I attended and presented our recent developments in the 2D DEM dataset at the 8th International Conference of Earthquake Geotechnical Engineering in Osaka, Japan.
Our short format paper is published open-access here: Chiama, K., Bednarz, W., Moss, R., Shaw, J.H. (2024). Identification and analysis of ground surface rupture patterns from thrust and reverse fault earthquakes using geomechanical models. Japanese Geotechnical Society Special Publication, 2024, Volume 10, Issue 26, Pages 978-983, Released on J-STAGE June 17, 2024, Online ISSN 2188-8027, https://doi.org/10.3208/jgssp.v10.OS-15-06.
Summary and Implications:
We show how to use the machine learning based computer vision (CV) model to take measurements of the gorund surface deformation and how the model classifies each of the scarp types for each increment of slip in the DEM models. We trained the CV model on the 45 DEM experiments presented in Chiama et al. (2023) and the CV model increases the resolution, frequency, and precision of obtaining surface measurements.
I presented our new large suite of 2D DEM model results at the Annual Seismological Society of America (SSA) Conference in Anchorage, Alaska!
We performed ~3,500 2D DEM experiments to evaluate the influence of the fault dip, sediment depth, sediment density, and sediment strength on the resultant ground surface deformation observed in thrust and reverse fault earthquakes. We used a machine learning model based on computer vision to obtain measurements of the ground surface characteristics every 5 cm of slip up to 5 m of total slip in the model (~ Mag. 7.7 earthquake). We obtained ~350,000 measurements of the scarp height, width, and dip. The model code and full dataset are available on DesignSafe.
I was awarded one of the best student presentations at SSA! https://www.seismosoc.org/awards/student-awards/
I attended the Annual 2024 Seismological Society of America (SSA) Conference's field trip to the Prince Willaim Sound in Alaska! This was a wonderful field trip and I learned so much about the 1964 Earthquake as well as the Barry Glacier landslide!
Statewide California Earthquake Center - 2023 Annual Meeting
I presented our new, 2D DEM model that replicates the Dolan et al. (1996) trench of the Cucamonga Fault.
Implications:
DEM models offer a deterministic way to replicate trench-scale observations with realistic faulting and earthquake source parameters.
We can use the insights of DEM modeling to help guide the inference of earthquake source parameters from paleoseismic trench observations.
We can test paleoseismic trench restoration methods by restoring DEM models.
Please send us interesting trenches!
My first paper of my PhD was published in BSSA! It focuses on applying 2D DEM models to reconstruct the deformation observed in large magnitude thrust and reverse fault earthquakes.
Check it out:
Chiama, K., Chauvin, B., Plesch, A., Moss, R., Shaw J. H. (2023) Geomechanical modeling of ground surface deformation associated with thrust and reverse‐fault earthquakes: A distinct element approach. Bulletin of the Seismological Society of America; 113(4), 1702–1723. doi: https://doi.org/10.1785/0120220264.