Pre-stack Depth Migration Imaging Research Articles

From FWI to ultra-high-resolution imaging

In recent years, the development of time-lag full-waveform inversion (FWI) has enabled the use of the full wavefield (primary reflections, diving waves, and their multiples and ghosts) in the inversion process. With this advancement, it is possible to obtain a very detailed velocity model, ultimately reaching the point of deriving from the velocity a migration-like reflectivity image called the FWI image. When the FWI maximum frequency is increased, high-resolution velocity models are obtained, revealing superior reservoir information compared to conventional imaging results. Two case studies will be discussed in this paper. The first is in the Greater Castberg area where the 150 Hz FWI image greatly surpassed the Q Kirchhoff prestack depth migration image from the water-bottom level down to the reservoir (located at a depth of about 1.5 km). The second case study is over the Nordkapp Basin. The use of the full wavefield for the shallow ultra-high-resolution (UHR) FWI image (run at 200 Hz) revealed reverse faulting and pockmark details that were invisible with Kirchhoff prestack depth migration and reverse time migration. By using additional information present in multiples, ghosts, and diving waves, a spatial resolution down to 3 m was achieved. This made it possible to image very thin features without the need for a dedicated high-resolution acquisition design. The current UHR FWI image workflow provides velocity and reflectivity information in the near surface that is important in identifying optimal locations for various purposes such as well placement and wind-turbine installation.

Geophysical evidence for a serpentine mud volcano in the relict slow-spreading center of the South China Sea

The seamounts formed in spreading centers are of basaltic with prominent gravity peaks and high density. But Longmen Seamount, 1.5 km high and ∼ 20 km wide, located in the relict slow-spreading center of the South China Sea, shows a weak gravity peak with a low-amplitude gravity variation (∼19 mGal). To reveal the geological origin of the unusual seamount, we conducted pre-stack depth-migrated image and 3D forward gravity modeling for Longmen Seamount based on the constraint from reflection and refraction seismic data. Our results show that bulk density and seismic velocity of Longmen Seamount is as low as ∼2.1 g/cm3, and 1.6–3.2 km/s, respectively, equivalent to that of poorly consolidated sediments. Furthermore, typical structural elements of the mud volcano system, such as a reflection blank zone, up-bent and downward bending reflections, are observed beneath LM Seamount. For the crust beneath LM Seamount is thin (1.6 km) and highly faulted revealed by numerous clear discontinuous reflections, it is easy for seawater to infiltrate through the thin crust and leads to a high degree of serpentinization of the upper mantle rocks. Combined with serpentine mud volcanoes featuring low bulk density and low seismic velocity, Longmen Seamount is most probably a serpentine mud volcano in a slow-spreading mid-ocean center. For the discovered serpentine mud volcanoes are all restricted to the convergent plate margins, this study will lead to a new understanding of the mechanism of serpentine mud volcanism.

Structures and fluid flows inferred from the microseismic events around a low-resistivity anomaly in the Kakkonda geothermal field, Northeast Japan

A low-resistivity anomaly around granitic intrusive rocks, which could represent a possible supercritical geothermal system, was identified in the Kakkonda geothermal field in Northeast (NE) Japan. We aimed to understand the nature of the possible supercritical geothermal system from the perspective of seismicity in this study. We employed a microseismic monitoring network in this field in 2017 to reveal seismic structures in and around the low-resistivity anomaly. More than 10,000 seismic events were detected until May 2019; we determined precise hypocenter locations using a double-difference algorithm. The main feature that we revealed for the low-resistivity anomaly was the lack of seismic events within its core, which is indicative of ductile conditions. The corresponding temperature is > 370 °C for quartz-dominant rocks in the tectonically active region. Conversely, seismic events were located intensely along the periphery of the low-resistivity anomaly, suggesting the possibility of lateral fluid flow along margins of the low-resistive anomaly. Some of those seismic events had the non-double couple components often associated with a compressive fracturing, which were likely derived from fracture compaction. The lateral fluid flow could be partially derived from the water squeezed by compaction. Additionally, the seismic wave tomography method indicated a region featuring a low ratio of P- and S-wave velocities (VP/VS) in the upper part of the low-resistivity anomaly, indicating a conventional hot water dominant reservoir, although the core of the anomaly was not imaged by the tomography owing to a lack of ray-paths. Pre-stack depth migration imaging using P-coda waves suggested the presence of some reflectors between the upper part and the core of the low-resistivity anomaly, which is possibly a boundary between them. As a result, we concluded that the core of the low resistivity anomaly satisfied the supercritical condition for water, and the structure of the core differed from the structures of the top and periphery of the low-resistivity anomaly. A supercritical geothermal system could be formed in the core of the low-resistivity anomaly.

