Copula-based IDF curves and empirical rainfall thresholds for flash floods and rainfall-induced landslides

摘要

Floods, landslides and debris flows are natural events that occur all over the world and are often induced by extreme rainfall conditions. Several extreme events occurred in Slovenia (Europe) in the last 25 years that caused 18 casualties and approximately 500 million Euros of economic loss. The intensity-duration frequency (IDF) relationship was constructed using the Frank copula function for several rainfall stations using high-resolution rainfall data with an average subsample length of 34 years. The empirical rainfall threshold curves were also evaluated for selected extreme events. Post-event analyses showed that rainfall characteristics triggering flash floods and landslides are different. The sensitivity analysis results indicate that the inter-event time definition (IETD) and subsample definition methodology can have a significant influence on the position of rainfall events in the intensity-duration space, the constructed IDF curves and on the relationship between the empirical rainfall threshold curves and the IDF curves constructed using the copula approach. Furthermore, a combination of several empirical rainfall thresholds with an appropriate high-density rainfall measurement network can be used as part of the early warning system of the initiation of landslides and debris flows. However, different rainfall threshold curves should be used for lowland and mountainous areas in Slovenia. (C) 2016 Elsevier B.V. All rights reserved.