Seismic Model Research Articles

A seismic source characterization model of multi-station based on graph neural network

A seismic source characterization model of multi-station based on graph neural network

A quantitative analysis of acoustic field directionality from a seismic airgun array

Marine seismic reflection surveys generate high amplitude impulsive acoustic events using airgun arrays to study geophysical characteristics of the seabed. These data can be used beyond seismic imaging – modeling short range propagation, considering impacts on marine mammals, extracting seabed properties and their effect on the acoustic field, etc. Knowledge of the source characteristics is necessary to utilize these data and proper modeling of sound propagation from these arrays requires characterization of the array beam pattern. Complex simulations of airgun arrays have been used in the past to model airgun array spectra, but beam patterns have not been thoroughly considered in the literature. Delay-and-sum combinations of these airgun signatures provide a simple beam pattern estimate, but this approach ignores variability in airgun position, timing, amplitude, interactions between airguns, etc. The use of more complex notional airgun signatures can yield more accurate estimates, but these are more challenging to model and still ignores shot-to-shot variability. Experimentally-determined beam patterns are evaluated here and the variability in the results considered, showing both similarities and notable differences from simulated results. The experimental results indicate that the source array depth impacts the ghost-free array beam pattern and that variability between shots is enough to significantly alter beam patterns. Overall, the observations suggest that accurate simulation of array beam patterns may require more complexity than is currently considered and that inclusion of uncertainty due to environmental and airgun shot variability is essential.

Limited arc magmatism and seismicity due to extensive mantle wedge serpentinization in the Makran subduction zone

Subduction processes usually involve extensive seismicity and create voluminous magmatic arcs by mantle wedge melting caused by dehydration of the subducting slab, but the Makran subduction zone has anomalously low seismicity and magmatism. Here we explain these anomalous features by 60–65% serpentinization in the peridotitic shallow mantle wedge based on our new integrated seismic, magnetic, gravity and isostatic model across the Makran subduction zone. The low-angle, slow Makran subduction provides ample time for the slab to release sufficient amounts of fluids for creating a large volume of rheologically weak serpentinite. This reduces seismicity by lowering the friction between the slab and surrounding rocks. Further, very little fluid is left in the slab when it reaches the melting depth, which explains the limited arc magmatism. Around 100 km depth, the subduction switches from low-angle to almost vertical. Our model demonstrates the combined effects of subduction rate and dip on mantle serpentinization with implication for assessment of seismic and volcanic hazards in subduction systems.

Reference 1D Seismic Velocity Models for Volcano Monitoring and Imaging: Methods, Models, and Applications

Abstract Seismic velocity models of the crust are an integral part of earthquake monitoring systems at volcanoes. 1D models that vary only in depth are typically used for real-time hypocenter determination and serve as critical reference models for detailed 3D imaging studies and geomechanical modeling. Such models are usually computed using seismic tomographic methods that rely on P- and S-wave arrival-time picks from numerous earthquakes recorded at receivers around the volcano. Traditional linearized tomographic methods that jointly invert for source locations, velocity structure, and station corrections depend critically on having reasonable starting values for the unknown parameters, are susceptible to local misfit minima and divergence, and often do not provide adequate uncertainty information. These issues are often exacerbated by sparse seismic networks, inadequate distributions of seismicity, and/or poor data quality common at volcanoes. In contrast, modern probabilistic global search methods avoid these issues only at the cost of increased computation time. In this article, we review both approaches and present example applications and comparisons at several volcanoes in the United States, including Mount Hood (Oregon), Mount St. Helens (Washington), the Island of Hawai’i, and Mount Cleveland (Alaska). We provide guidance on the proper usage of these methods as relevant to challenges specific to volcano monitoring and imaging. Finally, we survey-published 1D P-wave velocity models from around the world and use them to derive a generic stratovolcano velocity model, which serves as a useful reference model for comparison and when local velocity information is sparse.

