Displays

Earthquake rupture on splay faults in subduction zones could pose a significant tsunami hazard, as they could accommodate more vertical displacement and are situated closer to the coast. To better understand this tsunami hazard, we model splay fault rupture dynamics and tsunami propagation and inundation constrained by a geodynamic seismic cycle (SC) model; building on work presented in Van Zelst et al. (2019). This two-dimensional modelling framework considers geodynamics, seismic cycles, dynamic ruptures, and tsunamis together for the first time. The SC model provides six blind splay fault geometries, self-consistent stress and strength conditions, and heterogeneous material properties in the domain. We find that all six splay faults are activated when the megathrust ruptures. The largest splay fault closest to the nucleation region ruptures immediately when the main rupture front passes the branching point. The other splay faults are activated through dynamic stress transfer from the main megathrust rupture or reflected waves from the surface. Splay fault rupture results in distinct peaks in the vertical surface displacements with a smaller wavelength and larger amplitudes. The effect of the vertical surface displacements also translates into the resulting tsunami, which consists of one large wave for the megathrust-only model and seven waves for the model including splay faults. Here, six of the waves can be attributed to the splay faults and the seventh wave results from the shallow tip of the megathrust. The waves from the rupture including splay faults have larger amplitudes and result in two episodes of coastal flooding. The first episode is due to the large wave caused by rupture on the largest splay fault nearest to the coast. The second flooding episode results from the combination and interference of the waves caused by the rest of the splay faults and the shallow megathrust tip. In contrast, the tsunami caused by rupture on only the megathrust has only one episode of flooding. Our results suggest that larger-than-expected tsunamis could be attributed to rupture on large splay faults. When multiple smaller splay faults rupture their effect on the tsunami might be hard to distinguish from a pure megathrust rupture. Considering the significant effects splay fault rupture can have on a tsunami, it is important to understand splay fault activation and to consider them in hazard assessment.

References:

Van Zelst, I., Wollherr, S., Madden, E. H. , Gabriel, A.-A., and Van Dinther, Y. (2019). Modeling megathrust earthquakes across scales: one-way coupling from geodynamics and seismic cycles to dynamic rupture. Journal of Geophysical Research: Solid Earth, 124, https://doi.org/10.1029/2019JB017539

How to cite: van Zelst, I., Rannabauer, L., Gabriel, A.-A., and van Dinther, Y.: The effect of multiple splay fault rupture on tsunamis, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-1379, https://doi.org/10.5194/egusphere-egu2020-1379, 2020.

Great earthquakes repeatedly occurred with different rupture processes in the Nankai trough, southwestern Japan. The 1944 Tonankai and the 1946 Nankai earthquakes (M ~8) caused serious tsunami damage over many areas along the coastline. The greatest earthquake in this region is the 1707 Hoei earthquake (M 8.4) that is believed to have ruptured the whole region (~600 km) of the Nankai Trough. The purpose of this study is to theoretically assess the tsunami height along the coasts excited by great earthquakes that can possibly occur in future in this region and simulate observable tsunami records during the earthquakes.

This study employed a new method for making various rupture scenarios. Based on a shear-stress distribution along the plate boundary estimated by the GNSS data analyses (Noda et al. 2018 JGR), we calculated coseismic slip distributions to release the accumulated stress for possible multi-segment rupture scenarios. Then, we used the strain energy released by the rupture to evaluate the possibility of each event. The released strain energy should be larger than the energy dissipated on the fault. However, for some scenarios, the released strain energy was smaller than the dissipated energy under the assumptions of friction laws. Such rupture scenarios are not likely to occur in the viewpoint of earthquake mechanics. This approach can provide necessary conditions of the strain energy or the accumulated stress levels for multi-segment rupture processes, while methods based on empirical or kinematic approaches do not treat stress or interseimsmic stress-accumulation periods required for ruptures.

Another distinctive point in our approach is that we theoretically synthesize ocean-bottom pressure changes caused by both seismic waves and tsunamis using a simulation method based on elastic and fluid dynamics (Saito and Tsushima 2016 JGR; Saito et al. 2019 Tectonophysics). Seismic wave contributions to ocean-bottom pressure changes are critically important for the synthetics in near-field or inside rupture areas because the seismic waves overlap with tsunami signals and work as noise for real-time tsunami monitoring. The records simulated in this study can be used to examine the monitoring ability of a deep-ocean observation network for megathrust earthquakes and tsunamis in this region.

How to cite: Saito, T. and Noda, A.: Tsunami scenarios for megathrust earthquakes: mechanics of multi-segment ruptures and elastic-fluid dynamics of wave propagation , EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-12273, https://doi.org/10.5194/egusphere-egu2020-12273, 2020.

The 28 September 2018 Mw 7.5 Sulawesi strike-slip earthquake generated an unexpected tsunami with devastating consequences. Since such strike-slip earthquakes are not expected to generate large tsunamis, the latter’s origin remains much debated. A key notable feature of this earthquake is that it ruptured at supershear speed, i.e., with a rupture speed greater than the shear wave speed of the host medium. Dunham and Bhat (2008) showed that such supershear ruptures, in half-space, produce two shock fronts (or Mach fronts) corresponding to an exceedance of shear and Rayleigh wave speeds. The Rayleigh Mach front carries significant vertical velocity along its front. We couple the ground motion produced by such a supershear earthquake to a 1D non-linear shallow water wave equation that accounts for both the time-dependent bathymetric displacement as well its velocity. We use an extension of Fourier-based PDE solvers called the Fourier Continuation (FC) method to numerically solve the system. The FC method enables high-order convergence of Fourier series approximations of non-periodic functions by resolving the well-known Gibbs “ringing” effect.  FC-based solvers offer limited numerical dispersion, high-order accuracy and mild CFL conditions—making them ideal to solve this system. Using the local bathymetric profile of Palu bay around the Pantoloan harbor tidal gauge, we have been able to clearly reproduce the observed tsunami with minimal tuning of parameters. We conclude that the Rayleigh Mach front, generated by a supershear earthquake combined with the Palu bay geometry, caused the tsunami.

