From surface wave inversion to seismic site response prediction: Beyond the 1D approach

Surface wave methods consist of the extraction and inversion of the Rayleigh wave phase-velocity dispersion curve to recover the (usually 1D) shear-wave velocity profile. In the literature, uncertainty due to data error has not received much attention, but the discussion about uncertainty due to model error is even poorer. Even with an unrealistic noise-free dataset and an exact forward model, an inappropriate parameterization can generate solutions very far from the actual soil structure. In general, the model used for the dispersion curve interpretation is 1D. Hence, when the velocity distribution is laterally heterogeneous, model errors can have significant consequences on the reliability of the resulting shear-wave velocity distribution. From a poor velocity reconstruction, an unsatisfactory, and often dangerous site response analysis follows. In fact, shear wave measurements play a relevant role in seismic ground motion amplification estimation. In this paper, we discuss the possibility of processing the seismograms using a multi-offset phase analysis (MOPA), in order to derive soil elastic parameters for weak motion predictions. This technique allows the detection and location of the lateral discontinuities, and a better model parameterization. In fact, once the discontinuities are identified, we can split the profile into several, truly 1D, parts. The use of the standard 1D dispersion curve extraction and inversion for each side of the heterogeneity generates velocity profiles that we can put side by side to get correct 2D reconstructions of the shear-wave distributions. From 2D velocity reconstruction, we can calculate the site response that may be significantly different from the site response generated from a traditional 1D analysis of the same seismograms. In this work, we discuss the site responses of two synthetic examples with lateral heterogeneities. We show how misleading a 1D analysis may be if applied to a truly 2D velocity distribution, particularly in terms of site response prediction.     

PP and PS joint AVO inversion and fluid prediction

The technique of amplitude variation with offset or angle (AVO or AVA) can be used to extract fluid and lithology information from prestack seismic data. Based on three-term AVO equations, three elastic parameters can be inverted for by linear AVO inversion. However, many theoretical and numerical studies have demonstrated that by using offset limited data, a three-term AVO inversion may have problems of instability and inaccuracy while inverting for the density term. We have searched for an elastic parameter that contains density information and inverted this parameter in a more stable manner using offset limited data. First, we test the sensitivity of elastic parameters to hydrocarbon reservoirs and select the optimal fluid factor (ρf) that contains density information and has an excellent performance as an inversion parameter used to detect hydrocarbons. Then, we derive approximate PP and PS reflection coefficient equations in terms of the fluid factor. The derived equations allow us to directly estimate the fluid factor of the reservoir. Finally, we apply these equations to synthetic data by employing a joint AVO inversion technique. The results show that the method is stable and unambiguous.     

Estimation of shear wave velocity profiles by the inversion of spatial autocorrelation coefficients

The subsurface shear-wave velocity (Vs) is considered to be a key parameter for site characterization and assessment of earthquake hazard because of its great influence on local ground-motion amplification. Array microtremor measurements are widely used for the estimation of shear-wave velocities. Compared to other methods such as frequency-wavenumber (f-k) methods, the spatial autocorrelation (SPAC) method requires fewer sensors and thus is relatively easier to implement and gives robust estimations of shear-wave velocity profiles for depths down to a few hundred meters. The quantity derived from observed data is the SPAC coefficient, which is a function of correlation distance, frequency and phase velocity. Generally, estimation of Vs profiles is a two stage process: Estimation of the dispersion data from the SPAC coefficients and inversion of the dispersion data for shear-wave velocity structure. In this study, instead of inverting dispersion curves, a more practical approach is used; that is, observed SPAC coefficients are directly inverted for the S-wave velocities. A synthetic case and a field data application are presented to test the potential of the inversion algorithm. We obtain an iterative damped least-squares solution with differential smoothing. The differential smoothing approach constrains the change in shear-wave velocities of the adjacent layers and thus stabilizes the inversion.     

