Seismic Attributes For Enhancing Structural And Stratigraphic Features: Application To N-Field, Malay Basin, Malaysia

Over two decades, analysis of seismic attributes had been an integral part of seismic reflection interpretation. Seismic attributes are an influential assistance to seismic interpretation, delivering geoscientists with alternative images of structural (faults) and stratigraphic features (channels), which can be utilised as mechanisms to identify prospects, ascertain depositional environment and structural deformation history more rapidly even provide direct hydrocarbon indicators. The additional steps are obligatory to compute and interpret the attributes of faults and channels from seismic images, which are often sensitive to noise due to the characteristically computed as discontinuities of seismic reflections. Furthermore, on a conventional seismic profile or poor quality data, faults and channels are hard to visible. The current research review these geological structures through a case study of 3D seismic data from N-field in the viewpoint of Malay Basin. This study aimed to characterise the structure and stratigraphic features by using seismic attributes on the N-field below seismic resolution. Also, two different methods are proposed to improve seismic reflections, i.e., faults and channels that are hard to see on the conventional 3D data set. The first method, to detect faults in seismic data, which this paper employs the ant tracking attribute as a unique algorithm to be an advanced forwarding that introduces a new tool in the interpretation of fault. The effective implementation of ant tracking can be achieved when the output of other faults sensitive attributes are used as input data. In this work, the seismic data used are carefully conditioned using a signal. Chaos and variance that are sensitive to faults are applied to the seismic data set, and the output from these processes are used as input data that run the ant tracking attribute, which the faults were seen difficult to display on the raw seismic data. Meanwhile, for the second method, spectral decomposition was adopted to deliberate the way its method could be utilised to augment stratigraphic features (channels) of the N-field, where the channel is ultimately considered being one of the largest formations of the petroleum entrapment. The spectral decomposition analysis method is an alternative practice concentrated on processing S-transform that can offer better results. Spectral decomposition has been completed over the Pleistocene channels, and results propose that application of its methods directs to dependable implications. Respective channel in this area stands out more obviously within the specific frequency range. The thinner layer demonstrates higher amplitude reading at a higher frequency, and the thicker channel displays higher amplitude reading at a lower frequency. Implementation of spectral decomposition assists in deciding the channels that were placed within incised valleys and helps in recognising the orientation and the relative thickness of each channel. By doing this, the ant tracking attribute and spectral decomposition approach have generated the details of subsurface geologic features through attributes by obtaining enhanced reflections and channels and sharpened faults, respectively.

Download Full-text

Zircon U-Pb ages of the Benom Complex: Implications for a Late Triassic uplift event in Central Peninsular Malaysia

Bulletin of the Geological Society of Malaysia ◽

10.7186/bgsm72202101 ◽

2021 ◽

Vol 72 ◽

pp. 1-13

Author(s):

Chee Meng Choong ◽

◽

Manuel Pubellier ◽

Benjamin Sautter ◽

Mirza Arshad Beg ◽

...

Keyword(s):

Seismic Data ◽

Spectral Decomposition ◽

Input Data ◽

Seismic Attributes ◽

Structural Deformation ◽

Quality Data ◽

Late Triassic ◽

Data Set ◽

Decomposition Approach ◽

Seismic Reflections

Download Full-text

Seismic illumination of small-offset seismogenic faults, Anadarko Basin, Oklahoma

Interpretation ◽

10.1190/int-2020-0135.1 ◽

2021 ◽

pp. 1-50

Author(s):

Swetal Patel ◽

Folarin Kolawole ◽

Jacob I. Walter ◽

Xiaowei Chen ◽

Kurt J. Marfurt

Keyword(s):