Optimized seismic hazard and structural vulnerability model considering macroseismic intensity measures

Macroseismic intensity measures are fundamental quantitative indicators for evaluating the vulnerability and resilience of regional building clusters. The vulnerability model of building portfolios established can provide a noteworthy reference for the development of large-scale regional earthquake risk distributions. However, there are differences in the quantitative scales of the macroseismic intensity standards used in multiple regions. The positive effects of instrument intensity and probabilistic seismic hazard theory on the vulnerability assessment of regional buildings are rarely considered, resulting in significant differences in their assessment results. This paper comprehensively considers the impact of instrument intensity and seismic hazard models on traditional macroseismic intensity. An updated macroseismic intensity measure (UCMS-20) is proposed for evaluating regional buildings, which is combined with the latest version of China's seismic intensity standards. The actual seismic damage sample dataset (31,649 buildings) of four types of building clusters in three typical earthquakes in China (Wenchuan 2008, Yushu 2010, and Jiuzhaigou 2017) has been expanded and updated. Using the UCMS-20, the European Macroseismic Scale (EMS-98), the Modified Mercalli intensity (MMI), and the Medvedev, Sponheuer, and Karnik Scale (MSK-64), vulnerability assessments are conducted on the extended empirical structural seismic damage dataset, and an empirical disaster matrix model considering updating damage states is established. A nonlinear vulnerability prediction model considering updated macrointensity measures is proposed, and the optimized model is validated and compared against the developed multiscale empirical structure dataset. The multiscale vulnerability index calculation model is improved, and novel optimization curves, point clouds, and matrix probability risk models are established, considering the expansion of empirical structural vulnerability databases.

Simulation of January 22, 2024 Mw=7.0 Uqturpan earthquake consequences with the "Extremum" system application

This paper reports the use of intensity attenuation equations for the territory of the Northern Tien Shan by modeling the impact of a powerful destructive earthquake of January 22, 2024 in China. The relevance of the study follows from the needed reliable estimates of possible losses due to earthquakes in order to aid the decision making process for the response and the proper choice of a search and rescue strategy for the heavily affected settlements. The paper has as its goal the calibration of the Extremum system models for the territory of the Northern Tien Shan. Our study is the first to analyze the seismic intensity attenuation equations obtained by researchers at different times with a view to their applicability to near real time loss assessment for strong events occurring in the area of study. We investigate how the results of seismic attenuation modeling are affected by the ellipse compression factor k. and the source parameters of a strong event as determined by different seismological agencies in the alert mode. Computer simulation was applied to assess the possible impact of the January 22, 2024 earthquake using the "Extremum" system developed with our participation. This paper provides results of modeling the impact of the 2024 Uqturpan earthquake, as well as an assessment of the convergence achieved by calculated and observed intensities for various intensity prediction equations obtained before for China, Kyrgyzstan and adjacent areas within the Gissar-Kokshaalsky seismically active belt.

EARA2024: A New Radially Anisotropic Seismic Velocity Model for the Crust and Upper Mantle beneath East Asia and Northwestern Pacific Subduction Zones

EARA2024: A New Radially Anisotropic Seismic Velocity Model for the Crust and Upper Mantle beneath East Asia and Northwestern Pacific Subduction Zones

Analytical seismic model for a tunnel with segmental liner buried in the half-space soil

In this paper, by using the wave function expansion (WFE) and the expansion of the cylindrical wave into plane wave (ECPW) methods, an analytical method for the dynamic response of a circular tunnel with segmental liner buried in the half-space soil to harmonic elastic waves is developed. The liner of the tunnel is supposed to consist of several segments and joints. Both the segments and joints are treated as open cylindrical shells, and the segment and joint shells together thus form an equivalent continuous shell (ECS) liner, and the thin shell theory is utilized to describe its vibration. The scattered wave field due to the presence of the tunnel and boundary of the half-space soil is divided into two parts, namely, the direct scattered wave field and secondary scattered wave field. The expressions for the direct scattered cylindrical waves in the soil is determined via the WFE method, while the secondary scattered waves form the boundary of the half-space soil are obtained by the ECPW method together with the application of the boundary condition along the soil surface. Applying the cylindrical shell theory on the ECS liner and using the Fourier series expansions for the variables and parameters of the ECS liner along the azimuthal direction together with introducing the Fourier component constitutive relation for the ECS liner, a system of equations for the Fourier component ECS displacements are derived with the differential equations for the ECS liner displacements. By using the system of equations, the expressions for the free and scattered wave fields and the continuity conditions at the interface between the liner and soil, the system of equations for the ECS liner coupled with the soil are derived. By using the developed analytical method for the tunnel, some results of the ECS tunnel under different incident waves are given.

