Adjoint-based uncertainty quantification for inhomogeneous friction on a slow-slipping fault

TL;DR

This paper introduces an adjoint-based data assimilation method to quantify the uncertainty of spatially inhomogeneous frictional features on slow-slip faults, enhancing seismic motion prediction and understanding of fault mechanics.

Contribution

It develops a novel variational data assimilation approach using a second-order adjoint method for high-resolution uncertainty quantification of fault frictional features.

Findings

Successfully quantified spatial uncertainty of frictional features.

Revealed correlation between slow-slip dynamics and frictional parameters.

Provides insights for improved seismic prediction models.

Abstract

Long-term slow-slip events (LSSEs) usually occur on the deep, shallow parts of subducting plates and have substantial relation with adjacent megathrust fault motion. Conventional techniques of quantifying slow earthquake frictional features show that these features may be indicative of predictive seismic motion; however, quantifying high-accuracy uncertainty of the frictional fields has not yet been achieved. We therefore propose a method of uncertainty quantification for spatially inhomogeneous frictional features from slip motion on an LSSE fault--megathrust fault complex in southwestern Japan. By combining a fault motion model that mimics slow-slip motion and a variational data assimilation (DA) technique using a second-order adjoint method, we have succeeded in quantifying the spatial distribution of the uncertainty of the frictional features. Further, evaluation of the spatial distribution in high-resolution reveals the correlation between the dynamics of the slow-slip motion and the important components of the frictional features, which is valuable information for observational DA design. Findings from this study are expected to advance the theoretical foundation of applied seismic motion prediction techniques using slow-slip frictional features as stress meters for megaquakes, as well as improve understanding of the relationship between the slow-slip motion and frictional parameters of a fault.