Focal Mechanism ◽

Seismic Attributes ◽

Seismic Monitoring ◽

Set Covering ◽

Structural Deformation ◽

Data Set ◽

Anadarko Basin ◽

Seismic Reflection Data ◽

The Past ◽

Basement Faults

In the past decade across the Central and Eastern U.S., there has been a substantial increase in the seismicity rate, which scientists broadly attribute to wastewater disposal and, to a lesser extent, hydraulic fracturing. Active clusters of seismicity illuminate linear fault segments within the crystalline basement that were unknown until seismicity began. Such surprises are due to the limited availability of 3D surface seismic surveys and the difficulty of imaging relatively shallow earthquake events from sparse seismic monitoring arrays. The Sooner Trend Anadarko Basin Canadian Kingfisher Counties (STACK) play of Central Oklahoma provides an opportunity to map such basement faults. Modern high-quality surface seismic data acquired to map the Meramec and Woodford unconventional resource plays enable us to image the basement faults and intrusions. Furthermore, because of increased earthquake risk from anthropogenic activities in the past decade, state regulatory agencies have deployed a relatively dense array of seismic monitoring stations, which allows us to integrate earthquake data into subsurface fault analysis. We have mapped structural deformation using a suite of seismic attributes, including multispectral coherence, volumetric curvature, and aberrancy, in a 3D seismic reflection data set covering 625 sq mi of the STACK area of the Anadarko Basin, Oklahoma. To unravel the relationship between the structures and seismicity, we use relocated locally recorded earthquakes and compute the focal mechanism solution for the relocated events. Our results reveal previously unmapped fault segments with dominant north–south, northwest, and northeast trends, most of which extend into the shallower sedimentary Hunton and Woodford Formations. Because of the small offset, we find that aberrancy and the curvature attribute best illuminate the basement-rooted faults in the study area. Fault segments with significant offset are better illuminated by band-limited multispectral coherence. We argue that the inherited structure of these faults makes them easily illuminable by flexure-related seismic attributes, especially within the sedimentary cover. The integration of the illuminated faults with relocated earthquakes and focal mechanism solutions shows that some of the illuminated faults that have hosted intrasedimentary and/or basement seismicity are reactivated strike-slip faults. We hypothesize that careful attribute mapping of faults and flexures, coupled with an understanding of the local stress and geomechanical properties, calibrated with recent seismic activity in the area, can help mitigate seismic hazards in tectonic settings where small-offset faults predominate.

Download Full-text

Deep learning-driven velocity model building workflow

The Leading Edge ◽

10.1190/tle38110872a1.1 ◽

2019 ◽

Vol 38 (11) ◽

pp. 872a1-872a9 ◽

Cited By ~ 4

Author(s):

Mauricio Araya-Polo ◽

Stuart Farris ◽

Manuel Florez

Keyword(s):

Deep Learning ◽

Seismic Data ◽

Model Building ◽

Ground Truth ◽

Velocity Model ◽

Training Data ◽

Quality Data ◽

Generative Adversarial Network ◽

Data Set ◽

Velocity Models

Exploration seismic data are heavily manipulated before human interpreters are able to extract meaningful information regarding subsurface structures. This manipulation adds modeling and human biases and is limited by methodological shortcomings. Alternatively, using seismic data directly is becoming possible thanks to deep learning (DL) techniques. A DL-based workflow is introduced that uses analog velocity models and realistic raw seismic waveforms as input and produces subsurface velocity models as output. When insufficient data are used for training, DL algorithms tend to overfit or fail. Gathering large amounts of labeled and standardized seismic data sets is not straightforward. This shortage of quality data is addressed by building a generative adversarial network (GAN) to augment the original training data set, which is then used by DL-driven seismic tomography as input. The DL tomographic operator predicts velocity models with high statistical and structural accuracy after being trained with GAN-generated velocity models. Beyond the field of exploration geophysics, the use of machine learning in earth science is challenged by the lack of labeled data or properly interpreted ground truth, since we seldom know what truly exists beneath the earth's surface. The unsupervised approach (using GANs to generate labeled data)illustrates a way to mitigate this problem and opens geology, geophysics, and planetary sciences to more DL applications.

