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   

Inversion-based deblending using migration operators     

Amr Ibrahim  Daniel Trad 《Geophysical Prospecting》2020,68(8):2459-2470

In this paper, we compare the denoising- and inversion-based deblending methods using Stolt migration operators. We use Stolt operator as a kernel to efficiently compute apex-shifted hyperbolic Radon transform. Sparsity promoting transforms, such as Radon transform, can focus seismic data into a sparse model to separate signals, remove noise or interpolate missing traces. Therefore, Radon transforms are a suitable tool for either the denoising- or the inversion-based deblending methods. The denoising-based deblending treats blending interferences as random noise by sorting the data into new gathers, such as common receiver gather. In these gathers, blending interferences exhibit random structures due to the randomization of the source firing times. Alternatively, the inversion-based deblending treats blending interferences as a signal, and the transform models this signal by incorporating the blending operator to formulate an inversion problem. We compare both methods using a robust inversion algorithm with sparse regularization. Results of synthetic and field data examples show that the inversion-based deblending can produce more accurate signal separation for highly blended data.  相似文献   

基于贝叶斯方法的深层充电效应风险模型评估          下载免费PDF全文

李程  陈东 《地球物理学报》2019,62(6):1991-2000

高能电子穿透航天器并在其内部沉积电荷从而引发深层充电效应,是导致卫星故障的重要因素之一.为了评估深层充电效应诱发卫星异常的风险,本文基于贝叶斯方法,使用一颗地球同步轨道卫星的异常数据和GOES-8卫星的电子通量探测数据,计算了不同能量阈值及累积时间的电子注量、不同卫星配置下模拟仿真的沉积电荷,并分别与卫星异常建立一系列概率风险模型.本文从模型中随机抽样得到模拟异常,并与实测异常构造混淆矩阵以评估模型拟合优度,结果表明1.0MeV电子3日累积注量-卫星异常概率风险模型为该卫星最优模型.本文利用最优模型对该卫星深层充电效应风险进行了计算,在1.0MeV电子3日累积注量达到2.0×10~(10)cm~(-2)·sr~(-1)时,该卫星发生深层充电异常的平均后验概率为27%,且95%最小可信值为22%.根据最优模型,我们对该卫星最可能导致异常的部件的材料和结构等特征做出了推断.  相似文献   

中国西部地区利用烈度数据估计地震动参数的方法   总被引：1，自引：0，他引：1       下载免费PDF全文

陈鲲  俞言祥  高孟潭  亢川川 《地震地质》2014,36(4)

通过最近收集到的中国西部地区的6次地震的地震动参数和烈度数据,建立了该地区适用于贝叶斯方法的各烈度档P(I|GM)的经验分布。利用2011年发生的四川炉霍地震和新疆伽师地震资料验证了依据烈度数据用贝叶斯方法估算峰值加速度的可行性。研究结果还表明,用贝叶斯方法估计的参数精度与选择的先验概率(衰减关系)显著相关。针对这2次地震,用各烈度档地震动参数的均值法、地震动衰减关系法及贝叶斯法3种方法估算的峰值加速度值与峰值加速度观测值之间的均方根比较表明,用贝叶斯方法估算的参数精度优于用另2种方法所获结果。  相似文献   

Anthropogenic land-use signals propagate through stream food webs in a California,USA, watershed     

《Limnologica》2014

Human development of watersheds can change aquatic ecosystems via multiple pathways. For instance, human rural development may add nutrients to ecosystems. We used naturally occurring stable isotopes in stream food webs to investigate how land use affects stream ecosystems across a gradient of land development in the San Lorenzo watershed, California. Road density was used as a proxy for land development. We found that streams in watersheds with higher road densities had elevated concentrations of phosphate and nitrate. Furthermore, algal δ¹⁵N values increased as a function of nitrate concentration, but saturated at approximately 6‰. This saturating pattern was consistent with a two-source mixing model with anthropogenic and watershed sources, fit using Bayesian model fitting. In sites that had >2.6 km roads km⁻², anthropogenic sources of N were estimated to represent >90% of the N pool. This anthropogenic N signal was propagated to stream consumers: rainbow trout (Oncorhynchus mykiss), signal crayfish (Pacifasticus leniusculus), and benthic invertebrate δ¹⁵N were positively correlated with algal δ¹⁵N. Even relatively low density rural human land use may have substantial impacts on nutrient cycling of stream ecosystems.  相似文献   

IntCal13 calibrated ages of the Vedde and Saksunarvatn ashes and the Younger Dryas boundaries from Kråkenes,western Norway          下载免费PDF全文

ØYSTEIN S. LOHNE  JAN MANGERUD  HILARY H. BIRKS 《第四纪科学杂志》2014,29(5):506-507

  相似文献   

Copula models for frequency analysis what can be learned from a Bayesian perspective?     

