SUMMARY Wouldn't it be beneficial if we could predict the time-series at a seismic station even no longer exists? In geophysical data analysis, this capability would enhance our ability to study and monitor events noise, particularly in regions with incomplete coverage or where stations are temporarily offline. This introduces novel adaption of encoder–decoder networks from subfield deep learning, modified development wave fields between two stations. Using 1-D measurements, algorithm aims...

The separation of the diffracted wavefield is a notorious challenge in both seismic and ground-penetrating-radar (GPR) subsurface imaging. Over last decade, numerous studies have attempted to address this with various deterministic schemes. While each these schemes has specific advantages disadvantages, they all require an adaptation corresponding processing parameters for application, especially when crossing scales between electromagnetic measurements. In recent years, convolutional neural...

Single-station waveforms of teleseismic earthquakes are highly complex, because they a superposition numerous phases corresponding to different wave types and propagation paths. Moreover, data recorded at single stations is contaminated by noise, which often has similar or larger amplitudes than the arrivals earthquakes, especially in densely-populated areas. For high precision research facilities, for example field particle physics gravity detection, precise knowledge seismic wavefield...

The common-reflection surface (CRS) method represents a multidimensional stacking approach; i.e., the is determined in midpoint and offset directions. In 2D case, three attributes span surface, thus requiring three-parameter search contrary to one-parameter classic common-midpoint stack. However, CRS wavefront use data redundancy direction as well, which makes them very useful several seismic applications, e.g., preconditioning, velocity model building, migration. Contrary previous works, we...

The classic common-midpoint (CMP) stack, which sums along offsets, suffers in challenging environments the acquisition is sparse. In past, several multiparameter stacking techniques were introduced that incorporate many neighboring CMPs during summation. This increases data redundancy and reduces noise. Multiparameter methods can be parameterized by same wavefront attributes are multifocusing (MF), common-reflection-surface (CRS), implicit CRS, nonhyperbolic CRS (nCRS). CRS-type operators...

Seismic attributes play a crucial role in fault interpretation and mapping fracture density. Conventionally, seismic derived from migrated reflections are used for this purpose. The the other counterparts of recorded wavefield often ignored excluded categorization. We have performed categorization diffracted part combine them into new attribute class, which we call diffractivity attributes. extraction is based on 3D Kirchhoff time migration operator that includes dynamic muting. distinguish...

Summary The Common Reflection Surface (CRS) stack method improves the signal to noise ratio significantly. Furthermore its attributes allow for additional applications like prestack data enhancement. However, CRS suffers from conflicting dip situations. Current approaches determine an initial set of parameters multiple operators per sample. This can lead errors especially in complex settings. We propose use a global optimization namely differential evolution where search space is split into...

The process of wavefield decomposition is a notorious challenge in seismic and electromagnetic imaging, because often only specific components the full measured are targeted during processing, while other undesired. This specifically applies to separation reflected diffracted wavefields. While has been treated as noise past, recent advances have demonstrated its importance for both high-resolution imaging subsurface heterogeneities such faults depthvelocity model building multi-channel...

Summary The CRS operator improves the signal to noise ratio significantly due consideration of neighboring midpoints as well offset. determination required attributes for is often done by pragmatic approach get initial values that are refined a local optimization. This works reasonable most parts, however in more complex structures like salt bodies result not reliable anymore. Additionally does perform particularly presence conflicting dips. Therefore we propose use genetic algorithm based...

Summary In literature four multi-parameter stacking approaches can be found that stack along midpoint and offset direction: multifocusing, common-reflection-surface (CRS), implicit CRS non-hyperbolic CRS. recent years some comparisons were published. However, in these works methods missing conflicting dips, mostly caused by diffractions, not considered. Diffractions are of higher-order therefore benefit the most from methods. Schwarz et al. introduced a new parametrization for CRS-type...

SUMMARY Standard seismic acquisition and processing require appropriate source–receiver offsets. P-cable technology represents the opposite, namely, very short offsets at price of increased spatial lateral resolution with a high-frequency source. To use this advantage, flow excluding offset information is required. This aim can be achieved tuned to diffractions because point scatter same in midpoint direction. Usually, are small amplitude events careful diffraction separation required as...