Direct time-domain soil profile reconstruction for one-dimensional semi-infinite domains

We discuss the inverse medium problem associated with the reconstruction of the heterogeneous material profile of a semi-infinite (layered) soil medium, directly in the time domain, based on the complete waveform response of the medium to interrogating waves. To tackle the inversion process, we use a partial-differential-equation-constrained optimization approach, supplemented with a time-dependent regularization scheme. We introduce an absorbing boundary to truncate the semi-infinite extent of the physical domain, and propose two schemes to refine the reconstructed profiles: the first is based on iteratively re-positioning the truncation boundary until convergence, and the second is based on optimizing the observation period, so as to exclude records with information beyond the truncation boundary. We present numerical results that attest to the efficacy of the proposed schemes in reconstructing sharp profiles of semi-infinite soil domains using both noise-free and noisy data, while in the presence of absorbing boundaries.     

Global Optimisation of Time Domain Electromagnetic Data Using Very Fast Simulated Annealing

--Efficacy of the global optimisation technique is demonstrated in the inversion of time domain electromagnetic data. Transient EM responses observed using a coincident loops system along several profiles and different time channels over a plate like conducting body are inverted, using very fast simulated annealing (VFSA) as an optimisation tool. Three time channels for each profile are considered in the inversion. Study reveals that only one run of global inversion considering a single multi-channel profile is not enough to resolve all the nine model parameters of a plate-like conducting body. However, the global inversion of a single multi-channel profile with several runs yields a mean model that is quite close to the true model. Considering many profiles and time channels together in the global inversion can yield reliable estimates of all the parameters. Computationally, this is not an efficient procedure. Analysis of the results shows, however, that two distant profiles are enough to yield all the model parameters reliably, even after one run of the global inversion. Both noise-free and noisy synthetic data are used in the inversion. Finally, field data are also inverted to study the performance of the global inversion.     

航空电磁拟三维模型空间约束反演

为了克服时间域航空电磁数据单点反演结果中常见的电阻率或层厚度横向突变造成数据难以解释的问题,通过引入双向约束实现航空电磁拟三维空间约束反演.除考虑沿测线方向相邻测点之间的横向约束外,同时还考虑了垂直测线方向测点在空间上的相互约束.为此,首先设计拟三维模型中固定层厚和可变层厚两种空间约束反演方案,然后通过在目标函数中引入沿测线和垂直测线方向上的模型参数约束矩阵,并使用L-BFGS算法使目标函数最小化,获得最优拟三维模型空间反演解.基于理论模型和实测数据反演,对单点反演与两种空间约束反演方案的有效性进行比较,证明本文空间约束反演算法对于噪声的压制效果好,反演的界面连续光滑,同时内存需求和反演时间少,是一种快速有效的反演策略.     

Simultaneous time-lapse electrical resistivity inversion

Time-lapse monitoring is a powerful tool for observing dynamic changes in the subsurface. In particular it offers the potential for achieving inversion results with increased fidelity through the inclusion of complementary information from multiple time-steps. This inclusion of complementary information can reduce the need for spatial smoothing, without adding inversion artifacts to the resulting images. Commonly used time-lapse inversion methods include the ratio method, cascaded time-lapse inversion, difference inversion and differencing independent inversions. We introduce two additional methods in which both time-lapse data sets are inverted simultaneously. In the first, called temporally constrained time-lapse inversion, inversion of both datasets is done under a single optimization procedure and constraints are added to the regularization to ensure that the changes from one time to another are smooth. In the second method, called simultaneous time-lapse inversion, the inversions at time 1 and time 2 are performed simultaneously and constraints of smoothness and closeness to a reference model are applied to the difference image produced at each iteration, and subsequently, the constraints are updated at each iteration. Through both a numerical and a field example we compare the results of common time-lapse inversion methods as well as the introduced approaches. We found that of the commonly used time-lapse inversion methods the difference inversion method produced the best resolution of time-lapse changes and was the most robust in the presence of noise. However, we found that the alternative approach of simultaneous time-lapse inversion produced the best reconstruction of modeled EC changes in the numerical example and easily interpretable high resolution difference images in the field example. Moreover, there was less tailoring of regularization parameters with our simultaneous time-lapse approach, suggesting that it will lend itself well to an automated inversion code.     

利用大地电磁三维反演方法获得二维剖面附近三维电阻率结构的可行性

大地电磁野外实测数据月前大多为二维剖面数据.如何反演这些二维剖面数据获得较为接近实际地电情况的结果,是多数大地电磁工作者关心的问题.我们通过对理论模型的三维响应进行分析和对合成数据及实测资料的反演结果进行对比研究,讨论了利用三维反演的方法来获得大地电磁二维剖面附近三维电阻率结构的可行性.结果表明:可用三维反演的方法来解...     

