Data-Driven Ground Motion Synthesis using Deep Generative Models

Abstract:

Robust estimation of ground motions generated by scenario earthquakes is critical for many engineering applications. We leverage recent advances in Generative Adversarial Networks (GANs) to develop a new framework for synthesizing earthquake acceleration time histories from data. Our approach extends the Wasserstein GAN formulation to allow for the generation of ground-motions conditioned on a set of continuous physical variables. Our model is trained to approximate the intrinsic probability distribution of a massive set of strong-motion recordings from Japan. We show that the trained generator model can synthesize realistic 3-Component accelerograms conditioned on Magnitude, Distance, and Vs30 as a proxy for site response. Our model captures most of the relevant statistical features of the acceleration spectra and waveform envelopes. The output seismograms display clear P and S-wave arrivals with the appropriate energy content and relative timing of onsets. The synthesized Peak Ground Acceleration (PGA) estimates are also consistent with observations. We develop a set of metrics that allow us to assess the stability of the training process and tune model hyperparameters. We further show that the trained generator network can interpolate to regions where no earthquake ground motion recordings exist. Our approach allows the on-demand synthesis of seismograms for engineering applications, using a small set of input parameters, it is possible to synthesize a full range of potential ground motions, consistent with available data.