Estimating temporal probability for landslide initiation along transportation routes based on rainfall thresholds

详细信息    查看全文

• 作者：Pankaj Jaiswal ; Cees J. van Westen
• 关键词：Rainfall threshold ; Temporal probability ; Transportation route ; Nilgiri
• 刊名：Geomorphology
• 出版年：2009
• 出版时间：1 November 2009
• 年：2009
• 卷：112
• 期：1-2
• 页码：96-105
• 全文大小：1948 K

文摘

The estimation of the temporal probability of landslide initiation is an essential component in landslide hazard assessment. In this paper a temporal probability model is presented for the initiation of shallow translational debris slides and debris flows along cut slopes of a railroad sector in southern India, for which an extensive landslide database was available, covering a time span of 15 years. The model is based on rainfall thresholds and gives the likelihood of occurrence of rainfall that can trigger landslides with a certain density. The temporal probability was calculated as the joint probability of annual exceedance probability of the rainfall threshold, determined using a Poisson probability model and the probability of landslide occurrence once the threshold had been exceeded. The model was tested for a 19-km long railroad alignment in the Nilgiri hills, which was divided into a number of sections on the basis of terrain characteristics. A landslide inventory, containing dates of occurrence, was prepared from historical records for the period 1987 to 2007. Daily rainfall for the same period was collected from 15 rain gauges. Rainfall thresholds were established for the sections based on the relationship between daily and antecedent rainfalls. Four thresholds were defined for rainfall events that can trigger one or more landslides within each section and one threshold that can trigger 15 or more landslides along the entire route. The annual temporal probability varies from 0.27 to 0.49. The model was also found useful in predicting landslides in a nearby road with similar characteristics. The result indicates that more than 60 % of the recorded landslides along the road occurred within the sections with high temporal probability values (> 0.40). The temporal probability derived from the model forms the basis for future landslide risk analysis along the transportation routes.