Establishing a threshold for rainfall‐induced landslides by a kinetic energy–duration relationship

Abstract

Many investigators have attempted to define the threshold of landslide failure, that is, the level of the selected climatic variable above which a rainfall‐induced landslide occurs. Intensity–duration (I–d) relationships are the most common type of empirical thresholds proposed in the literature for predicting landslide occurrence induced by rainfall. Recent studies propose the use of the kinetic power per unit volume of rainfall (J m−2 mm−1) to quantify the threshold of landslides induced by rainfall. In this paper, the relationship between rainfall duration and kinetic power corresponding to landslides triggered by rain was used to propose a new approach to define the threshold for predicting landslide occurrence. In particular, for the first time, a kinetic power per unit volume of rainfall–duration relationship is proposed for defining the minimum threshold needed for landslide failure. This new method can be applied using commonly used relationship for estimating the kinetic power per unit volume of rainfall and a new equation based on the measured raindrop size distribution. The applicability of this last method was tested using the data of rainfall intensity, duration and median volume diameter for 51 landslides in Taiwan. For the 51 landslides, the comparison between the measured pairs' kinetic power–duration and all selected relationships demonstrated that the equation based on the measured raindrop size distribution is the best method to define the landslide occurrence threshold, as it is both a process‐oriented approach and is characterized by the best statistical performance. This last method has also the advantage to allow the forecasting of landslide hazard before the end of the rainfall event, since the rainfall kinetic power threshold value can be exceeded for a time interval less than the event duration.

Advances in a measurement method of rainfall kinetic power and momentum affecting soil erosion processes

“…According to Carollo, Ferro, et al (2018), kinetic power and momentum, both depending on drop size distribution (DSD) and terminal velocity (Epema & Riezebos, 1983; Gilley & Finker, 1985), are related each other and are able to represent the rainfall erosivity in Mediterranean areas. Few authors (Carollo, Serio, et al, 2018; Lim et al, 2015; Park et al, 1983; Sanchez‐Moreno et al, 2012) proposed relationships to estimate rainfall momentum by rainfall intensity, while most of authors suggested P n – I relationships (Ellison, 1944; Van Dijk et al, 2002; Nanko et al, 2008; Carollo & Ferro, 2015; Carollo et al, 2016; Nanko et al, 2016; Carollo et al, 2017; Serio et al, 2019b, Ferro et al, 2020; Johannsen et al, 2020). All these relationships use rainfall intensity as a predictor, which is easily and widely measured by the recording rain gauge networks (Serio et al, 2019a, 2019b).…”

Optical Precipitation Gauge in the Rainfall Kinetic Energy Measurements

“…Information about kinetic energy stored in falling raindrops from high altitudes is important in solving a number of scientific, practical and economic tasks. The most important information about the kinetic energy of precipitation is for the agro-industrial sector of the economy, since such energy causes not only soil erosion, but also landslides [1,2]. High values of kinetic energy brought by liquid atmospheric precipitation can provoke the occurrence of landslides [1].…”

“…High values of kinetic energy brought by liquid atmospheric precipitation can provoke the occurrence of landslides [1]. The advantage of predicting landslide hazards based on energy flow control is that the probability of a landslide may occur before the end of high-intensity precipitation, since the threshold value of kinetic energy transmitted to the soil by drops can be exceeded in a time interval shorter than the duration of rain [2].…”

Time-Domain Correlation Quantitative Analysis Method of Regional Rainfall-Landslide Displacement Responses Based on a Time-Domain Correlation Model

“…In view of these limitations, the combined use of geomechanics, physical mechanics, and numerical analysis models to evaluate the stability of landslides during rainfall events can help predict the development trend of landslides, but the corresponding workload is often large, leading to delays in warnings. As an alternative approach, the monitoring data of landslide deformation characteristics and the rainfall factor, which is the main influence on the landslide process, often have obvious correlations (Ferro et al 2020, Yin et al 2016, De-ying et al 2019. Therefore, it is feasible to predict the risk of landslides by using widely available rainfall monitoring data.…”