Estimating spatio-temporal variable parameters of Epidemic Type Aftershock Sequence model in a region with limited seismic network coverage: a case study of the East African Rift System

SUMMARY The Epidemic-Type Aftershock Sequence (ETAS) model is currently the most powerful statistical seismicity model that reproduces the general characteristics of earthquake clustering in space and time. However, its application can be hampered by biased parameter estimations related to earthquake catalogue deficiencies, particularly in regions where the spatial coverage of local recording networks is relatively poor. Here, we systematically investigate the possible influences of the effect introduced by data truncation through the choice of the cut-off magnitude (${{m}_{\rm cut}})$ and missing events due to heterogeneity of the seismic network on ETAS parameter estimates along the East African Rift System (EARS). After dividing the region into six source zones based on rheological and mechanical behaviours, the ETAS model is fitted to the earthquakes within each zone using the Davidon–Fletcher–Powell optimization algorithm. The fits and variations in parameter estimates are compared for each zone to the others and the seismological implications are discussed. We found that some parameters vary as a function of ${{m}_{\rm cut}}$ primarily driven by changes in catalogue size. Additionally, a systematic regional dependency of ETAS parameters is found across source zones. Furthermore, a median heat flow value for each analysed source zone in the EARS is calculated. In contrast to previous findings in other tectonic settings, the results reveal no significant correlations between the crustal heat flows and the ETAS parameters describing earthquake productivity (${{K}_0}$) and the relative efficiency of an earthquake with magnitude M to produce aftershocks ($\alpha $). Our findings have significant implications for understanding the mechanisms of earthquake interaction and, therefore, provide tight constraints on the model's parameters that may serve as a testbed for existing earthquake forecasting models in this region where the vulnerability of local buildings and structures exacerbate seismic risk.

Study on seismic failure mechanism and damage model of unconstrained adobe block walls

Through the quasi-static cyclic in-plane loading test of three kinds of unconstrained adobe block walls, namely plain adobe wall, modified mud adobe wall and modified adobe wall, the failure mode and seismic performance of each specimen were clarified. Using residual deformation and cumulative energy dissipation as failure parameters, a two-parameter seismic damage model of unconstrained adobe block walls was established. and a calculation method for the adjustment coefficients of energy consumption items was provided. The results showed that the failure mode of the typical unconstrained adobe block wall under earthquake action was not affected by the modification of adobe materials, and the final failure mode was shear failure along the mud crack and the inclined crack extending through the block. Furthermore, the proposed seismic damage model can well reflect the evolution law of earthquake damage of adobe block walls, and the recommended damage index boundary values for five performance levels were given, providing a reference for performance-based seismic design and damage assessment of adobe block walls.

Lithospheric resistivity structure of the Xuefeng Orogenic Belt and its implications for intracontinental deformation in South China

The Xuefeng Orogenic Belt (XFOB), located in the central part of the South China Block, is a typical Mesozoic intracontinental orogen in the central Jiangnan Orogenic Belt. By collecting magnetotelluric (MT) data across the XFOB, we obtained the resistivity structure of the lithosphere, which sheds light on the Mesozoic intracontinental orogenic processes in the XFOB. The resistivity structure reveals a low-resistivity body (<10 Ω∙m), beneath the XFOB, dipping southeast wards from a depth of 10 km to the bottom of the crust. This conductor is interpreted as a relic of the lower detachment zone, which coincides with low-density areas obtained from joint inversion of seismic models. It is believed to result from mineral fluids migrating along the thrust fault and squeezing sulfides into folds. Four low-resistivity bodies were identified at three extensional locations along the Jiangshan-Shaoxing Fault and at the Cili-Baojing Fault. The low-resistivity body (<10 Ω∙m) at the junction of the Shaoyang and the Hengyang Basin is located at the point where the Moho depth thins. The variation trend of the terrestrial heat flow values, with this low-resistivity body as the plate boundary, is consistent with the average variation of the terrestrial heat flow values within the block. We propose that the low-resistivity body under the Qidong-Yongzhou-Guilin fault conforms to the characteristics of the suture zone in the resistivity structure. Its existence indicates that the missing location of the Jiangshan-Shaoxing suture zone of the Yangtze and Cathaysia Block in the middle-southwest section of the South China Block is the Qidong-Yongzhou-Guilin fault. The Yangtze Block and the Hengyang Basin show high resistivity, the depth of which reaches 100 km and 40 km, respectively. Based on the resistivity model and geological data, the XFOB experienced Triassic compression, leading to basement decollement, thrusting, and nappe structures due to low-angle Paleo-Pacific Plate subduction. This compression also led to the uplift of the orogenic belt. Moreover, under the tension caused by the high-angle retreat of the Paleo-Pacific Plate, the Cretaceous extensional tectonics led to detachment along the thrust faults, forming half-graben and basin structures along the margins.