Download Full-text

A statistical methodology for deriving reservoir properties from seismic data

Geophysics ◽

10.1190/1.1443878 ◽

1995 ◽

Vol 60 (5) ◽

pp. 1437-1450 ◽

Cited By ~ 30

Author(s):

Frédérique Fournier ◽

Jean‐François Derain

Keyword(s):

Seismic Data ◽

Large Scale ◽

Seismic Attributes ◽

Seismic Facies ◽

Reservoir Properties ◽

Data Set ◽

Analysis Technique ◽

Seismic Facies Analysis ◽

Statistical Relationships

The use of seismic data to better constrain the reservoir model between wells has become an important goal for seismic interpretation. We propose a methodology for deriving soft geologic information from seismic data and discuss its application through a case study in offshore Congo. The methodology combines seismic facies analysis and statistical calibration techniques applied to seismic attributes characterizing the traces at the reservoir level. We built statistical relationships between seismic attributes and reservoir properties from a calibration population consisting of wells and their adjacent traces. The correlation studies are based on the canonical correlation analysis technique, while the statistical model comes from a multivariate regression between the canonical seismic variables and the reservoir properties, whenever they are predictable. In the case study, we predicted estimates and associated uncertainties on the lithofacies thicknesses cumulated over the reservoir interval from the seismic information. We carried out a seismic facies identification and compared the geological prediction results in the cases of a calibration on the whole data set and a calibration done independently on the traces (and wells) related to each seismic facies. The later approach produces a significant improvement in the geological estimation from the seismic information, mainly because the large scale geological variations (and associated seismic ones) over the field can be accounted for.

Download Full-text

Phase and the Hilbert transform

The Leading Edge ◽

10.1190/tle33101164.1 ◽

2014 ◽

Vol 33 (10) ◽

pp. 1164-1166 ◽

Cited By ~ 13

Author(s):

Steve Purves

Keyword(s):

Signal Processing ◽

Data Processing ◽

Hilbert Transform ◽

Seismic Data ◽

Seismic Attributes ◽

Seismic Data Processing ◽

Data Set ◽

The Hilbert Transform

The concept of phase permeates seismic data processing and signal processing in general, but it can be awkward to understand, and manipulating it directly can lead to surprising results. It doesn't help that the word phase is used to mean a variety of things, depending on whether we refer to the propagating wavelet, the observed wavelet, poststack seismic attributes, or an entire seismic data set. Several publications have discussed the concepts and ambiguities (e.g., Roden and Sepúlveda, 1999 ; Liner, 2002 ; Simm and White, 2002 ).

Download Full-text

Analysis of Source Parameters relationships for clusters of similar events recorded in Central Apennines

10.5194/egusphere-egu2020-17019 ◽

2020 ◽

Author(s):

Daniele Spallarossa ◽

Paola Morasca ◽

Dino Bindi ◽

Matteo Picozzi ◽

Kevin Mayeda

Keyword(s):

Pulse Duration ◽

Cross Correlation ◽

Source Parameters ◽

Quality Data ◽

Corner Frequency ◽

Central Apennines ◽

Data Set ◽

Small Magnitude ◽

Decomposition Approach ◽

Magnitude Range

Aim of this study is to investigate the relationship between moment magnitude (Mw) and source duration (i.e. corner frequency) for moderate to small magnitude earthquakes recorded in Central Apennines, Italy, including the 2016-2017 Amatrice-Norcia-Visso&#160;sequence. A data-set of ~ 6000&#160;events in the magnitude range ~1 and&#160; 6.5 was used to retrieve a reference data set of source parameters by applying spectral decomposition approach (Generalized Inversion Techniques). The large population of analyzed earthquakes allowed us to investigate the scaling of the source parameters with the earthquake size, their variability with hypocentral depth and to characterize the scaling between local and moment magnitudes in the magnitude range from 1 to 6.5 (Deichmann 2017). Analyzing the same data-set and taking advantage of the available high quality data for small events recorded in the area, we focus on the scaling properties of clustered events in the&#160;magnitude range between ~1 and&#160; 3.5. By applying different methodologies, relying on cross-correlation analysis, we detect a preliminary set of clusters. Then, events within 2 km from the geographic location of each cluster were extracted from a very large (more than 500000 events) high-resolution earthquake parametric catalog. New cross-correlation analyses were carried out on stations within 50 km from the centroid of each previously identified clusters to pad each ones with low magnitude events (below 2). This multi-steps procedure allowed us to identified 2933 events belonging to 45 clusters. For an in-deep analysis of source properties, we focus on three clusters selected on the basis of the number of events and different hypocentral depth distributions. For each cluster, the P-waves pulse duration (equivalent to corner frequency) of the events were compared each other on different stations. Results clearly show that below Ml ~ 2 the pulses duration remains nearly constant also for stations with low kappa values, showing a saturation effects. For a comparison with the GIT and cross-correlation results we also evaluate source parameters using a method based on coda-envelope amplitude measurements (Mayeda et al. 2003) applying site and path parameters previously calibrated for Central Apennines by Morasca et al. 2019. This comparison from independent and completely different methodologies applied on the same clusters well agrees with the saturation observed in pulse duration, strengthen the results and allowed us to define, for the given network geometry and earthquake distribution, the magnitude threshold below which we believe it is not possible to estimate source parameters. Moreover, our analysis of two clusters co-located on map but with different depth highlights a variation in stress drop with depth;