Feasibility of waveform inversion of Rayleigh waves for shallow shear-wave velocity using a genetic algorithm

Conventional surface wave inversion for shallow shear (S)-wave velocity relies on the generation of dispersion curves of Rayleigh waves. This constrains the method to only laterally homogeneous (or very smooth laterally heterogeneous) earth models. Waveform inversion directly fits waveforms on seismograms, hence, does not have such a limitation. Waveforms of Rayleigh waves are highly related to S-wave velocities. By inverting the waveforms of Rayleigh waves on a near-surface seismogram, shallow S-wave velocities can be estimated for earth models with strong lateral heterogeneity. We employ genetic algorithm (GA) to perform waveform inversion of Rayleigh waves for S-wave velocities. The forward problem is solved by finite-difference modeling in the time domain. The model space is updated by generating offspring models using GA. Final solutions can be found through an iterative waveform-fitting scheme. Inversions based on synthetic records show that the S-wave velocities can be recovered successfully with errors no more than 10% for several typical near-surface earth models. For layered earth models, the proposed method can generate one-dimensional S-wave velocity profiles without the knowledge of initial models. For earth models containing lateral heterogeneity in which case conventional dispersion-curve-based inversion methods are challenging, it is feasible to produce high-resolution S-wave velocity sections by GA waveform inversion with appropriate priori information. The synthetic tests indicate that the GA waveform inversion of Rayleigh waves has the great potential for shallow S-wave velocity imaging with the existence of strong lateral heterogeneity.     

Shallow velocity–depth model imaging by refraction tomography1

A tomographic imaging technique combined with coherence inversion is proposed for constructing a near-surface model from refraction events. A model obtained from coherence inversion serves as a good background model for the tomographic reconstruction. A simultaneous iterative reconstruction technique (SIRT) algorithm was used for this purpose. This is a simple algorithm and can be easily adapted to irregular acquisition geometry and limited angular aperture. Using synthetic data it was shown that the proposed procedure can be used for determination of local velocity anomalies in a shallow subsurface. The technique was also tested on a real data set.     

简化混合域全波形反演多GPU加速策略

全波形反演利用地震记录中的振幅、走时和相位等信息,通过拟合实际地震记录和计算波场来定量提取地下介质的弹性参数,进而为勘探地震成像、速度建模以及大尺度构造演化分析等提供可靠依据.但全波形反演计算量巨大,特别是应用于三维大区块叠前数据时,生产成本仍然很高.本文介绍并比较了时间域和频率域的全波形反演方法,综合两者的优点,最终采用混合域的反演算法,并且在此基础上做了进一步的简化以提高计算效率.针对全波形反演方法应用于大规模叠前数据时易陷入局部极小值的问题,我们提出对模型数据进行分割,同时在数个小模型内进行梯度搜索,然后对比各个局域的梯度,最终找出合适的全局下降方向,以克服局部极小的隐患.该方法能够充分利用GPU的硬件特性.在GPU环境下实现本文所提出的简化混合域全波形反演算法.数值计算实例体现出新方法具有良好的计算效率、反演精度和算法可扩展性.     

地下固体介质弹性模量反演

地下固体介质弹性模量是固体地球定量表征的重要参数,是探测地球内部结构、地下流体分布、设计合理工程方案的关键之一.本文首先简要分析常用的流体饱和孔隙介质岩石物理模型,在此基础上通过数值计算,详细讨论了流体饱和多孔介质等效弹性模量对固体基质、孔隙度、孔隙结构和孔隙流体的敏感性,探讨了利用等效介质弹性模量反演固体基质弹性模量的可行性,建立了固体基质弹性模量的非线性方程系统,提出了该方程系统的数值迭代求解方法,并通过对三类岩石样本实验室测量数据的反演和结果分析,验证了方法的有效性.文中同时通过不同条件下对实验数据的反演,探讨了孔隙结构、孔隙流体和等效介质弹性模量对反演效果的影响,为方法的合理有效应用提供依据.     