Postseismic processes in the region of the july 29, 2021 Chigni earthquake, Alaska. Part I: modeling results

We analyze the postseismic processes in the region of the Mw 8.2 Chignik earthquake, which occurred on July 29, 2021. Using the seismic rupture surface model constructed in our previous paper (Konvisar et al., 2023), we have simulated the viscoelastic relaxation process. The results of the simulation have shown that reducing the viscosity of the asthenosphere to 1018 Pa ‧ s in the calculations gives displacement velocities close to those recorded at the GPS coastal points. However, the displacements on islands close to the earthquake source region differ significantly not only in magnitude but also in direction. At the same time, the constructed model of the postseismic creep is closely consistent with the GPS displacement data and with the LOS (line-of-sight) displacement map derived from radar images acquired from the descending orbit of Sentinel 1A satellite. We also analyze temporal variations of the gravity field in the earthquake region. The obtained coseismic anomaly agrees with the anomaly calculated from the rupture surface model. Due to the insufficiently long series of gravity models after the earthquake, it is not yet possible to isolate the postseismic anomaly. The analysis of the postseismic processes is continued in the second part of this work (Smirnov et al., 2024), where we present the compare the time evolution of the postseismic displacements of various GPS sites with the aftershock activity, which allows us to draw conclusions about the creep nature of the postseismic processes in the source region of the Chignik earthquake.

Decoding Complexities: Seismic, Geological, and Dynamic Modeling of Fault and Reef Influence on Bubble Point Variations

Decoding Complexities: Seismic, Geological, and Dynamic Modeling of Fault and Reef Influence on Bubble Point Variations

Constraints on the Fate of Delaminated Lithosphere in the Upper and Mid‐Mantle

AbstractDelamination of lower continental lithosphere is known to have occurred under different tectonic settings. However, its fate in the mantle is poorly understood. By analyzing global seismic models, we find that most of likely lithosphere that delaminated during the Cenozoic and Mesozoic is preserved in the mantle transition zone, especially beneath North America and Africa. Numerical experiments indicates that delaminated lithosphere can remain stagnant in the mantle transition zone for tens of millions of years, followed by its potential sinking into the lower mantle or re‐rising to shallower depths depending on its density, the Clapeyron slope of the spinel‐to‐post‐spinel phase change and increase in mantle viscosity at ∼660–1,000 km depths. Re‐ascent occurs when delaminated lithosphere is reheated so that its effective density becomes lower than its surrounding ambient mantle after ∼100 Myr. Delaminated fragments can also potentially be mobilized by underlying global mantle flow to move horizontally away from plume regions.

Density structures of the upper mantle in the East African Rift System: implications for the evolution of intracontinental rifting

The East African Rift System (EARS) provides an ideal natural laboratory for studying the mechanisms of tectonic plate breakup and continental drift, as well as a unique perspective for exploring the maturation process of continental rifting and its drivers. This study combines high-resolution satellite gravity data and seismic tomography model with an integrated geophysical approach to reveal the density structures in the upper mantle of the EARS. The results show that the northeastern to central Congo and Zimbabwe Craton exhibit significant high-density anomalies extending up to 250 km, which is indicative of a thicker and more intact lithosphere. In contrast, the Uganda, Tanzania, eastern and southern Congo, and Kaapvaal Craton show shallow high-density anomalies underlain by low-density anomalies that are clearly derived from the deeper mantle, indicating a thining of the lithosphere with some degree of possible melting at the base. The various rift segments of the EARS exhibit different rift morphologies. The Main Ethiopian Rift and the Kenya Rift of the Eastern Rift Branch show strong low-density anomalies, indicating intense melting, which is much stronger than that observed in the Western Rift Branch. However, the two rifts may have originated from the same mantle uplift in which the low-density anomalies of the Eastern and Western Rift Branches connected in the deep upper mantle. The lower portion of the Malawi Rift exhibits weaker low-denstiy anomalies, which can be observed to the south of the Malawi Rift, extending further south as a continuation of the EARS. Combining the results of previous kinetics simulations and our density perturbation results, it can be inferred that the Eastern Rift Branch is mainly affected by active rifting, while the Western Rift Branch is affected by both active and passive rifting.