Download Full-text

Seismic attributes and analogs to characterize a large fold in the Taranaki Basin

Interpretation ◽

10.1190/int-2020-0018.1 ◽

2020 ◽

Vol 8 (4) ◽

pp. SR27-SR31

Author(s):

Karelia La Marca Molina ◽

Heather Bedle ◽

Jerson Tellez

Keyword(s):

New Zealand ◽

Seismic Data ◽

Large Scale ◽

Seismic Attributes ◽

Depositional Environments ◽

Analysis Tool ◽

Data Set ◽

Reflection Seismic ◽

Channel Belt ◽

Large Channel

The Taranaki Basin lies in the western portion of New Zealand, onshore and offshore. It is a Cretaceous rift basin that is filled with up to approximately 10 km thick deposits from marine to deepwater depositional environments from the Cretaceous (approximately 93 ma) to the Neogene (approximately 15 ma). This basin underwent important tectonic events that resulted in large-scale features such as faults and folds and the deposition of turbidites such as channels and channel belts. These features easily are recognizable in seismic data. When analyzing the offshore 3D Pipeline data set, we recognized a peculiar fault-like feature with large-scale dimensions (approximately 15 km long and approximately 1 km wide) within the sequence. The alignment was perpendicular to the direction of deposition in the basin (southeast–northwest) as identified by previous studies and subparallel to the main structures in the area (southwest–northeast). We interpreted the seismic character of the funny-looking thing (FLT) likely as (1) a fault, (2) a fold, or (3) a large channel belt within the basin. We use seismic attributes such as coherence (Sobel filter), dip, cosine of phase, and curvature to characterize this feature geomorphologically. The geologic background of the area and analog settings aided in understanding and distinguishing the nature of this large structure. Monocline examples in seismic data are rare to find, and we want to show how to avoid misinterpretations. Geological feature: Fault-bend fold or large-amplitude fold (possibly monocline) Seismic appearance: Large, discontinuous, high-variance feature Alternative interpretations: Fault, fold Features with a similar appearance: Fault, fold, wide straight channel belt (time or horizon slice) Formation: Rift sequence of the Taranaki Basin Age: Eocene Location: Taranaki Basin, Western offshore New Zealand Seismic data: Provided by New Zealand Petroleum and Minerals Contributors: Karelia La Marca, Heather Bedle, Jerson Tellez; School of Geosciences; University of Oklahoma, Norman, OK, USA Analysis tool: 3D reflection seismic, geometric seismic attributes

Download Full-text

Ordovician glacial paleogeography: Integration of seismic spectral decomposition, well sedimentological data, and glacial modern analogs in the Murzuq Basin, Libya

Interpretation ◽

10.1190/int-2018-0069.1 ◽

2019 ◽

Vol 7 (2) ◽

pp. T383-T408 ◽

Cited By ~ 2

Author(s):

Francisco J. Bataller ◽

Neil McDougall ◽

Andrea Moscariello

Keyword(s):

