |
|
