Combining discrete cosine transform and convolutional neural networks to speed up the Hamiltonian Monte Carlo inversion of pre-stack seismic data     

Mattia Aleardi 《Geophysical Prospecting》2020,68(9):2738-2761

Markov chain Monte Carlo algorithms are commonly employed for accurate uncertainty appraisals in non-linear inverse problems. The downside of these algorithms is the considerable number of samples needed to achieve reliable posterior estimations, especially in high-dimensional model spaces. To overcome this issue, the Hamiltonian Monte Carlo algorithm has recently been introduced to solve geophysical inversions. Different from classical Markov chain Monte Carlo algorithms, this approach exploits the derivative information of the target posterior probability density to guide the sampling of the model space. However, its main downside is the computational cost for the derivative computation (i.e. the computation of the Jacobian matrix around each sampled model). Possible strategies to mitigate this issue are the reduction of the dimensionality of the model space and/or the use of efficient methods to compute the gradient of the target density. Here we focus the attention to the estimation of elastic properties (P-, S-wave velocities and density) from pre-stack data through a non-linear amplitude versus angle inversion in which the Hamiltonian Monte Carlo algorithm is used to sample the posterior probability. To decrease the computational cost of the inversion procedure, we employ the discrete cosine transform to reparametrize the model space, and we train a convolutional neural network to predict the Jacobian matrix around each sampled model. The training data set for the network is also parametrized in the discrete cosine transform space, thus allowing for a reduction of the number of parameters to be optimized during the learning phase. Once trained the network can be used to compute the Jacobian matrix associated with each sampled model in real time. The outcomes of the proposed approach are compared and validated with the predictions of Hamiltonian Monte Carlo inversions in which a quite computationally expensive, but accurate finite-difference scheme is used to compute the Jacobian matrix and with those obtained by replacing the Jacobian with a matrix operator derived from a linear approximation of the Zoeppritz equations. Synthetic and field inversion experiments demonstrate that the proposed approach dramatically reduces the cost of the Hamiltonian Monte Carlo inversion while preserving an accurate and efficient sampling of the posterior probability.  相似文献   

Upper ocean high resolution regional modeling of the Arabian Sea and Bay of Bengal     

Suneet Dwivedi  Alok Kumar Mishr  Atul Srivastava 《海洋学报(英文版)》2019,38(5):32-50

In this paper, effort is made to demonstrate the quality of high-resolution regional ocean circulation model in realistically simulating the circulation and variability properties of the northern Indian Ocean(10°S–25°N,45°–100°E) covering the Arabian Sea(AS) and Bay of Bengal(BoB). The model run using the open boundary conditions is carried out at 10 km horizontal resolution and highest vertical resolution of 2 m in the upper ocean.The surface and sub-surface structure of hydrographic variables(temperature and salinity) and currents is compared against the observations during 1998–2014(17 years). In particular, the seasonal variability of the sea surface temperature, sea surface salinity, and surface currents over the model domain is studied. The highresolution model's ability in correct estimation of the spatio-temporal mixed layer depth(MLD) variability of the AS and BoB is also shown. The lowest MLD values are observed during spring(March-April-May) and highest during winter(December-January-February) seasons. The maximum MLD in the AS(BoB) during December to February reaches 150 m (67 m). On the other hand, the minimum MLD in these regions during March-April-May becomes as low as 11–12 m. The influence of wind stress, net heat flux and freshwater flux on the seasonal variability of the MLD is discussed. The physical processes controlling the seasonal cycle of sea surface temperature are investigated by carrying out mixed layer heat budget analysis. It is found that air-sea fluxes play a dominant role in the seasonal evolution of sea surface temperature of the northern Indian Ocean and the contribution of horizontal advection, vertical entrainment and diffusion processes is small. The upper ocean zonal and meridional volume transport across different sections in the AS and BoB is also computed. The seasonal variability of the transports is studied in the context of monsoonal currents.  相似文献   

厚层切叠砂体自然层描述方法:以北一区断东葡一组1-4小层为例     

吴迪  马世忠  景丽娟  周志国 《世界地质》2020,39(1):106-112

将多期切叠河道砂体归类合并,建立一个“垂向连续,横向联通的表外砂岩空间体”(其中砂岩间夹层厚度≤0.4 m)的自然层概念来控制多期河道复合切叠厚砂体。以北一匹断东萄一组1-4小层为例,利用自然层间砂体厚度、切叠程度、测井曲线形态、相叠加类型及砂体叠加期次将自然层分为5类;再依据砂体间切叠位置、切叠程度和切叠形态的差异建立自然层剖面表征方法;依据砂体叠加期次,建立自然层在平面上表征模式。  相似文献   

Research note: deblended-data reconstruction using generalized blending and deblending models     

