Accounting for uncertain fault geometry in earthquake source inversion

Abstract : Inferences of the history of fault slip are subject to biases induced by unavoidable imperfections in the assumed forward model. For example, we commonly simplify assumed model of crustal properties and the geometry of the fault. The impacts of these choices are rarely investigated or quantified. Here, we explore the impact of uncertainties related to the choice of a fault geometry. To do so, we develop an augmented misfit covariance matrix which approximates the uncertainty related to the choice of a given fault geometry, following a previously implemented method exploring the impact of uncertainties on the elastic properties of our models. We validate this approach with the simplified case of a fault that extends infinitely along strike, investigating the impact of uncertainty in fault dip and location. We apply our methodology to the 2016 Mw 6.2 Amatrice earthquake, Central Italy. These different tests show that introducing uncertainties in fault geometry in the static inversion results in more sensible slip models. In practice, this augmented misfit covariance matrix reduces the confidence in the data points which are more sensitive to geometrical uncertainties as well as allowing for correlated misfits that are expected from the use of imperfect forward models. For most events, the uncertainties in both fault geometry and crustal structure will have a significant impact on the retrieved models, but the effect is expected to be stronger for large earthquakes (M>7) as epistemic uncertainties tend to scale with the amplitude of slip.
Complete list of metadatas
Contributor : Publications Géoazur <>
Submitted on : Friday, November 8, 2019 - 3:22:36 PM
Last modification on : Saturday, November 9, 2019 - 1:51:35 AM


  • HAL Id : hal-02356067, version 1



T. Ragon, A. Sladen, M. Simons. Accounting for uncertain fault geometry in earthquake source inversion. American Geophysical Union, Dec 2018, Washington, United States. pp.S41A-01. ⟨hal-02356067⟩



Record views