Paleokarst reservoirs: Efficient and flexible characterization using point-spread-function-based convolution modeling

Paleokarst originates from the collapse, degradation, and infill of karstified rock, and it typically features spatially heterogeneous elements such as breakdown products, sediment infills, and preserved open cavities on all scales. Paleokarst may further contain aquifer or hydrocarbon reservoirs and may pose a drilling hazard during exploration. Seismic characterization of paleokarst reservoirs therefore remains a challenging and important task. We have determined how the application of 2D (3D) spatial convolution operators, referred to as point-spread functions (PSFs), allows for seismic modeling of complex and heterogeneous paleokarst geology at a cost equivalent to conventional repeated 1D convolution. Unlike the latter, which only considers vertical resolution effects, PSF-based convolution modeling yields simulated prestack depth-migrated images accounting for 3D resolution effects vertically and laterally caused by acquisition geometries, frequency-band limitations, and propagation effects in the overburden. We confirm the validity of the approach by a comparison of modeled results to results obtained from a published physical modeling experiment. Finally, we present four additional separate case studies to highlight the usability and flexibility of the approach by assessing different issues and challenges pertaining to characterizing and interpreting seismic features of paleokarst. Through PSF-based convolution modeling, geoscientists working with paleokarst seismic data may be better able to understand how various acquisition and modeling parameters affect seismic images of paleokarst geology.

A complex gas hydrate system imaged jointly using MCS and OBS data from the northern South China Sea

Much of our knowledge of gas hydrate in the Dongsha area of the South China Sea has been gained through intensive geophysical surveys and drilling. However, many factors remain unclear, such as the co-existence of shallow gas hydrate and deeper gas hydrate. This lack of clarity is partially due to the lack of a depth-domain velocity model and accurate imagery of the gas hydrate-bearing sediments. In this study, pre-stack depth migration (PSDM) is used to produce subsurface images using a depth-domain velocity model from OBS tomography as the migration velocity. A simple initial velocity model was built using the seafloor depth information derived from multi-channel seismic (MCS) data. This simple model was used to build a more accurate velocity model by using first arrival time tomography from OBS data. Three different approaches were used to assess the reliability of the velocity model. Resolution tests and uncertainties analysis were used to further evaluate the model. The final PSDM image revealed some features more clearly than they were seen on the time migration image. These features help better explain the complex gas hydrate system. The depth of a bottom-simulating reflector (BSR) was picked directly from the PSDM image. This BSR is very consistent with the drilling result. Our work highlights the importance of PSDM in the study of gas hydrate in complex settings and highlights the potential and value of using OBS refraction in building shallow velocity models.

Seismic velocity structure of seaward-dipping reflectors on the South American continental margin

Seaward dipping reflectors (SDRs) are a key feature within the continent to ocean transition zone of volcanic passive margins. Here we conduct an automated pre-stack depth-migration imaging analysis of commercial seismic data from the volcanic margins of South America. The method used an isotropic, ray-based approach of iterative velocity model building based on the travel time inversion of residual pre-stack depth migration move-out. We find two distinct seismic velocity patterns within the SDRs. While both types show a general increase in velocity with depth consistent with expected compaction and alteration/metamorphic trends, those SDRs that lie within faulted half grabens also have high velocity zones at their down-dip ends. The velocity anomalies are generally concordant with the reflectivity and so we attribute them to the presence of dolerite sills that were injected into the lava pile. The sills therefore result from late-stage melt delivery along the large landward-dipping faults that bound them. In contrast the more outboard SDRs show no velocity anomalies, are more uniform spatially and have unfaulted basal contacts. Our observations imply that the SDRs document a major change in rift architecture, with magmatism linked with early extension and faulting of the upper brittle crust transitioning into more organised, dike-fed eruptions similar to seafloor spreading.

