Kilometer-wide volumetric deformation of the shallow crust associated with strike-slip continental earthquakes and its relation with coseismic shallow slip deficits

Abstract

<p>Surface deformation associated with continental earthquakes separates into a localized component occurring on faults, and distributed deformation affecting the surrounding medium, referred to as off-fault deformation (OFD). This OFD includes both displacement discrete on secondary faults and cracks, and more diffuse deformation affecting the bulk volume of the crust. Although the deformation occurring on faults and cracks can be observed and measured in the field, diffuse deformation is more challenging to detect because it generates kilometer-scale continuous gradients of displacement without any visible disruption of the ground surface. Consequently, surface displacements measured in the field generally underestimate the total surface displacement of the earthquakes. Moreover, results from inversions of geodetic and/or seismic data suggest that, for many earthquakes, the amount of coseismic slip occurring at depth (3-7 km) is larger than what occurs in the shallower part (<3 km). This is referred to as the shallow slip deficit (SSD). So far, because diffuse deformation is not explicitly considered in earthquake displacement budgets, and because the origin of the SSD remains debated, it is difficult to directly compare directly surface observations with modeling results. In this study, we use a set of complementary geodetic data (InSAR, GPS, high-resolution optical data) to jointly invert for the coseismic slip of the 2019 Ridgecrest earthquake sequence in Southern California (M<sub>w</sub>6.4 and 7.1). To reproduce the rupture complexity observed in the high-resolution optical data, we use a complex fault model with increased resolution in the uppermost crust. We pay special attention that our preferred model fits both with the fault slip distribution observed at the surface in the high-resolution optical imagery data, and regional-scale displacement data from InSAR and GPS. In our best model, we estimate a 30% SSD in the upper 3 km. This value of 30% matches the amount of diffuse deformation we measured around the ruptures at the surface directly on the high-resolution optical data. From these observations, we propose that SSD is entirely balanced by the volumetric diffuse deformation, and more generally, that diffuse surface deformation is proportional to SSD. Finally, based on a compilation of published data, we show that SSD and diffuse deformation are both inversely proportional to the earthquake magnitude. Indeed, for large magnitude earthquakes, SSD and diffuse deformation are close to 0%. Conversely, for earthquakes that do not break the surface, diffuse deformation might be close to 100%. However, in this latter case, it still needs to be determined whether the diffuse deformation is only elastic, or not.</p>

Impact of optical imagery and topography data resolution on the measurement of surface fault displacement using sub-pixel image correlation

“…Field data report sparse surface ruptures with up to 2.6-2.9 m of horizontal displacement detected locally along primary fault strands (Pan et al, 2022;Ren et al, 2022Ren et al, , 2021Xie et al, 2022;Yuan et al, 2022). The total horizontal surface displacement measured from OIC is 2.27-2.35 m, and occurs over a FZW of 30 m to 2.15 km (Antoine, 2024;Fig. 1).…”

“…Continental earthquakes often generate surface displacements that can be directly observed using satellite imagery. Measuring the amount and spatial distribution of the earthquake surface displacement then allows for untangling the shallow Fault Zone (FZ) coseismic deformation processes (Antoine et al, 2024(Antoine et al, , 2022(Antoine et al, , 2021Barnhart et al, 2020;Li et al, 2023;Milliner et al, 2021;Zinke et al, 2019), bringing insights into the FZ mechanical behavior. Surface displacement measurements also enable constraining the earthquake rupture processes at depth using numerical inversion methods (e.g., Fialko et al, 2001;Jin and Fialko, 2021;Jolivet et al, 2014;Ragon and Simons, 2020;Segall, 2010;Simons et al, 2002;Wang et al, 2020).…”

“…The broad scale coseismic deformation, of several tens of kilometers wide, primarily arises from the elastic response of the crust to the deep earthquake rupture (Avouac, 2015;Segall, 2010). The deformation that occurred on the faults and in their proximate vicinity, at a spatial scale of ~1-100 m, often leads to inelastic and permanent deformation of the crust which can take place through a combination of slip along primary faults, distributed deformation along secondary faults and/or fractures (e.g., DuRoss et al, 2020;Klinger et al, 2005;Rockwell et al, 2002;Teran et al, 2015;Yuan et al, 2022), and diffuse deformation of the surrounding medium (Antoine et al, 2024(Antoine et al, , 2022(Antoine et al, , 2021. The regions of surface deformation on-and near the fault correspond to what is referred to as the fault zone (FZ).…”

“…OIC (e.g., Aati et al, 2022;Leprince et al, 2007;Rosu et al, 2015), and differential LiDAR (e.g., Borsa and Minster, 2012;Nissen et al, 2014;Scott et al, 2018) techniques, on the other hand, allow for three-dimensional (3-D), spatially continuous, and possibly high-resolution (submeterscale) measurement of the earthquake surface displacement field, even in regions of highdisplacement gradients. Between the two techniques, OIC is more commonly used (e.g., Antoine et al, 2024Antoine et al, , 2022Antoine et al, , 2021Delorme et al, 2020;Milliner et al, 2021Milliner et al, , 2016Teran et al, 2015;Zinke et al, 2019) as satellite optical data cover wide regions (> 15). LiDAR is generally limited to small study area because of the high-cost associated with LiDAR data acquisitions.…”

“…1a). High-resolution imagery (HR; <3 m; e.g., Pleiades, WorldView, SPOT5-7) was though shown to be a more reliable source of information, especially for mapping the fault ruptures (e.g., Klinger et al, 2005;Reitman et al, 2023), and assessing the different components of the near-fault surface deformation (e.g., Antoine, 2024Antoine, , 2021Zhou et al, 2018;Fig. 1b) and the fault zone width (FZW; Ajorlou et al, 2021) using OIC methods.…”