How to cite: Amlani, F. and Bhat, H.: Tsunami Generation due to Supershear Earthquakes: A Case Study, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-6939, https://doi.org/10.5194/egusphere-egu2020-6939, 2020.

Seismological data provide compelling evidence of a depth-dependent rupture behavior of megathrust earthquakes. Relative to deeper events of similar magnitude, shallow earthquake ruptures have larger slip and longer duration, radiate energy that is depleted in high frequencies and have a larger discrepancy between their surface wave and moment magnitudes (M_S and M_W, respectively). These source properties make them prone to generating devastating tsunamis without clear warning signs. The origin of the observed differences has been a long-lasting matter of debate. Here we first show that the overall depth trends of all these observations can be explained by worldwide average variations of the elastic properties of the rock body overriding the megathrust fault, which deforms by dynamic stress transfer during co-seismic slip, and we discuss some general implications for tsunami hazard assessment. Second, we test this conceptual model for the particular case of the 1992 Nicaragua tsunami earthquake (M_S7.2 and M_W7.8). This event nucleated at ~20 km-deep but it appears to have released most of its seismic moment near the trench. This earthquake caused mild shaking and little damage, so that tsunami hazard based on human perception was underestimated and the destructive tsunami hit the coast unexpectedly. We use a set of 2D seismic data to map the P-wave seismic velocity above the inter-plate boundary, and we combine it with previously estimated moment release distribution to calculate slip and stress drop distributions and moment-rate spectra that are compatible with both the seismological and the geophysical data. The models confirm that slip concentrated in the shallow megathrust, with two patches of maximum slip exceeding 10-12 m in the near-trench zone that can explain the observed tsunami run-up, while the average stress drop is ~3 MPa. The low rigidity of the upper plate in the zone of maximum slip explains the high frequency depletion and the resulting M_W-M_S discrepancy without need to consider anomalous rupture properties or fault mechanics

How to cite: Sallares, V., Ranero, C. R., Prada, M., and Calahorrano B., A.: Earthquake rupture properties and tsunamigenesis in the shallowest megathrust, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-5809, https://doi.org/10.5194/egusphere-egu2020-5809, 2020.

Despite the growing spatio-temporal density of geophysical observations, our understanding of the megathrust earthquake cycle continues to be limited by a series of factors, in particular the short observation time compared to mega-earthquake recurrence and the partial spatial coverage of geodetic data. Here, we attempt to compensate for these natural limitations by simulating dozens of seismic cycles in a laboratory-scale analogue model of subduction. The model creates analog earthquakes of magnitude Mw 6.2–8.3, with a coefficient of variation in recurrence intervals of 0.5, similar to real subduction megathrusts. Using a digital image correlation technique, we measure coseismic and interseismic deformation – this is akin to having a dense continuous geodetic network homogeneously distributed over the whole margin. We show how, by deciphering the spatially and temporally complex surface deformation history, machine learning can predict the timing and size of analog earthquakes. Then, we investigate data characteristics that maximize the performance of a machine learning binary classifier predicting slip-events imminence. We show how this framing can be used for designing an efficient geodetic network, and defining the minimum space-time coverage requirements for analog earthquake prediction. Converting the laboratory scale to the natural scale, we found that a 70-85 km wide coastal swath gives the most important information on slip imminence and that model performance is mainly 
influenced by the alarm duration, with density of stations and record length playing a secondary role. Under optimal monitoring conditions, about ten seismic cycles long record is enough to predict alarm periods in good agreement with those observed.

How to cite: Corbi, F., Bedford, J., Sandri, L., Funiciello, F., Gualandi, A., and Rosenau, M.: Predictability of large subduction earthquakes: insights from analog models and machine learning, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-6805, https://doi.org/10.5194/egusphere-egu2020-6805, 2020.

Subduction zones are the most seismically active regions on the globe and about 90% of historical events, including the largest ones with the magnitude M>9, occurred along these regions (Hayes et al., 2018). Most of these events were followed by devastating tsunamis with, in some cases, perhaps unexpected wave height distributions. Observation of events in the megathrust environment reveals that some earthquakes are characterized by slip concentration on the very shallow part of the subduction zone. This shallow slip phenomenon was repeatedly observed in the last two decades for both ordinary megathrust events (e.g. 2010 Maule and 2011 Tohoku) and tsunami earthquakes (2006 Java and 2010 Mentawai). Shallow ruptures feature depleted short–period energy release and very slow rupture velocity possibly due to the presence of (hydrated) sediments (Lay et al., 2011; Lay 2014; Polet and Kanamori, 2000). Associated long rupture durations have been explained with fault mechanics-related rigidity and stress drop variation with depth (Bilek and Lay, 1999) or, more recently, with lower rigidity of surrounding materials (Sallares and Ranero, 2019).

The characteristics of co-seismic slip distribution have an important impact on tsunami hazard. There are numerous methods that have been proposed to generate stochastic slip distributions, also including shallow slip amplification (Le Veque et al., 2016; Sepulveda et al., 2017; Scala et al., 2019). However, these models need to be calibrated against slip models estimated for real events.

Here, we investigate similarities and differences between the synthetic slip distributions provided by Scala et al. (2019) and a suite of 144 slip models of real events that occurred in different subduction zones (Ye et al.,2016). In particular, Scala et al. (2019) model features shallow slip amplification in single events, whose relative probabilities are balanced to restore cumulative slip homogeneity on the fault plane over multiple seismic cycles. This study also aims to improve and/or calibrate this model to account for the behavior observed from real events.

How to cite: Bayraktar, H. B., Scala, A., Festa, G., and Lorito, S.: Improving the Co-seismic Slip Distributions of Synthetic Catalogs With Real Observations, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-1050, https://doi.org/10.5194/egusphere-egu2020-1050, 2020.

When modelling tsunamis and assessing tsunami hazard, it is frequently necessary to make simplifying assumptions in order to reduce the problem to one which is computationally tractable within a reasonable period of time. In this paper, we examine the key factors controlling the generation of the initial sea surface wave and present a series of clear and simple guidelines for real-world problems. We also provide number of computational resources (a tsunami loader) which may be utilised with existing tsunami propagation codes (e.g. COMCOT) to modify the initial sea-surface way, where necessary.

