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

A. A. Saoulis 1,2, T.-S. Phạm 3, A. M. G. Ferreira 2 1 Department of Physics & Astronomy, University College London, Gower Street, London, WC1E 6BT, United Kingdom 2 Department of Earth Sciences, University College London, 5 Gower Place, London, WC1E 6BS, United Kingdom 3 Research School of Earth Sciences, The Australian National University, Canberra, ACT, Australia · 2026

Read Improving moment tensor solutions under Earth structure uncertainty with simulation-based inference by A. A. Saoulis 1,2, T.-S. Phạm 3, A. M. G. Ferreira 2 1 Department of Physics & Astronomy, University College London, Gower Street, London, WC1E 6BT, United Kingdom 2 Department of Earth Sciences, University College London, 5 Gower Place, London, WC1E 6BS, United Kingdom 3 Research School of Earth Sciences, The Australian National University, Canberra, ACT, Australia (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-18925.txt.

Introduction

Saoulis and co-authors improve earthquake source estimates when Earth's inner structure is not known for sure. Standard fits often assume a simple bell shape for model error. They show that this shape breaks down under small shifts in a layered Earth model. Their method uses simulation-based learning to model the errors as they truly fall. Two forms are offered, one guided by physics and one end-to-end learned. Tests show the old shape biases results and understates doubt, most at short periods.

Catalog record

CREATORA. A. Saoulis 1,2, T.-S. Phạm 3, A. M. G. Ferreira 2 1 Department of Physics & Astronomy, University College London, Gower Street, London, WC1E 6BT, United Kingdom 2 Department of Earth Sciences, University College London, 5 Gower Place, London, WC1E 6BS, United Kingdom 3 Research School of Earth Sciences, The Australian National University, Canberra, ACT, Australia
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE97634
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2603.18925

