A. A. Saoulis

Improving moment tensor solutions under Earth structure uncertainty with simulation-based inference

A. A. Saoulis, T. -S. Pham, A. M. G. Ferreira

This paper addresses a fundamental problem in earthquake analysis: traditional methods for determining earthquake characteristics (moment tensor solutions) make oversimplified assumptions about measurement uncertainties caused by incomplete knowledge of Earth's structure. The researchers introduce a machine learning technique called simulation-based inference that can better account for these complex, non-linear uncertainties without making restrictive assumptions, leading to more accurate and reliable earthquake source estimates.