Summary Seismic data contains valuable information about the subsurface, but also noise that can be distracting. Denoising is an important step in seismic processing. We used a U-Net-shaped encoder-decoder network with ResNeXt blocks and added self-attention to deeper layers of network. Additionally, we implemented attention gates as second mechanism order attend global features next local ones. evaluated impact on machine-learning-based processing by comparing results obtained without...

Summary Diffraction separation remains a challenge in both seismic and ground-penetrating radar (GPR) imaging. Different deterministic methods have been proposed for this purpose. Although these are of different nature, they common that processing parameters to be adapted each application and, particular, when changing scales between seismics GPR. In the recent years, convolutional neural networks proven powerful tool data analysis. However, their performance strongly depends on training...

Seismic waveforms of teleseismic earthquakes are highly complex since they a superposition numerous phases that correspond to different wave types and propagation paths. In addition, measured contain noise contributions from the surroundings measuring station. The regional distribution seismological stations is often relatively sparse, in particular regions with low seismic hazard such as Northern Germany. However, detailed knowledge wavefield generated by large can be crucial for precise...

Summary The diffracted wavefield plays an important role in the processing and interpretation of both seismic ground-penetrating radar (GPR) data, as it encodes information about small-scale subsurface heterogeneities such faults, small objects or water intrusions glaciers. Separating faint data that is often dominated by higher amplitude reflected arrivals therefore a key challenge GPR data. When using deterministic methods for diffraction separation, parameters have to be adapted every...

Wouldn't it be beneficial if we could predict the time series at a seismic station even no longer exists? In geophysical data analysis, this capability would enhance our ability to study and monitor events noise, particularly in regions with incomplete coverage or where stations are temporarily offline. This introduces novel adaption of encoder-decoder networks from subfield Deep Learning, modified development wave fields between two stations. Using one-dimensional measurements, algorithm...

Summary P-cable data have the special characteristic of short source-receiver offsets and a high frequency source. These characteristics lead to resolution with issues in velocity-model building. Conventional methods can not be applied due an insufficient offset coverage. Point diffractions potential allow building without appropriate information available. This is case their scattering nature, moveout midpoint direction equivalent. Therefore, velocity point scatterer extracted direction. We...

While seismic data contains an abundance of useful information about the subsurface, it is also contaminated with disturbing recorded energy, coherent and random noise. A crucial step in processing, therefore, denoising. In this work, we use a U-Net-based encoder-decoder network that uses ResNeXt blocks rather than traditional convolutions as denoising tool. Furthermore, add self-attention to ResNext deeper part neural architecture. We compare results those obtained same network, but without...

Summary In the recent years, machine learning and artificial intelligence have gained increasing importance across all fields of Earth sciences. particular convolutional neural networks (CNNs) proven to be a powerful tool for different types data analysis, such as pattern recognition, image segmentation, denoising reconstruction. context seismic electromagnetic imaging, decomposition measured wavefield into its reflective diffractive components is notorious challenge that has been approached...

The separation of the reflected and diffracted wavefields has been a crucial challenge in both seismic ground- penetrating radar (GPR) data processing for many years. Different deterministic methods based on wavefront attributes, adaptive subtraction reflections or plane-wave destruction have proposed applied this purpose. While all these are characterized by different advantages drawbacks they generally common that choice parameters to be adapted each application – particularly scales GPR...

Summary Noise is a major concern in seismic data and influences the processing interpretability of at various steps. However, noise has certain pattern, which can be exploited by machine learning algorithms, that rose drastically popularity within last decade. We aim to remove random an early stage workflow shot-gather domain. use unsupervised approach without time consuming necessity generating labels. In our work, we autoencoder, resembles U-Net structure but uses ResNext50 encoder...

Summary Most current implementations of the CRS operator suffer from occurrence conflicting dip situations in acquired data. To address this properly we apply idea CDS. We use iCRS that can be related to operator, and show, dips resolved well multi-parameter processing. The results are promising reveal a lot potential for further applications. This is shown by diffraction separation technique applied field data obtained Levantine Basin.

Summary Most current implementations of the CRS operator suffer from occurrence conflicting dip situations in acquired data. To address this properly we apply idea CDS. We use i-CRS that can be related to operator, and show, dips resolved well multiparameter processing. The results are promising reveal a lot potential for further applications. This is shown by diffraction separation technique applied field data obtained Levantine Basin.