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Improving moment tensor solutions under Earth structure uncertainty with simulation-based inference A. A. Saoulis 1,2, T.-S. Phạm 3, A. M. G. Ferreira 2 1 Department of Physics & Astronomy, University College London, Gower Street, London, WC1E 6BT, United Kingdom 2 Department of Earth Sciences, University College London, 5 Gower Place, London, WC1E 6BS, United Kingdom 3 Research School of Earth Sciences, The Australian National University, Canberra, ACT, Australia Abstract \pagerange Improving moment tensor solutions under Earth structure uncertainty with simulation-based inference – References keywords Earthquake source observations – Waveform inversion – Bayesian inference – Inverse theory – Machine learning {summary} Bayesian inference represents a principled way to incorporate Earth structure uncertainty in full-waveform moment tensor inversions, but traditional approaches generally require significant approximations that risk biasing the resulting solutions. We introduce a robust method for handling theory errors using simulation-based inference (SBI), a machine learning approach that empirically models their impact on the observations. This framework retains the rigour of Bayesian inference while avoiding restrictive assumptions about the functional form of the uncertainties. We begin by demonstrating that the common Gaussian parametrisation of theory errors breaks down under minor ( 1-3\% ) 1-D Earth model uncertainty. To address this issue, we develop two formalisms for utilising SBI to improve the quality of the moment tensor solutions: one using physics-based insights into the theory errors, and another utilising an end-to-end deep learning algorithm. We then compare the results of moment tensor inversion with the standard Gaussian approach and SBI, and demonstrate that Gaussian assumptions induce bias and significantly under-report moment tensor uncertainties. We also show that these effects are particularly problematic when inverting short period data and for shallow, isotropic events. On the other hand, SBI produces more reliable, better calibrated posteriors of the earthquake source mechanism. Finally, we successfully apply our methodology to two well studied moderate magnitude earthquakes: one from the 1997 Long Valley Caldera volcanic earthquake sequence, and the 2020 Zagreb earthquake. 1 Introduction Focal mechanism solutions of full-waveform moment tensor inversion have been shown to rely on high-quality forward modelling, which primarily depends on accurate Earth structure models ( Hardebeck & Shearer 2002 ; Hjörleifsdóttir & Ekström 2010 ; Takemura et al. 2020 ; Rösler et al. 2021 ; Song et al. 2022 ; Simutė et al. 2023 , see Phạm et al. 2025 for a recent review). These effects are particularly pronounced in regions with strongly heterogeneous or poorly constrained crustal structure ( Abercrombie & Ekström 2001 ; Křížová et al. 2013 ; Petersen et al. 2021 ) . Reliable moment tensor solutions are also critical for downstream applications such as seismic tomography ( Blom et al. 2023 ) . Beyond characterising individual events, heterogeneities can lead to systematic biases in 1-D Earth model derived source catalogs ( Hingee et al. 2011 ; Hejrani et al. 2017 ; Rösler et al. 2024a ) . These inaccuracies can induce biases in, and specious trade-offs between, moment tensor components ( Dziewonski & Woodhouse 1983 ; Ferreira & Woodhouse 2006 ; Weston et al. 2011 ; Ferreira et al. 2011 ) . Such effects have been shown to cause, for instance, systematic spurious non-double-couple components in retrieved seismic source solutions for tectonic events ( Sawade et al. 2022 ; Rösler et al. 2024b ) . Moreover, unresolved Earth structure uncertainty continues to pose significant challenges for interpreting seismic source solutions ( Akuhara et al. 2025 , e.g.,) . An appealing solution is Bayesian inference, which provides a framework for treating theory or mismodelling errors directly in the uncertainty estimation procedure ( Tarantola 2005 ) . We assume the uncertain 1-D Earth model to be the primary source of theory errors in the context of this work. Including the effect of theory errors in source inversions can significantly improve inversion quality ( Yagi & Fukahata 2011 ; Minson et al. 2013 ; Minson et al. 2014 ; Duputel et al. 2015 ; Phạm & Tkalčić 2021 , e.g.,) . This is often achieved by treating theory errors as an additional term in a Gaussian likelihood function, following Tarantola 2005 . Several studies approximate the theory errors by sampling over the modelled Earth structure uncertainty empirically and estimating the resulting variability in the observations ( Duputel et al. 2012 ; Vasyura-Bathke et al. 2021 ; Phạm & Tkalčić 2021 ; Poppeliers & Preston 2021 ; Phạm et al. 2024 ) , while others rely on theoretical approximations for the effect of theory errors on the observation covariance structure ( Duputel et al. 2014 ; Hallo & Gallovič 2016 ) . All these prior works show that incorporating theory error contributions yields improved solutions over the naïve approach. However, they perform significant approximations to make the inverse problem tractable by traditional means. The question of whether the resulting solutions are genuinely accurate and trustworthy thus still remains. There are many other hybrid or empirically motivated approaches. Grid-search strategies have proven effective for tracing likelihood contours for uncertainty quantification ( Sokos & Zahradník 2013 ; Vackář et al. 2017 ; Zahradník & Sokos 2025 ; Thurin et al. 2025 ) , though likelihood contours often do not have a clear probabilistic interpretation. Additionally, making this approach tractable for holistically treating theory errors may prove challenging. Another family of approaches relies on building solution ensembles through hybrid techniques, including bootstrapping ( Wéber 2006 ; Valentine & Trampert 2012 ) and by injecting empirically motivated theory errors ( Poppeliers & Preston 2022 ) . A relevant example to this study is Stähler & Sigloch 2014 ; Stähler & Sigloch 2016 , which used earthquake source catalogues to build empirical likelihood functions that better account for mismodelling errors. While these approaches remain valuable, particularly as a means for dealing with unknown (or challenging to model) theory uncertainties, it is desirable to continue to develop a complementary approach that directly integrates our ability to model theory errors. Simulation-based inference (SBI) is a promising alternative to fully treat the effects of theory errors without compromising (or abandoning) a rigorous Bayesian approach to moment tensor inversion. SBI uses machine learning (ML) models to learn the observational uncertainties empirically through direct simulation, in principle allowing arbitrarily complex sources of uncertainty to be modelled without making parametric assumptions about their form ( Alsing et al. 2019 ; Cranmer et al. 2020 ; Zammit-Mangion et al. 2025 ) . Saoulis et al. 2025 demonstrated that standard Gaussian likelihood assumptions about full-waveform data errors yielded inaccurate moment tensor solutions. SBI, on the other hand, produced accurate results at a fraction of the computational cost. This work applies SBI to full-waveform moment tensor inversion while incorporating theory errors. SBI has received significant attention across the physical sciences for enabling fast and accurate Bayesian inference ( Gonçalves et al. 2020 ; Stockman et al. 2024 ; von Wietersheim-Kramsta et al. 2025 ; Atlas Collaboration 2025 ; Dax et al. 2025 ) . A key practical challenge in adapting SBI for a given domain is preparing the observational data to make it amenable to ML-based probabilistic modelling (see e.g. Deistler et al. 2025 , for a practical perspective) . This is typically straightforward in low dimensions, but high-dimensional observations, such as full-waveform multi-station seismic data, can

Read the complete text