Ground-Motion modeling as an image processing task
- We construct and examine the prototype of a deep learning-based ground-motion model (GMM) that is both fully data driven and nonergodic. We formulate ground-motion modeling as an image processing task, in which a specific type of neural network, the U-Net, relates continuous, horizontal maps of earthquake predictive parameters to sparse observations of a ground-motion intensity measure (IM). The processing of map-shaped data allows the natural incorporation of absolute earthquake source and observation site coordinates, and is, therefore, well suited to include site-, source-, and path-specific amplification effects in a nonergodic GMM. Data-driven interpolation of the IM between observation points is an inherent feature of the U-Net and requires no a priori assumptions. We evaluate our model using both a synthetic dataset and a subset of observations from the KiK-net strong motion network in the Kanto basin in Japan. We find that the U-Net model is capable of learning the magnitude???distance scaling, as well as site-, source-, andWe construct and examine the prototype of a deep learning-based ground-motion model (GMM) that is both fully data driven and nonergodic. We formulate ground-motion modeling as an image processing task, in which a specific type of neural network, the U-Net, relates continuous, horizontal maps of earthquake predictive parameters to sparse observations of a ground-motion intensity measure (IM). The processing of map-shaped data allows the natural incorporation of absolute earthquake source and observation site coordinates, and is, therefore, well suited to include site-, source-, and path-specific amplification effects in a nonergodic GMM. Data-driven interpolation of the IM between observation points is an inherent feature of the U-Net and requires no a priori assumptions. We evaluate our model using both a synthetic dataset and a subset of observations from the KiK-net strong motion network in the Kanto basin in Japan. We find that the U-Net model is capable of learning the magnitude???distance scaling, as well as site-, source-, and path-specific amplification effects from a strong motion dataset. The interpolation scheme is evaluated using a fivefold cross validation and is found to provide on average unbiased predictions. The magnitude???distance scaling as well as the site amplification of response spectral acceleration at a period of 1 s obtained for the Kanto basin are comparable to previous regional studies.…
Author details: | Henning LilienkampORCiDGND, Sebastian von SpechtORCiDGND, Graeme WeatherillORCiD, Giuseppe CaireORCiD, Fabrice CottonORCiDGND |
---|---|
DOI: | https://doi.org/10.1785/0120220008 |
ISSN: | 0037-1106 |
ISSN: | 1943-3573 |
Title of parent work (English): | Bulletin of the Seismological Society of America |
Subtitle (English): | introducing a neural network based, fully data-driven, and nonergodic |
Publisher: | Seismological Society of America |
Place of publishing: | Albany |
Publication type: | Article |
Language: | English |
Date of first publication: | 2022/03/07 |
Publication year: | 2022 |
Release date: | 2024/03/08 |
Volume: | 112 |
Issue: | 3 |
Number of pages: | 18 |
First page: | 1565 |
Last Page: | 1582 |
Funding institution: | Helmholtz Einstein International Berlin Research School in Data Science |
Organizational units: | Mathematisch-Naturwissenschaftliche Fakultät / Institut für Mathematik |
Mathematisch-Naturwissenschaftliche Fakultät / Institut für Geowissenschaften | |
DDC classification: | 5 Naturwissenschaften und Mathematik / 55 Geowissenschaften, Geologie |
Peer review: | Referiert |
External remark: | Erratum: https://doi.org/10.1785/0120220131 |