← Back to Sessions

Dissecting Seismic Signals to Estimate Landslide Volume

Description: 

Determining the size of a landslide based solely on the seismic signal it produces could be a useful tool for rapid event characterization, response, and early-warning systems. In this study, we work towards establishing a general relationship between the seismic signal and the physical properties of large, rapid landslides that are detected by regional seismic networks. To do this, we analyze the seismic signals for known landslide events and determine which combination of seismic parameters best predict landslide characteristics of interest, primarily volume.

We analyze events from the Exotic Seismic Events Catalog (ESEC), a published catalog including information on the seismic detection and physical properties (e.g., volume, runout distance, and drop height) of 290 seismogenic landslides. This collection spans several continents, which provides us with a diverse dataset to work with. We focus on large (>105 m3) mass movements with fall, slide, and avalanche movement types because they are seismically detectable at regional scales and generate observable long-period (>20 s) signals that we can invert to obtain the force history of the landslide’s bulk motions. We analyze parameters in the high-frequency range (1-5 Hz) to represent the portion of seismic energy produced by smaller-scale processes, such as disrupted flow and granular interactions. We also include characteristics of the force histories inverted from the long-period signals which represent the bulk movement of the failure mass. We find that envelope rise time, envelope area, and maximum force scale logarithmically with volume, whereas peak frequency and maximum amplitude have a less-defined relationship. Using this information, we perform a multivariate linear regression to develop an empirical relationship between these parameters and independently estimated volumes (usually via remote sensing), as reported in the ESEC. We find that a combination of envelope rise time, envelope area, and maximum force, derived solely from seismic analysis, provide a reasonable estimate of landslide volume with well-defined uncertainties.

Session: Detecting, Characterizing and Monitoring Mass Movements - II

Type: Oral

Date: 5/2/2024

Presentation Time: 11:00 AM (local time)

Presenting Author: Elaine

Student Presenter: No

Invited Presentation: 

Authors