Acoustic wave simulation in strongly heterogeneous models using a discontinuous Galerkin method

In recent years, the discontinuous Galerkin method (DGM) has been rapidly developed for the numerical simulation of seismic waves. For wavefield propagation between two adjacent elements, it is common practice to apply a numerical flux to the boundary of each element to propagate waves between adjacent elements. Several fluxes, including the center, penalty, local Lax-Friedrich (LLF), upwind, and Rankine-Hugoniot jump condition-based (RH condition) fluxes, are widely used in numerical seismic wave simulation. However, some fluxes do not account for media differences between adjacent elements. Although different fluxes have been successfully used in DGM for many velocity models, it is unclear whether they can produce sufficiently accurate or stable results for strongly heterogeneous models, such as checkerboard models commonly used in tomographic studies. We test different fluxes using the acoustic wave equation. We analyze the accuracy of the penalty, LLF, upwind, and RH-condition fluxes based on the results of the numerical simulations of the homogeneous and two-layer models. We conduct simulations using checkerboard models, and the results indicate that the LLF, penalty, and upwind fluxes may have instability problems in heterogeneous models with long-time simulations. We observe instability issues in the LLF, penalty, and upwind fluxes when the wave-impedance contrast is high at the media interface. However, the results of the RH-condition flux remain consistently stable. The series of numerical examples presented in this work provide insights into the characteristics and application of fluxes for seismic wave modeling.

Seismic Reservoir Characterization With Implicit Neural Representations

AbstractSeismic reservoir characterization refers to a set of techniques aimed at estimating static and/or dynamic physical properties of subsurface geological formations from seismic data, with applications ranging from exploration and development of energy resources, geothermal production, carbon capture and storage, and the assessment of geological hazards. Extracting quantitative information from seismic recordings, such as an acoustic impedance model, is however a highly ill‐posed inverse problem due to the band‐limited and noisy nature of the seismic data. As such, seismic inversion techniques strongly rely on additional prior information that penalizes (or promotes) some features in the recovered model. This paper introduces IntraSeismic, a novel hybrid seismic inversion method that combines implicit neural representations as an effective way to parameterize a subsurface model with the physics of the seismic modeling operator. We demonstrate its effectiveness in both pre‐stack and 3D/4D post‐stack seismic inversion using synthetic and field data sets. Key features of IntraSeismic are (a) unparalleled performance in 2D and 3D pre‐stack/post‐stack dynamic and static seismic inversion, (b) rapid convergence rates, (c) ability to seamlessly include hard constraints (i.e., well data) and perform uncertainty quantification, and (d) potential data compression and fast randomized access to specific portions of the inverted model.

Post-earthquake damage probability assessment of high-speed railway simply supported bridge based on traffic capacity

Railway transportation is an important channel for emergency rescue and disaster relief. Traditionally, the method of evaluating the damage degree of bridges under different seismic intensities from the perspective of structural damage cannot be directly used to judge the traffic capacity of trains on bridges after earthquakes. It is impossible to guarantee the running safety and affect the transportation of materials after the earthquake. Because the damage state of track structure directly affects the running safety of trains, this study combines the relationship between the damage of track structure and the deformation of bridge structure after earthquake, and puts forward the transverse relative displacement of girder end (TRDGE) as the damage index of train running function of railway bridge after earthquake. Then, the OpenSees-TRBF (Train-Railway-Bridge-Foundation Coupled System Dynamic Analysis Platform) co-simulation method is used to carry out the train-track-bridge coupling dynamic calculation under different seismic damage states, and the running safety analysis database under different TRDGEs is established. Then, the Particle Swarm Optimization (PSO) algorithm is used to classify the bridge damage index of traffic capacity through different target safety speeds. Based on the probabilistic seismic demand model, the exceedance probability curve of traffic capacity damage under different seismic intensities is established. Finally, based on the curve, the running safety risk domain of trains on the bridge under different seismic intensities is obtained with a confidence level of 95 %. This study proposes a post-earthquake damage assessment method for railway bridges based on the assurance of running safety probabilities, evaluating the damage degree of the bridge with the residual traffic capacity of the bridge after the earthquake. This study can be applied to assess earthquake damage on existing railway bridges and provide references for the safe speed of trains on the bridge after an earthquake.