Most tsunami modelling codes operate under the assumption that the initial sea surface wave is identical to the seafloor perturbation. Yet this is only true for large tsunami sources (Kajiura 1963). With our tsunami loader we model the tsunamigensis process and the formation of the initial sea-surface wave. Critically, the diffusive effect of the water column above the deforming seafloor is accurately addressed, which can result in a substantial decrease in the energy in the initial sea-surface wave.

For example, let us consider a rectangular uplifting patch on the seafloor, at a depth of 4km. For a 4x4km square patch, the diffusive effect will result in an energy reduction of 90%. Even if one of those dimensions is 100 times larger, such that we have a relatively large 400x4 km uplifting region, the energy reduction is still 70%. We find the shortest dimension of the uplifting patch provides a strong control on the energy of the initial sea-surface wave, and consequential tsunami. If we move to a 40x40 km square patch we find the reduction is now 20%, and 400x40 km patch is now a relatively modest, but non-negligible 12%.

We also include other effects such as the time-dependence of seafloor deformation, which also reduces the potential tsunami energy, and horizontal advection of topography, which conversely increases the potential tsunami energy, in our analysis of the tsunamigenesis process. Currently implemented for fault sources, we are working to include landslide and volcanic sources.

How to cite: Moore, J., Hubbard, J., Felix, R., Lythgoe, K., and Switzer, A.: Tsunamigenesis Revisited, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-21113, https://doi.org/10.5194/egusphere-egu2020-21113, 2020.

We investigate the spatial variations of coda attenuation (Qc) structure in the tectonically complex Andaman–Nicobar subduction zone (ANSZ), which is one of the most seismically active subduction zones on the Earth. The region constitutes the northernmost part of the Sunda subduction zone, where the Indian plate disappears beneath the Burmese plate along the Burma and Andaman arcs to the east. This is probably the first attempt to map the Qc variations across the whole ANSZ. In a seismically active area, the spatial distribution of Qc is important to evaluate the seismic hazard in relation to tectonics and seismicity.

A total of 289 high-quality events recorded at a network of broad-band stations operational since 2009 are considered for the analysis. The variations in attenuation characteristics at different frequencies reveal a marked contrast from the northern to the southern Andaman region, consistent with the geotectonic diversity of the region. At low frequencies, low Qc values are observed in the northern part of ANSZ in the vicinity of the Narcondum volcanic island, which does not appear in the high-frequency image. The low values are in agreement with the 3-D tomogram, which suggests a distinct low-velocity structure below this volcanic island. The Andaman trench also exhibits a relatively low Qc, which is well correlated with the low-Vp zone. The spatial distributions of Q0 (Qc at 1 Hz) structure of the region are further projected onto three east–west profiles to capture the detailed attenuation characteristics from north to south. Results show that the northernmost part of ANSZ is more attenuative than the southern part, which may be indicative of the changes in physical properties of the crust. The frequency relation parameter (n) shows an inverse correlation with the observed Q0 values. Furthermore, we have observed a good correlation between the Q0 variation and the seismicity pattern of the area that enables us to enhance our understanding about the role of crustal heterogeneity in the earthquake occurrence in this area.

How to cite: Singh, C., Biswas, R., Jaiswal, N., and Kumar, M. R.: Spatial variations of coda wave attenuation in Andaman-Nicobar subduction zone, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-32, https://doi.org/10.5194/egusphere-egu2020-32, 2020.

3D imaging reveals striking along-trench structural variations of subduction zones world-wide (e.g., Han et al, JGR 2018). Subduction zones include basins, sediments, splay and back-thrusting faults that evolve over a large time span due to tectonic processes, and may crucially affect earthquake dynamics and tsunami genesis. Such features should be taken into account for realistic hazard assessment. Numerical modeling bridges time scales of millions of years of subduction evolution to seconds governing dynamic earthquake rupture, as well as spatial scales of hundreds of kilometers of megathrust geometry to meters of an earthquake rupture front.

Recently, an innovative framework linking long-term geodynamic subduction and seismic cycle models to dynamic rupture models of the earthquake process and seismic wave propagation at coseismic timescales was presented (van Zelst et al., JGR 2019). This workflow was extended in a simple test case to link the 2D seismic cycle model to a three-dimensional earthquake rupture mode, which was then linked to a tsunami model  (Madden et al., EarthArxiv, doi:10.31223/osf.io/rzvn2). Here, we couple a 2D seismic cycle model to 3D earthquake and tsunami models and assess the geophysical aspects of this coupling. We extract all 2D material properties, stresses and the strength of the megathrust, and its geometry, from the seismic cycling model at a time step right before a typical megathrust event to use as initial conditions for the 3D dynamic rupture models. We explore the effects of along-arc variations of megathrust curvature, sediment content, and closeness to failure of the wedge on earthquake dynamics by studying the effects on slip, rupture velocity, stress drop and seafloor deformation.

In a next step, the dynamic seafloor displacements are linked to tsunami simulations that use depth-integrated (hydrostatic) shallow water equations. This approach efficiently models wave propagations and large-scale horizontal flows. We also present novel, fully coupled 3D dynamic rupture-tsunami simulations (Krenz et al., AGU19; Abrahams et al., AGU19; Lotto and Dunham et al., 2015, Computational Geosciences) which solve simultaneously for the solid earth and ocean response, taking gravity into account via a modified free surface boundary condition.

Earthquake rupture modeling and the fully-coupled tsunami modeling utilize SeisSol (www.seissol.org), a flagship code of the ChEESE project (www.cheese-coe.eu). SeisSol is an open source software package using unstructured tetrahedral meshes that are optimally suited for the complex geometries of subduction zones. The here presented links between geodynamic subduction and seismic cycling model with earthquake dynamics and tsunami models better account for the complexity of subduction zones and help evaluate the effects of along arc heterogeneities on earthquake and tsunami behavior and advance physics-based assessments of earthquake-tsunami hazards.

