Joint estimate of the co-seismic 2011 Tohoku earthquake fault slip and post-seismic viscoelastic relaxation by GRACE data inversion

Abstract

SUMMARY Satellite-derived gravity data offer a novel perspective for understanding the physics of megathrust earthquakes at subduction zones. Nonetheless, their temporal resolution and observational errors make it difficult to discern the different phases of the seismic cycle, as the elastostatic deformation (co-seismic) and the stress relaxation by viscous flow (post-seismic). To overcome these difficulties and to take advantage of the physical constraints on the temporal evolution and on the spatial pattern of the earthquake-induced gravity disturbances, we have jointly estimated the fault slip of the 2011 Tohoku earthquake and the rheological stratification by means of a Bayesian inversion of GRACE data time series and within the framework of spherically symmetric self-gravitating compressible viscoelastic Earth models. This approach, in addition to improve the exploitation of satellite-derived gravity data, allows us (i) to constrain the fault slip taking advantage of information from both the co- and post-seismic signatures and (ii) to investigate the trade-off between the fault slip and the shallow rheological stratification. In this respect, it can be used to improve the modelling of crustal displacements from GPS data, even if their higher accuracy and temporal resolution allow to discriminate well the co-seismic signature from the others.

On Earthquake Detectability by the Next-Generation Gravity Mission

“…9, but after Gaussian filtering with filter radius 2 • to the long (rather than short) wavelengths of the gravity field, the long wavelengths being more accurate than the short ones. In particular, as already shown by Cambiotti (2020), even the less accurate GRACE data, when supplemented by their covariance matrix, can provide physically sound information on the co-seismic slip of the 2011 Tohoku earthquake, without the need of any spatial filtering, the latter being necessary only to limit the trade-off between the shallow rheological stratification and the trends due to geophysical process other than the earthquake.…”

“…Indeed, the possibility of its detection cannot be simply excluded on the basis of an analysis of the signal-to-noise ratio. Rather, as we are going to discuss in the following sections and as already shown by Cambiotti (2020) for the case of the 2011 Tohoku earthquake using GRACE data, the earthquake signature can become detectable by looking for the specific spatio-temporal pattern of the earthquake signature in the whole data time series. Thus, the more data we have, the smaller signals can be recognized, even smaller than the observational errors.…”

“…Such a physico-mathematical model has been already applied successfully for investigating the earthquake signatures at the long wavelengths observed by GRACE and GOCE missions (Cambiotti and Sabadini 2013;Fuchs et al 2016;Cambiotti 2020) and, so, it has been judged as adequate for the modelling of the earthquake signatures. As for the updated ESA-ESM, the calculations have been performed for the years 1995 until 2006 with a temporal resolution of 6 hours and up to SH degree and order 180.…”

“…In the present work, after having evaluated the expected performance provided by the NGGM and after having designed a simulator for generating NGGM synthetic data during a mission lifetime of almost 12 years, we discuss the detectability of the co-and postseismic gravity signatures associated with elastostatic deformation and stress relaxation by viscous flow caused by earthquakes. Real gravity data from GRACE and GOCE missions have already been exploited for studying mass rearrangement caused by megathrust earthquakes, such as the 2004 Sumatra and 2011 Tohoku earthquakes (De Linage et al 2009;Cambiotti and Sabadini 2013;Fuchs et al 2016;Cambiotti 2020) and, in some cases, earthquakes of magnitude as low as M W = 8.3 have been detected as well (Chao and Liau 2019). Hereinafter, in order to investigate the potential of the NGGM for seismological purposes, we assess the lowest earthquake magnitude threshold detectable by this new satellite mission by considering earthquakes of magnitude as low as M W = 7 , comparable with the 1980 Irpinia intraplate earthquake.…”

Geophysical Challenges for Future Satellite Gravity Missions: Assessing the Impact of MOCASS Mission

“…For instance, water movements in wide inland basins, as the Amazonas (Han et al, 2010), as well as over the oceans (Jeon et al, 2018) and glaciers (Zemp et al, 2019) were successfully detected and studied from the time variation of GRACE data: the analysis revealed the presence of both periodic annual components and long term gravity trends connected to the underground water flow (Döll et al, 2014). GRACE and GOCE satellites were also able to observe the gravity field variations due to the co-seismic and post-seismic deformations of large earthquakes, such as the Sumatra-Andaman (Han, 2006) earthquake or the Tohoku event (Cambiotti, 2020;Han et al, 2015). Plurennial trends in the gravity field variations have been recorded and studied in the Himalaya region (Braitenberg & Shum, 2017;Matsuo & Heki, 2010;Sun et al, 2009) and interpreted as due to the combination of hydrologic, isostatic and tectonic effects.…”