← Back to Sessions

Incorporating Numerical Landslide Models Into Broadband Synthetic Seismogram Simulations of Large, Rapid Landslides

Description: 

The ability to generate broadband synthetic seismograms of large, rapid landslides would be useful for forensic analysis of real events, and to generate test data of landslide scenarios for detection and early warning systems. Depth-averaged numerical landslide models are increasingly being used to model and try to match observed seismic data, but only for the simplest, lowest frequency (less than ~0.1 Hz) part of the spectrum. Depth-averaged models do not explicitly model the higher-frequency source processes (e.g., individual impacts), yet most of the energy in a landslide seismic signal is at higher frequencies. In this study, we establish a quantitative way to empirically model the higher frequency energy by examining the relationship between radiated higher frequency seismic energy from real landslides, and flow parameters from numerical landslide models. We do this using two depth-averaged numerical landslide models: SHALTOP, unique in explicitly outputting the entire basal traction field, and D-Claw, which can model two-phase flows (water and solids). We calibrate the models to the runout and the force-time history derived from long-period seismograms of well-characterized landslides from the Exotic Seismic Event Catalog (ESEC), focusing primarily on the 2019 11 million m³ Iliamna, AK ice and rock avalanche, and the 2009 48.5 million m³ Mount Meager, BC landslide. Once we establish a good model fit, we examine the relationship between the observed intermediate and high frequency (> 0.1 Hz) wavefield and spatiotemporally varying bulk flow characteristics that each model can produce (e.g., the distribution of basal tractions, frictional work rate, momentum, and inertial number). We then examine these findings in context of a greater set of empirical data of large, rapid landslides from the ESEC to propose a general approach for relating the higher frequency part of the spectrum to bulk landslide dynamics and characteristics.

Session: Detecting, Locating, Characterizing and Monitoring Non-earthquake Seismoacoustic Sources

Type: Oral

Date: 4/19/2023

Presentation Time: 08:30 AM (local time)

Presenting Author: Kate Allstadt

Student Presenter: No

Invited Presentation: 

Authors