How to cite: Wirp, S. A., Gabriel, A.-A., Madden, E. H., van Zelst, I., Krenz, L., and van Dinther, Y.: Linking geodynamic subduction models to self-consistent 3D dynamic earthquake rupture and tsunami simulations, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-1675, https://doi.org/10.5194/egusphere-egu2020-1675, 2020.

We present a 2D p-wave velocity (Vp) model and a coincident multichannel seismic reflection profile mapping the structure of the southern Costa Rica margin and incoming Cocos Ridge. The seismic profiles image the ocean and overriding plates from the trench across the entire offshore margin, including the structures involved in the 2002 Mw6.4 Osa earthquake. The overriding plate consists of three domains: Domain I at the margin front displays thin-skinned deformation of an imbricated-thrust system composed of fractured rocks with relatively low Vp. Domain II under the middle continental shelf is comparatively less fractured, showing ~15 km long landward-dipping reflection packages and discrete active deformation of the shelf sediment and seafloor. Domain III in the inner shelf is little fractured and appears to be dominated by elastic deformation, with inactive structures of an extensional basin consisting of tilted blocks overlain by ~2 km-thick gently landward-dipping strata. The velocity structure supports the argument that the bulk of the margin is highly consolidated rock possibly similar to outcrops in the Osa Peninsula. Thick-skinned tectonics probably causes the uplift of Domains II and III. The incoming oceanic plate shows crustal thickness variations from ~14 km at the trench (Cocos Ridge) to 6-7 km beneath the continental shelf. We combine (1) inter-plate geometry and velocity-derived fracturing degree at the base of the overriding plate, (2) tectonic stresses and brittle strain above the inter-plate boundary extracted from 3D numerical models, and (3) earthquake locations, to investigate potential relationships between structure and earthquake generation. The 2002 Osa earthquake and its aftershocks appear to have nucleated at the leading flank of two subducting seamounts, coinciding with the area of highest tectonic overpressure in numerical models. Both estimated rock fracturing and modelled brittle strain, steadily increase from the leading flank of the subducting seamounts to their top, which we interpret to reflect the progressive damage caused by the incoming plate relief. Therefore, the analysis supports a spatial and temporal relationship between subducting seamount location, upper plate fracturing, brittle strain, tectonic overpressure, and earthquake nucleation.

How to cite: Martínez-Loriente, S., Sallarès, V., R. Ranero, C., B. Ruh, J., Barckhausen, U., Grevemeyer, I., and Nangs, N.: Influence of incoming plate relief on overriding plate deformation and earthquake nucleation: Cocos Ridge subduction (Costa Rica), EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-2940, https://doi.org/10.5194/egusphere-egu2020-2940, 2020.

The megathrust topography is key in conditioning the structural integrity of the overriding plate, and thus, the generation of tsunamigenic structures. Our objective is to investigate the Rivera subduction zone, offshore the Mexican Pacific coast, known for hosting large megathrust tsunamigenic earthquakes (Mw > 7.5), and where little is known regarding the distribution of tsunamigenic structures along the margin. Our working hypothesis is that there is an interaction between the megathrust relief at the surface of the subducted slab (Rivera Plate) and the existence of tsunamigenic structures in the above unsubducted plate (North America). To investigate this interaction, we used seismic methods to characterize the variations of the physical properties of the overriding plate, generally related to tectonic (faults) structures that are sources of tsunamis, with the reliefs of the deeper subducted plate obtained with the same method. Here, we use spatially coincident 2D multichannel seismic (MCS, 5.85 km long-streamer) and active marine wide-angle seismic (WAS) data acquired during the TSUJAL survey in 2014 offshore west of Mexico to measure structural variations of the overriding plate and the megathrust interface. We have jointly inverted refracted and reflected travel-times (TT) from both MCS and WAS data to constrain the P-wave velocity (Vp) structure of the overriding plate and the geometry of the megathrust. Before the inversion and to increase the amount of refracted TT we have applied the downward continuation technique to MCS field data allowing to better image the refracted waves in the records. MCS data has a higher spatial sampling than OBS data, which translates into a higher density sampling of the refracted waves and hence the tomographic resolution. Therefore, the resulting tomographic model displays small-scale velocity structure variations of the overriding plate and the megathrust relief that would not be resolved with TT from OBS data only. We used further refracted and reflected TT from OBS data to constrain the Vp structure of the subducting oceanic plate and the geometry of the oceanic Moho. The inverted megathrust interface obtained with the tomography shows clear topographic features in its shallow portion (<~10 km from the trench). Such topographic variations are smaller than the average size of seamounts of the Rivera plate, but they are similar to the seafloor fabric generated by a relict East Pacific Rise segment identified west of the trench in the bathymetry map of the region. Time-migrated images were also obtained after processing the MCS data to constrain the tectonic framework of the shallow subduction zone regardless of the tomographic models. The seismic sections reveal the lack of an extensive accretionary prism, implying that subduction-erosion dominates the structure of the margin in this region. Integrating all the data results, we find that megathrust highs correlate with low-velocity anomalies, suggesting the presence of fluids, and correlate with the presence of extensional faults in the overriding plate as well. This correlation demonstrates the control that megathrust topography exerts on the formation of tsunamigenic structures along the Rivera plate boundary.

How to cite: Bartolome, R., Prada, M., Gras, C., Begovic, S., Bandy, W., and Dañobeitia, J. J.: Interaction between interplate fault topography and tsunamigenic structures at the subduction zone offshore West Mexico, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-3593, https://doi.org/10.5194/egusphere-egu2020-3593, 2020.

The Nankai Trough megasplay fault likely hosts different modes of fault slip, from slow to megathrust earthquakes, and is responsible for related phenomena such as tsunamis and submarine landslides. All types of slip events require some kind of frictional weakening process (e.g. slip and/or velocity weakening) in order to nucleate and propagate. Most frictional earthquake studies analyze the velocity dependence of friction but disregard the slip dependence of friction observed in experimental friction studies.