基于walkaway VSP下行P波的TTI各向异性参数反演

本文基于弱各向异性（WA）介质的正反演公式和qP波的坐标变换,推导了利用qP波反演任意倾斜对称轴的横向各向同性（TTI）介质的各向异性参数和对称轴方向的公式.理论和数值实验表明,利用2个相互正交的变井源距垂直地震剖面（walkaway VSP）可以完全确定钻井中TTI介质qP波的3个WA参数和对称轴的2个方向参数.我们完成了几个由不同数量剖面组成的walkaway VSP模拟实验,使用TTI模型和一般各向异性模型对模拟数据进行了反演,证明了反演公式的正确性和可靠性.使用这些公式,对来自Java Sea的由3条剖面组成的walkaway VSP观测数据进行了各向异性反演,获得了钻井中接收点处介质的WA参数.     

Nonlinear Rayleigh wave inversion based on the shuffled frog-leaping algorithm

At present, near-surface shear wave velocities are mainly calculated through Rayleigh wave dispersion-curve inversions in engineering surface investigations, but the required calculations pose a highly nonlinear global optimization problem. In order to alleviate the risk of falling into a local optimal solution, this paper introduces a new global optimization method, the shuffle frog-leaping algorithm (SFLA), into the Rayleigh wave dispersion-curve inversion process. SFLA is a swarm-intelligence-based algorithm that simulates a group of frogs searching for food. It uses a few parameters, achieves rapid convergence, and is capability of effective global searching. In order to test the reliability and calculation performance of SFLA, noise-free and noisy synthetic datasets were inverted. We conducted a comparative analysis with other established algorithms using the noise-free dataset, and then tested the ability of SFLA to cope with data noise. Finally, we inverted a real-world example to examine the applicability of SFLA. Results from both synthetic and field data demonstrated the effectiveness of SFLA in the interpretation of Rayleigh wave dispersion curves. We found that SFLA is superior to the established methods in terms of both reliability and computational efficiency, so it offers great potential to improve our ability to solve geophysical inversion problems.     

各向异性TI介质qP反射波走时层析成像

地震走时层析成像是反演地层各向异性参数分布的有效方法,但是关于地震各向异性介质走时层析成像的研究并不多,其技术远远没有达到成熟的阶段.在野外数据采集时,地表反射波观测方式相对井间和垂直地震剖面观测方式的成本更低,利用qP反射波走时反演各向异性参数具有更加广泛的实用价值.本文实现的TI介质地震走时层析成像方法结合了TI介质反射波射线追踪算法、走时扰动方程和非线性共轭梯度算法,它可以对任意强度的TI介质模型进行反演,文中尝试利用qP反射波走时重建TI介质模型的参数图像.利用qP反射波对层状介质模型和块状异常体模型进行走时反演,由于qP波相速度对弹性模量参数和Thomsen参数的偏微分不同,所以可以分别反演弹性模量参数和Thomsen参数.数值模拟结果表明:利用qP反射波可以反演出TI介质模型的弹性模量参数与Thomsen参数,不同模型的走时迭代反演达到了较好的收敛效果,与各向同性介质走时反演结果相比较,各向异性介质走时反演结果具有较好的识别能力.     

Time‐lapse seismic imaging using regularized full‐waveform inversion with a prior model: which strategy?

Full‐waveform inversion is an appealing technique for time‐lapse imaging, especially when prior model information is included into the inversion workflow. Once the baseline reconstruction is achieved, several strategies can be used to assess the physical parameter changes, such as parallel difference (two separate inversions of baseline and monitor data sets), sequential difference (inversion of the monitor data set starting from the recovered baseline model) and double‐difference (inversion of the difference data starting from the recovered baseline model) strategies. Using synthetic Marmousi data sets, we investigate which strategy should be adopted to obtain more robust and more accurate time‐lapse velocity changes in noise‐free and noisy environments. This synthetic application demonstrates that the double‐difference strategy provides the more robust time‐lapse result. In addition, we propose a target‐oriented time‐lapse imaging using regularized full‐waveform inversion including a prior model and model weighting, if the prior information exists on the location of expected variations. This scheme applies strong prior model constraints outside of the expected areas of time‐lapse changes and relatively less prior constraints in the time‐lapse target zones. In application of this process to the Marmousi model data set, the local resolution analysis performed with spike tests shows that the target‐oriented inversion prevents the occurrence of artefacts outside the target areas, which could contaminate and compromise the reconstruction of the effective time‐lapse changes, especially when using the sequential difference strategy. In a strongly noisy case, the target‐oriented prior model weighting ensures the same behaviour for both time‐lapse strategies, the double‐difference and the sequential difference strategies and leads to a more robust reconstruction of the weak time‐lapse changes. The double‐difference strategy can deliver more accurate time‐lapse variation since it can focus to invert the difference data. However, the double‐difference strategy requires a preprocessing step on data sets such as time‐lapse binning to have a similar source/receiver location between two surveys, while the sequential difference needs less this requirement. If we have prior information about the area of changes, the target‐oriented sequential difference strategy can be an alternative and can provide the same robust result as the double‐difference strategy.     