Seismic Data ◽

Spectral Decomposition ◽

Depositional Environments ◽

Optimum Frequency ◽

Data Set ◽

Glacial Sediments ◽

Depositional System ◽

Significant Difference ◽

Different Depths ◽

Well Data

Ancient glacial sediments form major hydrocarbon plays in several parts of the world; most notably, North Africa, Latin America, and the Middle East. We have described a methodology for reconstructing broad-scale paleogeographies in just such a depositional system, using an extensive subsurface data set from the uppermost Ordovician glacial sediments of the Murzuq Basin of southwest Libya. Our workflow begins with the analysis of a large, high-quality 3D seismic data set, to understand the frequency content. Subsequently, optimum frequency bands are extracted, after applying spectral decomposition, and then recombined into an R (red) G (green) B (blue) blended cube. This volume is then treated as an image within which paleomorphological features can be distinguished and compared with modern glacial analogs. Mapping at different depths (time slices) of these features is then tied, by integration with core and image-log sedimentology, to specific depositional environments defined within the framework of a facies scheme developed using the well data and published outcrop studies. These depositional environments are extrapolated into areas with little or no well data using the spectral decomposition as a framework, always taking into account the significant difference in vertical resolution between the seismic data set and core-scale descriptions. The result of this methodology is a set of calibrated maps, at three different time depths (two-way time travel), indicating paleogeographic reconstructions of the glacial depositional environments in the study area and the evolution through time (at different depths/time slices 2D + 1) of these glacial settings.

Download Full-text

Footprint removal from seismic data with residual dictionary learning

Geophysics ◽

10.1190/geo2019-0482.1 ◽

2020 ◽

Vol 85 (4) ◽

pp. V355-V365

Author(s):

Julián L. Gómez ◽

Danilo R. Velis

Keyword(s):

Dictionary Learning ◽

Seismic Data ◽

Input Data ◽

Random Noise ◽

Coherent Noise ◽

Edge Preservation ◽

Data Set ◽

Edge Preserving ◽

Learned Dictionary ◽

Data Volume

Dictionary learning (DL) is a machine learning technique that can be used to find a sparse representation of a given data set by means of a relatively small set of atoms, which are learned from the input data. DL allows for the removal of random noise from seismic data very effectively. However, when seismic data are contaminated with footprint noise, the atoms of the learned dictionary are often a mixture of data and coherent noise patterns. In this scenario, DL requires carrying out a morphological attribute classification of the atoms to separate the noisy atoms from the dictionary. Instead, we have developed a novel DL strategy for the removal of footprint patterns in 3D seismic data that is based on an augmented dictionary built upon appropriately filtering the learned atoms. The resulting augmented dictionary, which contains the filtered atoms and their residuals, has a high discriminative power in separating signal and footprint atoms, thus precluding the use of any statistical classification strategy to segregate the atoms of the learned dictionary. We filter the atoms using a domain transform filtering approach, a very efficient edge-preserving smoothing algorithm. As in the so-called coherence-constrained DL method, the proposed DL strategy does not require the user to know or adjust the noise level or the sparsity of the solution for each data set. Furthermore, it only requires one pass of DL and is shown to produce successful transfer learning. This increases the speed of the denoising processing because the augmented dictionary does not need to be calculated for each time slice of the input data volume. Results on synthetic and 3D public-domain poststack field data demonstrate effective footprint removal with accurate edge preservation.

Download Full-text

Multiscale phase inversion of seismic data

Geophysics ◽

10.1190/geo2017-0353.1 ◽

2018 ◽

Vol 83 (2) ◽

pp. R159-R171 ◽

Cited By ~ 14

Author(s):

Lei Fu ◽

Bowen Guo ◽

Gerard T. Schuster

Keyword(s):

Seismic Data ◽

Phase Inversion ◽

Input Data ◽

Waveform Inversion ◽

Data Set ◽

Acoustic Data ◽

Full Waveform ◽

Elastic Data ◽

Mpi Implementation ◽

Starting Model

We present a scheme for multiscale phase inversion (MPI) of seismic data that is less sensitive than full-waveform inversion (FWI) to the unmodeled physics of wave propagation and to a poor starting model. To avoid cycle skipping, the multiscale strategy temporally integrates the traces several times, i.e., high-order integration, to produce low-boost seismograms that are used as input data for the initial iterations of MPI. As the iterations proceed, lower frequencies in the data are boosted by using integrated traces of lower order as the input data. The input data are also filtered into different narrow frequency bands for the MPI implementation. Numerical results with synthetic acoustic data indicate that, for the Marmousi model, MPI is more robust than conventional multiscale FWI when the initial model is moderately far from the true model. Results from synthetic viscoacoustic and elastic data indicate that MPI is less sensitive than FWI to some of the unmodeled physics. Inversion of marine data indicates that MPI is more robust and produces modestly more accurate results than FWI for this data set.

Download Full-text