Tomohide Ishiyama  Mohammed Y. Ali  Satoshi Ishikawa  Gerrit Blacquiere  Shotaro Nakayama 《Geophysical Prospecting》2019,67(7):1852-1866

We introduce a concept of generalized blending and deblending, develop its models and accordingly establish a method of deblended-data reconstruction using these models. The generalized models can handle real situations by including random encoding into the generalized operators both in the space and time domain, and both at the source and receiver side. We consider an iterative optimization scheme using a closed-loop approach with the generalized blending and deblending models, in which the former works for the forward modelling and the latter for the inverse modelling in the closed loop. We applied our method to existing real data acquired in Abu Dhabi. The results show that our method succeeded to fully reconstruct deblended data even from the fully generalized, thus quite complicated blended data. We discuss the complexity of blending properties on the deblending performance. In addition, we discuss the applicability to time-lapse seismic monitoring as it ensures high repeatability of the surveys. Conclusively, we should acquire blended data and reconstruct deblended data without serious problems but with the benefit of blended acquisition.  相似文献   

Some remarks on Q-compensated sparse deconvolution without knowing the quality factor Q     

Xintao Chai  Ronghua Peng  Genyang Tang  Wei Chen  Jingnan Li 《Geophysical Prospecting》2019,67(8):2003-2021

The subsurface media are not perfectly elastic, thus anelastic absorption, attenuation and dispersion (aka Q filtering) effects occur during wave propagation, diminishing seismic resolution. Compensating for anelastic effects is imperative for resolution enhancement. Q values are required for most of conventional Q-compensation methods, and the source wavelet is additionally required for some of them. Based on the previous work of non-stationary sparse reflectivity inversion, we evaluate a series of methods for Q-compensation with/without knowing Q and with/without knowing wavelet. We demonstrate that if Q-compensation takes the wavelet into account, it generates better results for the severely attenuated components, benefiting from the sparsity promotion. We then evaluate a two-phase Q-compensation method in the frequency domain to eliminate Q requirement. In phase 1, the observed seismogram is disintegrated into the least number of Q-filtered wavelets chosen from a dictionary by optimizing a basis pursuit denoising problem, where the dictionary is composed of the known wavelet with different propagation times, each filtered with a range of possible values. The elements of the dictionary are weighted by the infinity norm of the corresponding column and further preconditioned to provide wavelets of different values and different propagation times equal probability to entry into the solution space. In phase 2, we derive analytic solutions for estimates of reflectivity and Q and solve an over-determined equation to obtain the final reflectivity series and Q values, where both the amplitude and phase information are utilized to estimate the Q values. The evaluated inversion-based Q estimation method handles the wave-interference effects better than conventional spectral-ratio-based methods. For Q-compensation, we investigate why sparsity promoting does matter. Numerical and field data experiments indicate the feasibility of the evaluated method of Q-compensation without knowing Q but with wavelet given.  相似文献   

基于新型高频声学多普勒流速剖面仪的河口近底部边界层观测初探     

兰庭飞  吴昊  唐杰平  李占海  汪亚平 《海洋通报》2019,38(6):640-649

河口物质输运、能量交换与底边界层内的水动力过程密切相关,底边界层参数(如切应力、拖曳系数)的确定至关重要。挪威Nortek公司生产的新型声学多普勒流速剖面仪AD2CP相比传统ADCP具有高频、低噪的优点,可用于高频(16Hz)流速剖面观测,而被广泛应用于底边界层观测的ADV只能测量单点高频流速。本文采用AD2CP在长江口南槽最大浑浊带区域进行座底式观测,并与同步近底部三脚架上ADV的观测结果进行对比。结果表明,使用AD2CP测得的近底部平均流速与ADV的测量结果吻合良好;使用惯性耗散法计算了底切应力,基于ADV的单点高频流速数据计算结果为2.16×10~(-2)～5.69×10~(-1)N/m~2,基于AD2CP的结果为2.09×10~(-2)～4.26×10~(-1)N/m~2,二者范围大致相当。在此基础上,基于AD2CP数据计算出摩阻流速为4.55×10~(-3)～2.06×10~(-2)m/s、底拖曳系数范围为1.84×10~(-4)～2.49×10~(-3),与ADV的计算结果基本一致。此外,由于AD2CP可以获得高频的流速剖面数据,优于单点ADV,具备观测近底部边界层参数和边界层内湍流剖面的潜力。  相似文献   

p-y曲线对成层土体中大直径单桩的适用性研究          下载免费PDF全文

孟晓伟  翟恩地  许成顺 《海洋技术学报》2019,38(2)