Time-domain forward and inverse modeling of lossy soils with frequency-independent Q for near-surface applications

In this paper, we are concerned with a full-waveform-based methodology that allows the simultaneous imaging of the soil's stiffness and attenuating properties, using solely the soil's surficial response to probing waves.To date, field observations of small-strain wave attenuation in geomaterials at moderate spatial scales suggest that a commonly used metric of intrinsic and apparent attenuation, the seismic quality factor Q, is frequency-independent for a wide part of the frequency spectrum, including the frequency range of interest to seismic applications. We discuss first the forward simulation of waves in near-surface soil deposits directly in the time-domain using simplified models that adequately approximate nearly frequency-independent Q. To this end, we first review various attenuation models that aim at reproducing the frequency-independent Q behavior, and conclude, supported by site analyses, that, even though a generalized Maxwell body with eight Maxwell elements in parallel (GMB8) provides the best fit to frequency-independent Q, we favor a version of it with fewer parameters (GMB2), in order to reduce modeling complexity, while still retaining good agreement with the GMB8 model.We report on forward site analyses that lend credence to the choice of the GMB2 simplified model. We, then, use the GMB2 constitutive relation in the context of full-waveform inversion, and report on numerical experiments that lead to the imaging of the soil's properties in heterogeneous semi-infinite domains.     

Direct and indirect inversions

A bridge is highlighted between the direct inversion and the indirect inversion. They are based on fundamental different approaches: one is looking after a projection from the data space to the model space while the other one is reducing a misfit between observed data and synthetic data obtained from a given model. However, it is possible to obtain similar structures for model perturbation, and we shall focus on P-wave velocity reconstruction. This bridge is built up through the Born approximation linearizing the forward problem with respect to model perturbation and through asymptotic approximations of the Green functions of the wave propagation equation. We first describe the direct inversion and its ingredients and then we focus on a specific misfit function design leading to a indirect inversion. Finally, we shall compare this indirect inversion with more standard least-squares inversion as the FWI, enabling the focus on small weak velocity perturbations on one side and the speed-up of the velocity perturbation reconstruction on the other side. This bridge has been proposed by the group led by Raul Madariaga in the early nineties, emphasizing his leading role in efficient imaging workflows for seismic velocity reconstruction, a drastic requirement at that time.     

Comparison of acoustic and elastic full-waveform inversion of 2D towed-streamer data in the presence of salt

Over the last years, full-waveform inversion has become an important tool in the list of processing and imaging technologies available to the industry. For marine towed-streamer data, full-waveform inversion is typically applied using an acoustic approximation because S-waves do not propagate in water and elastic effects in recorded data are generally assumed to be small. We compare acoustic and elastic modelling and full-waveform inversion for a field data set acquired offshore Angola over sediments containing a salt body with significant topology. Forward modelling tests reveal that such geological structures lead to significant mode conversions at interfaces and, consequently, to significant relative amplitude differences when elastically and acoustically modelled traces are compared. Using an acoustic approach for modelling in full-waveform inversion therefore leads to problems matching the synthetic data with the field data, even for recorded pressure data and with trace normalization applied. Full-waveform inversion is unable to find consistent model updates. Applying elastic full-waveform inversion leads to more consistent and reliable model updates with less artefacts, at the expense of additional computation cost. Although two-dimensional marine towed-streamer data are least favourable for the application of full-waveform inversion compared to three-dimensional data or ocean-bottom data, it is recommended to check on the existence of elastic effects before deciding on the final processing and imaging approach.