We tested fluid-saturated powdered megasplay fault samples from Integrated Ocean Drilling Program Site C0004 in a direct shear apparatus under effective normal stresses from 2 – 18 MPa to investigate the velocity- and slip-dependence of friction of the megasplay fault. For every tested effective normal stress, we performed one velocity-step experiment and two constant velocity experiments (no velocity step). In the velocity-step experiments the samples were sheared to a total displacement of 10 mm, with an initial sliding velocity V₀ = 0.1 µm/s for the first ~5 mm (run-in) followed by a velocity step increase to V = 1.0 µm/s over the last 5 mm. During the constant velocity experiments, the shearing velocity (0.1 and 1.0 µm/s respectively) was held constant for 10 mm of displacement.

The velocity-stepping tests showed an evolution from velocity weakening at low effective normal stresses to velocity strengthening at high effective normal stresses. All experiments revealed strong slip-weakening behavior, with the slip dependence having a much larger effect on friction than the velocity dependence. The friction slip dependence is also controlled by the effective normal stress, showing large weakening rate at low effective normal stresses and smaller weakening rate at higher effective normal stresses. Therefore, both frictional weakening mechanisms on the megasplay fault become more effective at shallow depths. This may amplify seafloor deformation by shallow coseismic slip and could increase the tsunamigenic potential of the fault zone.

How to cite: Roesner, A., Ikari, M., Huepers, A., and Kopf, A.: Slip-dependent weakening revealed for a shallow megasplay fault in the Nankai subduction zone, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-5457, https://doi.org/10.5194/egusphere-egu2020-5457, 2020.

In the shallow region of subduction zones, topographic variations of the interplate interface condition the structural integrity of the upper plate, and thus the distribution of elastic properties in this region, which determines its tsunamigenic potential. Yet, we know little about the distribution of elastic properties in these shallow regions, which yields large uncertainty during tsunami hazard assessment.

Here we assess topographic variations of the interplate boundary as well as the distribution of elastic properties of the upper plate in two tsunamigenic regions of the Middle American Trench. We focus on the rupture area of three tsunami earthquakes, the 1992 Nicaragua event, and the 1932 and 1995 Jalisco-Colima earthquakes (Pacific Mexican coast).

We use 2D coincident wide-angle (WAS) and multichannel seismic (MCS) lines acquired across the rupture area of each event to jointly invert refracted and reflected travel-times (TT) and obtain the P-wave velocity (Vp) structure of the tsunamigenic region of the upper plate, and the geometry of the interplate boundary. Mixing both types of seismic data allowed for the first time to retrieve small-scale local topographic variations of the interplate that would have been omitted with the classical inversion of WAS TT. From Vp, we derive other elastic parameters namely, density, S-wave velocity, and rigidity using well-established empirical relationships.

The results show that the heterogeneous distribution of the elastic properties of the upper plate in the shallow tsunamigenic region correlates with topographic variations of the interplate in both margins. These results not only sustain the direct relationship between the interplate relief and the tectonic structure of the overriding plate as it has been already stated by previous authors, but they also allow to quantify the relationship between topographic highs of the subducted plate with low rigidity regions in the upper plate. This quantification is of paramount importance in these shallow regions of the subduction, because low rigidity implies high slip during coseismic deformation, and therefore, high tsunamigenic potential. The heterogeneous distribution of elastic properties inferred for the upper plate in this study should be considered during tsunami modeling, tsunami hazard assessment and tsunami early warning systems.

How to cite: Prada, M., Sallares, V., Bartolome, R., Melendez, A., Calahorrano, A., Begovic, S., Gras, C., and Ranero, C.: The heterogeneous distribution of elastic properties in the tsunamigenic region of subduction zones, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-8603, https://doi.org/10.5194/egusphere-egu2020-8603, 2020.

Historical earthquakes in the Java subduction zone have given genesis to tsunami affecting the southwest coasts of the island of Java, in Indonesia. The most recent earthquake on the 17^th of July 2006, has given rise to a tsunami that killed more than 600 people. The tsunami was difficult to escape due to the small amount of ground shaking, which could have acted as an early warning, and due to the epicentre being very close to the shorelines, giving insufficient time for response. Historical data and scientific studies give little evidence for mega-thrust events in the Java trench, however such possibilities are not excluded and could have a devastating impact in the region. This work aims to assess the tsunami hazard occurring from a range of earthquake scenarios in the subduction zone. Taking as a benchmark the 2006 event, we initially validate our modelling approach against the wave observations recorded at three tide gauges. We then expand our work to account for future earthquake scenarios and their tsunamigenic consequences in the southern coasts of Java island. Bathymetry displacement is computed using the Okada elastic dislocation model. The nonlinear shallow water equation solver JAGURS is employed for the modelling of wave propagation. Our objective is to quantify the uncertainty of such events by using statistical surrogates: fast stochastic approximations of the model that can explore the likelihood of thousands of tsunami scenarios in a few moments of time. Gaussian process emulators are utilised to predict maximum wave amplification occurring from varying parameter distributions such as the moment magnitude of an earthquake. The resulting tsunami hazard footprints can be used in conjunction with existing socio-demographic information to assess tsunami risk in vulnerable areas. The end-data can eventually be used to inform policy making for better disaster mitigation planning.

How to cite: Salmanidou, D., Heidarzadeh, M., and Guillas, S.: Assessing future uncertainties: earthquake tsunami hazard in the Java trench, Indonesia., EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-10762, https://doi.org/10.5194/egusphere-egu2020-10762, 2020.