Decoupled approximate qP‐ and qSV‐wave equations in attenuated transversely isotropic media

AbstractAccurate seismic models with anisotropy and attenuation characteristics are crucial to accurately imaging subsurface structures. However, the anisotropic viscoelastic equations are complex and require significant computational resources. In addition, the single‐mode waves have been sufficient for most practical exploration needs. However, separating the qP‐ and qSV‐waves in anisotropic viscoelastic wavefields is challenging. Thus, we propose a new method to approximate and efficiently separate the qP‐ and qSV‐waves in attenuated transversely isotropic media. First, we obtain the decoupled approximate phase velocities of qP‐ and qSV‐waves by a curve‐fitting method. Consequently, based on the average and maximum relative error analysis, our approximate qP‐ and qSV‐wave phase velocities are more accurate than the existing approximations. Additionally, our approximations have broader applicability, resulting in acceptable errors during their application. Second, based on the approximate qP‐ and qSV‐wave phase velocities, we derive the corresponding qP‐ and qSV‐wave equations for a complete decoupling of the qP‐ and qSV‐wave components in transversely isotropic media. Third, to combine the attenuation and anisotropy characteristics, we incorporate the Kelvin–Voigt attenuation model and obtain the decoupled qP‐ and qSV‐wave equations in attenuated transversely isotropic media. Then, we use an efficient and stable hybrid finite‐difference and pseudo‐spectral method to solve the new decoupled qP‐ and qSV‐wave equations. Finally, several numerical examples demonstrate the separability and high accuracy of the proposed qP‐ and qSV‐wave equations. We obtain a qP‐wave wavefield entirely devoid of SV‐wave artefacts. In addition, the decoupled approximate qP‐ and qSV‐wave equations are accurate and stable in heterogeneous media with different velocities and attenuation. The decoupled, approximated qP‐wave and qSV‐wave equations proposed in this paper can effectively separate the qP‐wave and qSV‐wave components, resulting in fully decoupled qP‐ and qSV‐wave wavefields in attenuated transversely isotropic media.

On accounting for the effects of crust and uppermost mantle structure in global scale full-waveform inversion

Summary Fundamental mode surface wave data have often been used to construct global shear velocity models of the upper mantle under the so-called “path average approximation”, an efficient approach from the computational point of view. With the advent of full-waveform inversion and numerical wavefield computations, such as afforded by the spectral element method, accounting for the effects of the crust accurately becomes challenging. Here, we assess the merits of accounting for crustal and uppermost mantle effects on surface and body waveforms using fundamental mode dispersion data and a smooth representation of the shallow structure. For this we take as reference a model obtained by full waveform inversion and wavefield computations using the spectral element method, model SEMUCB-WM1 (French and Romanowicz, 2014) and compare the waveform fits of synthetics to different parts of three component observed teleseismic records, in the period band 32-300 s for body waves and 40-300 s for surface waves and their overtones for three different models. The latter are: a dispersion-only based model (model Disp_20s_iter5), and two models modified from SEMUCB-WM1 by successively replacing the top 200 km (model Merged _200km) and top 80 km (model Merged _80km), respectively, by a model constrained solely by fundamental mode surface wave dispersion data between periods of 20 and 150 s. The crustal part of these 3 models (resp. SEMUCB-WM1) is constrained from dispersion data in the period range 20-60 s (resp. 25-60 s), using the concept of homogenization (e.g., Backus 1962, Capdeville &amp; Marigo 2007) which is tailored to simplify complex geological features, enhancing the computational efficiency of our seismic modeling. We evaluate the fits to observed waveforms provided by these 3 models compared to those of SEMUCB-WM1 by computing three component synthetics using the spectral element method for 5 globally distributed events recorded at 200+ stations, using several measures of misfit. While fits to waveforms for model 3 are similar to those for SEMUCB-WM1, the other two models provide increasingly poorer fits as the distance travelled by the corresponding seismic wave increases and/or as it samples deeper in the mantle. In particular, models 1 and 2 are biased towards fast shear velocities, on average. Our results suggest that, given a comparable frequency band, models constructed using fundamental mode surface wave data alone and the path average approximation, fail to provide acceptable fits to the corresponding waveforms. However, the shallow part of such a 3D radially anisotropic model can be a good starting model for further full waveform inversion using numerical wavefield computations. Moreover, the shallow part of such a model, including its smooth crustal model, and down to a maximum depth that depends on the frequency band considered, can be fixed in FWI iterations for deeper structure. This can save significant computational time when higher resolution is sought in the deeper mantle. In the future, additional constraints for the construction of the homogenized model of the crust can be implemented from independent short period studies, either globally or regionally.