A random three-dimensional (3D) porous medium can be reconstructed from a two-dimensional (2D) image by reconstructing an image from the original 2D image, and then repeatedly using the result to reconstruct the next 2D image. The reconstructed images are then stacked together to generate the entire reconstructed 3D porous medium. To perform this successfully, a very important issue must be addressed, i.e., controlling the continuity and variability among adjacent layers. Continuity and variability, which are consistent with the statistics characteristic of the training image (TI), ensure that the reconstructed result matches the TI. By selecting the number and location of the sampling points in the sampling process, the continuity and variability can be controlled directly, and thus the characteristics of the reconstructed image can be controlled indirectly. In this paper, we propose and develop an original sampling method called three-step sampling. In our sampling method, sampling points are extracted successively from the center of $5\times 5$ and $3\times 3$ sampling templates and the edge area based on a two-point correlation function. The continuity and variability of adjacent layers were considered during the three steps of the sampling process. Our method was tested on a Berea sandstone sample, and the reconstructed result was compared with the original sample, using tests involving porosity distribution, the lineal path function, the autocorrelation function, the pore and throat size distributions, and two-phase flow relative permeabilities. The comparison indicates that many statistical characteristics of the reconstructed result match with the TI and the reference 3D medium perfectly.

Direct sampling (DS) is a versatile multiple‐point statistics method for generating spatial‐temporal geostatistical models. DS is known for being able to address a variety of training images and hence spatiotemporal stochastic modeling problems. One limitation of DS is the central processing unit (CPU) time, mostly attributed to the use of a random search for patterns in the training image. To improve CPU performance, we propose a tree‐based direct sampling (TDS) method. In our method, training patterns are grouped according to their similarities combined with a clustering tree for fast lookup. Rather than patterns, we store locations in our database. During the simulation, TDS applies a tree‐driven search approach. Two objectives, similarity and diversity, are used to rapidly retrieve patterns and prevent trapping into local optima. We also introduce a way to speed up simulation by means of pasting patterns with adaptive size. The performance of our TDS is investigated using a 2‐D benchmark training image. Moreover, we apply the proposed method to two real cases including gap filling the bedrock topography in Antarctica from radar to better understand subglacial hydrology and creating 3‐D groundwater models in the Danish aquifer system. Based on several quantitative evaluations, we find the proposed TDS is comparable to DS in terms of simulation quality, while significantly saves CPU time.

Seismic inversion problems are well-known to be nonlinear and their misfit functions often involve many local minima. Global optimization methods are capable of converging to the global minimum of a misfit function, thus, they are promising in seismic inversion. As a global optimization method, multimutation differential evolution (MMDE) has been proven to be effective in solving high-dimensional seismic inversion problems. However, it is challenging to choose the optimal parameters for MMDE to achieve the best performance in seismic inversion. In this paper, we propose a new deep network based on MMDE and name it as MMDE-Net, which enables us to learn the optimal parameters by using a network training procedure rather than empirically choosing them. Benefiting from the learned parameters, MMDE-Net has advantages over MMDE in applications. Numerical examples based on synthetic and field data set clearly indicate that MMDE-Net can provide faster convergence speed and better inversion result than conventional methods in seismic inversion.

Three-dimensional (3D) structures are useful for studying the spatial structures and physical properties of porous media. A 3D structure can be reconstructed from a single two-dimensional (2D) training image (TI) by using mathematical modeling methods. Among many reconstruction algorithms, an optimal-based algorithm was developed and has strong stability. However, this type of algorithm generally uses an autocorrelation function (which is unable to accurately describe the morphological features of porous media) as its objective function. This has negatively affected further research on porous media. To accurately reconstruct 3D porous media, a pattern density function is proposed in this paper, which is based on a random variable employed to characterize image patterns. In addition, the paper proposes an original optimal-based algorithm called the pattern density function simulation; this algorithm uses a pattern density function as its objective function, and adopts a multiple-grid system. Meanwhile, to address the key point of algorithm reconstruction speed, we propose the use of neighborhood statistics, the adjacent grid and reversed phase method, and a simplified temperature-controlled mechanism. The pattern density function is a high-order statistical function; thus, when all grids in the reconstruction results converge in the objective functions, the morphological features and statistical properties of the reconstruction results will be consistent with those of the TI. The experiments include 2D reconstruction using one artificial structure, and 3D reconstruction using battery materials and cores. Hierarchical simulated annealing and single normal equation simulation are employed as the comparison algorithms. The autocorrelation function, linear path function, and pore network model are used as the quantitative measures. Comprehensive tests show that 3D porous media can be reconstructed accurately from a single 2D training image by using the method proposed in this paper.

The subglacial bed topography is critical for modelling the evolution of Thwaites Glacier in the Amundsen Sea Embayment (ASE), where rapid ice loss threatens the stability of the West Antarctic Ice Sheet. However, mapping of subglacial topography is subject to uncertainties of up to hundreds of metres, primarily due to large gaps of up to tens of kilometres in airborne ice-penetrating radar flight lines. Deterministic interpolation approaches do not reflect such spatial uncertainty. While traditional geostatistical simulations can model such uncertainty, they become difficult to apply because of the significant non-stationary spatial variation of topography over such large surface area. In this study, we develop a non-stationary multiple-point geostatistical (MPS) approach to interpolate large areas with irregular geophysical data and apply it to model the spatial uncertainty of entire ASE basal topography. We collect 166 high-quality topographic training images (TIs) of resolution 500 m to train the gap-filling of radar data gaps, thereby simulating realistic topography maps. The TIs are extensively sampled from deglaciated regions in the Arctic as well as Antarctica. To address the non-stationarity in topographic modelling, we introduce a Bayesian framework that models the posterior distribution of non-stationary TIs assigned to the local line data. Sampling from this distribution then provides candidate training images for local topographic modelling with uncertainty, constrained to radar flight line data. Compared to traditional MPS approaches that do not consider uncertain TI sampling, our approach results in a significant improvement in the topographic modelling quality and efficiency of the simulation algorithm. Finally, we simulate multiple realizations of high-resolution ASE topographic maps. We use the multiple realizations to investigate the impact of basal topography uncertainty on subglacial hydrological flow patterns.