Characterizing the time-dependent slip evolution along subduction megathrusts during seismic cycles is a key step to unfold the recurrence patterns of and hazard imposed by great earthquakes. However, having adequate geodetic observations in the appropriate locations is essential for reliably modelling both interseismic coupling and coseismic slip. The 2011 Tohoku-oki earthquake as a well-known trench breaking and tsunamigenic megathrust seismic event clearly demonstrated the limitations of the distributed slip models using land-limited geodetic instruments. In this study, we have set up a scaled analogue megathrust model to produce analogue earthquakes while Digital Image Correlation (DIC) and the Analogue Geodetic Slip Inversion Technique (AGSIT) have been applied to retrieve the model surface velocities (incremental displacement) and model coseismic slip distribution, respectively. We generated more than 20 slip models for a series of events by sequentially disregarding trenchward rows of virtual GPS stations (vGPSs) for slip modelling thereby systematically reducing simulated offshore coverage. The analogue earthquakes have been categorized to two different sets as non-trench-breaking and trench-breaking ruptures. Here we show how slip models of analogue earthquakes change as a function of offshore coverage quantitatively and qualitatively. The sensitivities with respect to a potential bimodality of slip distribution and up-dip limit of the slip distribution model have also been assessed.

How to cite: Kosari, E., Rosenau, M., Bedford, J., Rudolf, M., and Oncken, O.: On the relationship between offshore geodetic coverage and slip model uncertainty , EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-11165, https://doi.org/10.5194/egusphere-egu2020-11165, 2020.

Non-planar faults and surface reached rupture are seldom considered in the source modelling of subduction zone earthquakes. Here we present a new method for accounting for both phenomena in the generation of stochastic slip distribution while still maintaining self similar properties. To do this, we use the composite source model, which involves the placement of numerous circular dislocations on the fault plane. The fault plane is described by an unstructured mesh allowing for a non-planar surface while surface rupture is correctly accounted for by reflecting the slip from circular dislocations that intersect with the fault trace.

In a case study we demonstrate that the inclusion of rupture at the surface alters the ground or seafloor deformation both in terms of the magnitude (between 60%-20% in 5km zone near the fault trace) and the orientation of the deformation vectors (i.e. by up to 5 degrees). Such changes can have a significant effect on tsunami source and subsequent wave.

Additionally, with a prescribed rupture velocity model, complex source time functions can also be calculated for each element on the fault plane. Generally, rise time is assumed to be instantaneous in tsunami simulation.

We will also present preliminary results focused on comparing the tsunami wave height observed along nearby coastlines generated by the different source models (i.e. with/without surface reached rupture and variable source time functions).    

How to cite: Murphy, S., Herrero, A., Romano, F., and Lorito, S.: Application of stochastic fractal surface rupture on non-planar faults in tsunami simulation, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-11267, https://doi.org/10.5194/egusphere-egu2020-11267, 2020.

On 28 September 2018 a Mw 7.5 strike-slip earthquake occurred on the Palu-Koro fault system in the Sulawesi Island. Immediately after the earthquake a powerful tsunami hit the Palu Bay causing large damages and numerous fatalities.

Several works, inverting seismic or geodetic data, clearly estimated the slip distribution of this event, but the causative source of the tsunami is still not completely understood; indeed, the strike-slip mechanism of the seismic source alone might not be sufficient to explain the large runups observed (> 6 m) along the coast of the Palu Bay, and thus one or more additional non-seismic sources like a landslide could have contributed to generate the big tsunami. An insight of that can be found in an extraordinary collection of amateur videos, and on the only available tide gauge in the Bay, at Pantoloan, that showed evidence for a short period wave of at least 2-3 minutes, compatible with a landslide.

In this study, we attempt to discriminate the contribution in the tsunami generation of both the seismic source and  some supposed landslides distributed along the coast of the Bay.

In particular, we attempt to estimate the causative source of the tsunami by means of a nonlinear joint inversion of geodetic (InSAR) and runup data. We use a fault geometry consistent with the Sentinel-2 optical analysis results and analytically compute the geodetic Green’s functions. The same fault model is used to compute the initial condition for the seismic tsunami Green’s functions, including the contribution of the horizontal deformation due to the gradient of the bathymetry (10 m spatial resolution); the landslide tsunami Green’s functions are computed the software BingClaw by placing several hypothetical sources in the Bay. In both the cases the tsunami propagation is modelled by numerically solving the nonlinear shallow water equations.

In this work we also attempt to address the validity of Green’s functions approach (linearity) for earthquake and landslide sources as well as the wave amplitude offshore as predictor of nearby runup.

How to cite: Romano, F., Hasan, H., Lorito, S., Løvholt, F., Brizuela, B., Tolomei, C., and Piatanesi, A.: Study of the tsunami source in the Palu Bay following the Mw7.5 2018 Sulawesi earthquake, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-11535, https://doi.org/10.5194/egusphere-egu2020-11535, 2020.

Subduction zones host some of the greatest megathrust earthquakes in the world. Slow earthquakes have been discovered around the subduction zones of the Pacific rim very close to megathrust earthquakes. Investigating the lithosphere of the slow earthquake area versus non slow-earthquake area in subduction zones is crucial in understanding the role of the internal structure to control slow earthquakes. In this study, we investigate the lithospheric structure of stations in the slow earthquake area and non slow-earthquake areas in Chile using receiver function analysis and inversion method using teleseismic earthquakes. Here we focus on, especially the Vp/Vs ratios from both slow and non-slow earthquake areas, because the Vp/Vs ratio is sensitive to the fluid distribution in the lithosphere; the fluid distribution possibly controls the potential occurrence of slow earthquakes. Additionally, the nature of the slab can also play a crucial factor. The Vp/Vs ratio results across depth shows significantly higher value in the deeper oceanic slab region beneath the stations in the slow earthquake areas with higher contrast at the boundary.

How to cite: Mukherjee, P., Ito, Y., S. Garcia, E., Plata-Martinez, R., and Shibutani, T.: The Chile subduction zone : Nature of the lithosphere between alternating regions of slow earthquakes versus non slow earthquakes, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-4035, https://doi.org/10.5194/egusphere-egu2020-4035, 2020.

Slow earthquakes play important roles in the occurrence of megathrust earthquakes in subduction zones. An increasing number of seismic networks have contributed to significant findings and the detection of slow earthquake activities; however, it is still unclear what sort of seismological structures exhibits each slow earthquake activity. We have developed the multi-band receiver function (RF) method, in which the RFs are composed of different higher-frequency contents. We, here, reveal smaller-scale structures from the RFs from local deep-focus earthquakes around the Philippine Sea plate boundary in Southwestern Japan, where numerous slow earthquakes have been detected (e.g., Obara, 2002; Ito et al., 2007; Nishimura et al., 2013).