《Advances in water resources》2014

Large spring floods in the Québec region exhibit correlated peakflow, duration and volume. Consequently, traditional univariate hydrological frequency analyses must be complemented by multivariate probabilistic assessment to provide a meaningful design flood level as requested in hydrological engineering (based on return period evaluation of a single quantity of interest). In this paper we study 47 years of a peak/volume dataset for the Romaine River with a parametric copula model. The margins are modeled with a normal or gamma distribution and the dependence is depicted through a parametric family of copulas (Arch 12 or Arch 14). Parameter joint inference and model selection are performed under the Bayesian paradigm. This approach enlightens specific features of interest for hydrological engineering: (i) cross correlation between margin parameters are stronger than expected , (ii) marginal distributions cannot be forgotten in the model selection process and (iii) special attention must be addressed to model validation as far as extreme values are of concern.  相似文献   

A sequential sparse polynomial chaos expansion using Bayesian regression for geotechnical reliability estimations     

Qiujing Pan  Xingru Qu  Leilei Liu  Daniel Dias 《国际地质力学数值与分析法杂志》2020,44(6):874-889

Polynomial chaos expansions (PCEs) have been widely employed to estimate failure probabilities in geotechnical engineering. However, PCEs suffer from two deficiencies: (a) PCE coefficients are solved by the least-square minimization method which easily causes overfitting issues; (b) building a high order PCE is often computationally expensive. In order to overcome the aforementioned drawbacks, the Bayesian regression technique is employed to evaluate PCE coefficients, which not only provides a sparse solution but also avoids overfitting. With the aid of the predictive means and variances given by Bayesian analysis, a learning function is proposed to sequentially select the most informative samples that are critical to build a PCE. This sequential learning scheme can highly enhance the computational efficiency of PCEs. Besides, importance sampling (IS) is incorporated into the sequential learning (SL)-PCEs to deal with geotechnical problems with small failure probabilities. The proposed method of SL-PCE-IS is applied to three illustrative examples, which shows that the improved PCE method is more effective and efficient than the common PCEs method, leading to accurate estimations of small failure probabilities using fewer training samples.  相似文献   

长白山景区旅游安全风险动态评价研究   总被引：1，自引：0，他引：1  

孙滢悦  杨青山  陈鹏 《地理科学》2019,39(5):770-778

以长白山景区旅游安全为研究对象,以鱼骨图、动态贝叶斯、GIS技术等为基本研究方法,从研究区自然环境、社会环境及责任人为3个方面出发,筛选景区致险因子,构建景区旅游安全风险危险性评价指标体系,利用动态贝叶斯方法综合构建景区旅游安全风险动态评价模型;并以实测数据及景区统计数据为依据,划分景区旅游安全风险评价的4个动态时段,综合实现景区旅游安全风险动态风险评价。研究结果表明：① 中等以上风险区域呈条带状分布;② 高风险区域与主要景点重合;③ 长白山景区安全风险发生高概率的时段发生在第三个时段（12:00~14:00）;较高概率发生分别在第二个时段（10:00~12:00）与第四个时段（14:00~16:00）;中等概率发生较高的时段在第四个时段（14:00~16:00）;较低概率发生在第一个时段（8:00~10:00）。  相似文献   

基于加权贝叶斯网络的海洋灾害风险评估     

李明  张韧  洪梅 《海洋通报》2018,(2):121-128

全球气候变化背景下,海洋灾害的群发性、难以预见性和灾害链效应日显突出,造成的损失逐年上升,开展海洋灾害的风险评估工作至关重要。针对海洋灾害评估中的不确定问题,本文首先基于风险理论剖析了海洋灾害风险的不确定性特征,构建了灾害评估指标体系;然后基于贝叶斯网络模型,提出针对不确定性灾害评估的风险贝叶斯网络,进而基于主客观定权,构建了加权贝叶斯网络评估模型;最后对我国沿海地区海洋灾害开展评估研究。实验表明,该评估模型有效实现海洋灾害的风险评估,具有实际可操作性。  相似文献