API规范推荐的p-y曲线是由均质土体得到的,并未考虑土层间相互作用,Georgiadis基于柔性桩提出了等效深度法修正p-y曲线,把均质土p-y曲线延伸到了成层土体中。为了研究p-y曲线和等效深度法对于大直径单桩在成层土体中的适用性,取4种典型地质条件:成层砂土、砂土-黏土-砂土、成层黏土和黏土-砂土-黏土,通过L-PILE软件计算了6 m直径单桩基础的p-y曲线、桩顶水平荷载-位移曲线、桩身位移和弯矩。并与ABAQUS建立的单桩基础三维有限元模型计算结果进行比较。结果表明等效深度法对于成层砂土影响不大;对于成层黏土影响较大;对于中间为软弱土层的成层土体,在荷载较大时影响显著,等效深度法计算结果更加接近FEM结果。在成层土体中,p-y曲线应用于大直径单桩对于砂土高估了初始刚度,低估了极限抗力;对于黏土则低估了初始刚度和极限抗力。  相似文献   

番红砗磲外套膜色素提取条件优化及成分分析     

付正祎  刘二田  顾志峰  石耀华  王爱民  吴川良  刘春胜 《海洋科学》2019,43(6):71-77

以番红砗磲(Tridacna crocea)为实验对象,通过单因素对比实验对其外套膜色素的提取条件进行了优化,并利用薄层层析色谱法对色素组成进行了初步分析。结果表明:丙酮和乙醇为适宜的番红砗磲外套膜色素提取剂;使用丙酮为提取剂时,浸提时间为4 h,提取温度为25℃,料液比为1︰40。使用乙醇为提取剂时,浸提时间为4 h,提取温度为30℃,料液比为1︰200;利用薄层色谱法分离外套膜颜色为蓝色和红棕色的番红砗磲外套膜色素后发现,蓝色番红砗磲外套膜色素由9种色素组分构成,红棕色个体番红砗磲外套膜色素由10种色素组分构成。  相似文献   

Scattering pattern analysis and true-amplitude generalized Radon transform migration for acoustic transversely isotropic media with a vertical axis of symmetry     

Quan Liang  Wei Ouyang  Weijian Mao  Shijun Cheng 《Geophysical Prospecting》2020,68(4):1211-1227

In multi-parameter ray-based anisotropic migration/inversion, it is essential that we have an understanding of the scattering mechanism corresponding to parameter perturbations. Because the complex nonlinearity in the anisotropic inversion problem is intractable, the construction of true-amplitude linearized migration/inversion procedures is needed and important. By using the acoustic medium assumption for transversely isotropic media with a vertical axis of symmetry and representing the anisotropy with P-wave normal moveout velocity, Thomsen parameter δ and anelliptic parameter η, we formalize the linearized inverse scattering problem for three-dimensional pseudo-acoustic equations. Deploying the single-scattering approximation and an elliptically anisotropic background introduces a new linear integral operator that connects the discontinuous perturbation parameters with the multi-shot/multi-offset P-wave scattered data. We further apply the high-frequency asymptotic Green's function and its derivatives to the integral operator, and then the scattering pattern of each perturbation parameter can be explicitly presented. By naturally establishing a connection to generalized Radon transform, the pseudo-inverse of the integral operator can be solved by the generalized Radon transform inversion. In consideration of the structure of this pseudo-inverse operator, the computational implementation is done pointwise by shooting a fan of rays from the target imaging area towards the acquisition system. Results from two-dimensional numerical tests show amplitude-preserving images with high quality.  相似文献   

Separation of PP- and PS-wave reflected seismic data using two-dimensional finite offset common-reflection-surface traveltime approximation     

Raphael Di Carlo Silva dos Santos  João Carlos Ribeiro Cruz  Manuel de Jesus dos Santos Costa 《Geophysical Prospecting》2020,68(7):2046-2063

Recently, the interest in PS-converted waves has increased for several applications, such as sub-basalt layer imaging, impedance estimates and amplitude-versus-offset analysis. In this study, we consider the problem of separation of PP- and PS-waves from pre-stacked multicomponent seismic data in two-dimensional isotropic medium. We aim to demonstrate that the finite-offset common-reflection-surface traveltime approximation is a good alternative for separating PP- and PS-converted waves in common-offset and common shot configurations by considering a two-dimensional isotropic medium. The five parameters of the finite-offset common-reflection-surface are firstly estimated through the inversion methodology called very fast simulated annealing, which estimates all parameters simultaneously. Next, the emergence angle, one of the inverted parameters, is used to build an analytical separation function of PP and PS reflection separation based on the wave polarization equations. Once the PP- and PS-converted waves were separated, the sections are stacked to increase the signal-to-noise ratio using the special curves derived from finite-offset common-reflection-surface approximation. We applied this methodology to a synthetic dataset from simple-layered to complex-structured media. The numerical results showed that the inverted parameters of the finite offset common-reflection-surface and the separation function yield good results for separating PP- and PS-converted waves in noisy common-offset and common shot gathers.  相似文献