Deep-focus earthquakes, frequently occurring in the Pacific slab below Southwestern Japan, can be applicable to the multi-band RF analysis because the local deep events and teleseismic events are similar in the slowness of the first-arrival phases. Local deep-focus events, however, have different variations in back azimuths from teleseismic events, which enables us to estimate seismological structures in a wider range of azimuths by stacking traces from both events. We carefully select the deep-focus events with longer S-P time than 40 sec and exclude triplication phases from mantle transition zones. Here we apply this method to short-period 3-component seismograms of Hi-net (NIED, Japan) in the Northeastern Kii Peninsula, where short-term slow slip events (SSEs) and episodic tremors are very active (e.g., Obara et al., 2010; Nishimura et al., 2013; Yabe & Ide, 2014).

Cross-sections of higher-frequency RFs (up to 2 Hz) show sharp and strong negative phases from the plate interface shallower than 35 km depth, which is one of the most active regions of episodic tremors (Obara et al., 2010). At the deeper portion, the higher-frequency RFs exhibit the mantle wedge structure with obscure phases of the plate interface, where minor and continuous tremor activities have been reported (Obara et al., 2010). These results suggest that episodic tremors accompanied by short-term SSEs occur on the interface between the continental crust and the oceanic crust, whereas the source regions of minor tremors are between the oceanic crust and the mantle wedge as indicated in Kato et al. (2010).

How to cite: Sawaki, Y., Ito, Y., Ohta, K., Shibutani, T., and Iwata, T.: A New Approach to Clarify Slow Earthquake Source Regions: Multi-band Receiver Function Analysis Including Local Deep-focus Events, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-4921, https://doi.org/10.5194/egusphere-egu2020-4921, 2020.

Great subduction earthquakes occur along the seismogenic zone of the megathrust, a fault segment that is mechanically coupled so that seismic rupture can propagate. Numerous factors, including the rheology and structure of the plates, shear stress distribution, fluids pressure, or thermal structure, size and width of the coupled zone, have been proposed to play a role to determine and dynamic behavior of the rupture. These factors are conditioned, in turn, by the properties of the rocks undergoing deformation during seismic rupture and by the fault geometry and roughness. While the information on the 3D velocity field and on inter-plate geometry can potentially be extracted from travel-times of active seismic data, the experiments that are appropriate to define these parameters are scarce. Additionally, most 3D travel-time tomography codes use only first arrivals to define the velocity field, so that they do not provide information on the megathrust geometry. Here we combine for the first time ever wide-angle seismic data with a joint refraction and reflection travel-time tomography code (tomo3d), to retrieve a 3D velocity model of margin as well the geometry of the inter-plate boundary with unprecedented detail. In particular, we use data acquired in the French project “Esmeraldas” offshore NE Ecuador/SE Colombia in 2005. Our model, which builds on a previous one obtained by first arrival tomography, goes from the surface to 18-20 km depth, covering a substantial part of the seismogenic zone. This region, where the Nazca plate plunges beneath South America, has produced remarkable examples of variable earthquake rupture behavior. The entire ~500 km-long segment ruptured during the great tsunamigenic earthquake of 1906 (Mw = 8.8), and it was ruptured again by three smaller events, directly adjacent to one another, in 1942 (Mw = 7.8), 1958 (Mw = 7.7) and 1979 (Mw = 8.2). According to our results, the inter-plate boundary where these earthquakes took place is of variable dip and rough, spotted by 2-3 km-high and 10-15 km-wide features that resemble subducting seamounts. The velocity of the overriding plate just above the inter-plate boundary is strongly heterogeneous, showing velocity-derived rock rigidity variations of up to 30-40% both along- and across-strike. The presence of inter-plate relief and velocity changes is confirmed by parameter uncertainty analysis and data sensitivity tests. Interestingly, the sharpest velocity contrasts appear to follow the limits between crustal blocks of different origin and composition that, according to previous work, could correspond to crustal-scale faults acting as barriers during earthquake propagation. The combined effect of a rough inter-plate boundary and heterogeneous elastic properties on earthquake rupture and tsunamigenesis remains to be tested by dynamic rupture models

How to cite: Sallares, V., Melendez, A., Terzic, D., Buinheira, P., Charvis, P., Galve, A., Collot, J.-Y., and Calahorrano B., A.: 3D distribution of elastic properties and subduction inter-plate relief in NW Ecuador from joint refraction and inter-plate reflection travel-time tomography, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-5842, https://doi.org/10.5194/egusphere-egu2020-5842, 2020.

Large and destructive earthquakes (M_w≥ 7.5) occur worldwide particularly along the major subduction zones causing extensive damage and loss of life in the hinterland of epicentral region. Source models and rupture characteristics of these earthquakes (i.e. faulting geometry, focal depth, non-uniform finite-fault slip distributions) can be precisely determined by using seismological data and multidisciplinary earth-science observations. It is also known that earthquake source parameters play key roles in the modelling of secondary events such as earthquake-induced tsunamis. There are many studies emphasizing the importance of using heterogonous slip distribution models of earthquakes in mathematical tsunami simulations to predict synthetic tsunami waves more consistent with the observed ones. In this study, we obtained double-couple source mechanisms and slip distribution models of complex large earthquakes (M_w≥ 7.5) lately occurred at different parts of the Earth. For this purpose, we used point-source teleseismic P- and SH- body waveform inversion and kinematic slip distribution inversion techniques. Besides, azimuthal distributions of P- wave first motion polarities, which are recorded by near-field and regional seismic stations, are checked to approve obtained minimum misfit source mechanism parameters of earthquakes. We essentially observed that tsunamigenic earthquakes occurred at shallow focal depths (h ≤ 70 km) with dip-slip source mechanisms and rather complex slip distributions along the fault planes. However, in some cases, tsunami waves may be unexpectedly triggered due to the secondary effects of large strike-slip earthquakes (e.g., September 28, 2018 Palu, Indonesia - M_w7.5). Here, we discuss our inversion results, which reveal the significant contributions of earthquake source studies on resolving the relationships between the faulting geometry, rupture characteristics and tsunami generation. Furthermore, the necessity of high-resolution bathymetry data in numerical tsunami simulations is highlighted for the modelling of tsunami waves, in particular, recorded at the near-field tide-gauge stations. This study is partially supported by the Turkish Academy of Sciences (TÜBA) through GEBIP program.

How to cite: Yolsal-Çevikbilen, S. and Taymaz, T.: Seismological Constraints on Fault-Slip Source Models and Rupture Characteristics of Global Large Earthquakes (Mw ≥ 7.5) and Associated Tsunamis , EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-7922, https://doi.org/10.5194/egusphere-egu2020-7922, 2020.

The Lesser Antilles subduction zone is a challenging region when it comes to unravelling its seismogenic behaviour. Over the last century, it has been seismically quiet, with no large thrust events recorded, leading to the question whether this subduction zone is able to produce large interplate earthquakes or not. The slow subduction velocity of ~20 mm/yr complicates this even further, as mega-earthquake recurrence times would be up to many hundreds of years in the case of a fully locked subduction interface, and up to several thousands of years for a partially locked interface. The record of two large historical earthquakes, a M ~8 in 1839 and M ~8.5 in 1843, is often referred to as evidence supporting the seismic character of the Lesser Antilles subduction zone. It remains, however, questionable whether these events actually occurred along the subduction interface.

Here we use GPS data acquired on various islands within the Antilles to infer interseismic coupling along the Lesser Antilles Arc. Previous block models have suggested low coupling of the subduction interface, making the occurrence of large megathrust earthquakes less likely. However, the non-uniqueness of these inversions, as well as uncertainties related to the distance between GPS stations and the subduction trench, cast doubts on how well the inferred coupling represents the actual degree of locking along the subduction interface. In this study, we attempt to improve these estimates, by using a Bayesian approach to derive a meaningful set of uncertainties on the distribution of interseismic coupling. By exploring the entire range of model parameters, we are able to provide a probabilistic estimate of interseismic coupling. To further improve our analysis with respect to previous models, we incorporate a layered elastic structure, as well as a more realistic fault geometry, testing two different slab models.

Our results suggest that the subduction interface of the Lesser Antilles subduction zone is most likely to be uncoupled. A sensitivity analysis highlights the deeper part of the interface (i.e., 30-60 km depth) as the region with higher sensitivity, since the GPS stations are distributed mostly above that portion of the subduction. A test regarding the proposed 1843 rupture contour reveals that this area is very unlikely to be locked. This apparent aseismic character of the Lesser Antilles raises questions about the role of slow slip along the interface. We therefore also analyse GPS time series to assess the spatial and temporal distribution of transient deformation signals in the region.

How to cite: van Rijsingen, E., Calais, E., Jolivet, R., de Chabalier, J.-B., Jara, J., Symithe, S., Robertson, R., and Ryan, G.: Seismogenic behaviour in the Lesser Antilles: Insights from geodetic observations, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-7688, https://doi.org/10.5194/egusphere-egu2020-7688, 2020.

Tsunamis observed by offshore ocean-bottom pressure gauges have been used to infer fault models and stress drops for major (M > 7) offshore earthquakes, to understand the earthquake and tsunamigenesis (e.g., Satake et al. 2013). However, it is challenging to observe tsunamis due to moderate (M ~6) earthquakes with reasonable quality by those, previous, few and remote pressure gauge arrays. Recently, a new, dense and wide pressure gauge network, the Seafloor Observation Network for Earthquakes and Tsunamis along the Japan Trench (S-net), was constructed off eastern Japan (Kanazawa et al. 2016). This array observed tsunamis associated with a moderate (M~6) earthquake which occurred inside the array, with amplitudes of less than one cm. We analyzed these millimeter-scale tsunami records to infer the finite fault model and stress drop, and to examine its relationship with other interplate earthquake phenomena.

We analyzed the pressure data associated with an Mw 6.0 earthquake off Sanriku on August 20, 2016. This earthquake was located at the shallowest part of the plate boundary off Sanriku, Japan, near the northern edge of the rupture area of the 1896 Sanriku tsunami earthquake (Kanamori, 1972). Although the signal-to-noise ratio is not high, the westward tsunami propagation with the velocity of ~0.1 km/s could be recognized when the waveforms were aligned according to the station locations. Using these data, we constrained the rectangular fault model with a uniform slip across the fault. As a result, the fault model was located ~10 km to the west of the Global CMT centroid (a seismic moment M₀= 1.4 × 10¹⁸ Nm, Mw 6.0, and a stress drop of Δσ = 1.5 MPa). The stress drop seems not so small as expected in tsunami earthquakes such as the 1896 Sanriku tsunami earthquake (≪ ~1 MPa, e.g., Kanamori 1972) even if the uncertainty of the stress drop estimation is considered (Δσ > ~ 0.7 MPa). We also found the rupture area was unlikely to overlap with regions where slow earthquakes are active, such as low-frequency-tremors and very-low-frequency-earthquakes (e.g., Matsuzawa et al. 2015; Nishikawa et al. 2019; Tanaka et al. 2019).

This result demonstrates that the S-net new dense and wide pressure gauge array dramatically increases the detectability of a millimeter-scale tsunami and the constraints on earthquake source parameters of moderate earthquakes off eastern Japan. It is expected that more tsunamis due to minor-to-moderate offshore earthquakes are recorded by this new array, which will reveal the spatial variation of the stress drops, or mechanical properties, along the plate interface with much higher resolution than previously possible.

How to cite: Tatsuya, K., Saito, T., and Suzuki, W.: Fault modeling and stress drop estimation based on millimeter-scale tsunami records of an M6 earthquake detected by the dense and wide pressure gauge array, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-11999, https://doi.org/10.5194/egusphere-egu2020-11999, 2020.