The Imaging Method and Verification Experiment of Chang’E-5 Lunar Regolith Penetrating Array Radar

As a main payload of the lander of Chang’E-5, lunar regolith penetrating array radar (LRPR) is a multi-input and multioutput ground penetrating array radar system. The following are the main characteristics of LRPR: it works at a fixed location, the layout of the antennas is irregular, and the amount of data for imaging is very rare. A prestack depth migration imaging method based on Kirchhoff integral is presented, which overcomes the problems of irregular antenna layout and the layered medium. The issue of sparse matrix imaging for LRPR is solved successfully by the method. The laboratory experiment demonstrates that the proposed method is very suitable for LRPR imaging, in which the targets are clearly distinguished and their depths are also consistent with the reality.

Seismic imaging of magma sills beneath an ultramafic-hosted hydrothermal system

Hydrothermal circulation at mid-ocean ridge volcanic segments extracts heat from crustal magma bodies. However, the heat source driving hydrothermal circulation in ultramafic outcrops, where mantle rocks are exhumed in low-magma-supply environments, has remained enigmatic. Here we use a three-dimensional P -wave velocity model derived from active-source wide-angle refraction-reflection ocean bottom seismometer data and pre-stack depth-migrated images derived from multichannel seismic reflection data to investigate the internal structure of the Rainbow ultramafic massif, which is located in a non-transform discontinuity of the Mid-Atlantic Ridge. Seismic imaging reveals that the ultramafic rocks composing the Rainbow massif have been intruded by a large number of magmatic sills, distributed throughout the massif at depths of ∼2–10 km. These sills, which appear to be at varying stages of crystallization, can supply the heat needed to drive high-temperature hydrothermal circulation, and thus provide an explanation for the hydrothermal discharge observed in this ultramafic setting. Our results demonstrate that high-temperature hydrothermal systems can be driven by heat from deep-sourced magma even in exhumed ultramafic lithosphere with very low magma supply.

Estimation of uncertainties in fault lateral positioning on 3D PSDM seismic image: an example from the North West Shelf

This extended abstract presents a two-step sequence to estimate uncertainties in lateral positioning of fault planes on 3D PSDM (pre-stack depth migration) seismic images. This analysis can be applied to any localised detail on a seismic image but, in the majority of geological settings, it is most important for the faults. The first step provides an approximate evaluation of what causes the uncertainties, how the uncertainties are distributed in a 3D space, and what to expect within target zones. The authors assume that every complex detail within a 3D PSDM velocity model causes some uncertainties to the seismic image below. Thus, the uncertainties at a target level depend on the complexity of the overburden and the seismic acquisition parameters. At this step a qualitative 3D volume of lateral fault position uncertainties is created. In the second step the authors focus on a single fault of practical interest. Based on the results of the first step, the authors modify the existing 3D PSDM anisotropic velocity model by introducing additional anomalies that cause maximal changes to the lateral position of the fault on seismic image. Then the authors iteratively re-migrate a small sub-volume around the fault and check the PSDM images and residual moveout. The objective is to find out how far the velocity variations can move the image of the fault and still satisfy available seismic data. The second step gives more reliable quantitative estimations of the impact of velocity on fault positioning. A real multi-azimuth 3D seismic dataset from the North West Shelf is used to illustrate this sequence.

Ray-based seismic modeling of geologic models: Understanding and analyzing seismic images efficiently

Often, interpreters only have access to seismic sections and, at times, well data, when making an interpretation of structures and depositional features in the subsurface. The validity of the final interpretation is based on how well the seismic data are able to reproduce the actual geology, and seismic modeling can help constrain that. Ideally, modeling should create complete seismograms, which is often best achieved by finite-difference modeling with postprocessing to produce synthetic seismic sections for comparison purposes. Such extensive modeling is, however, not routinely affordable. A far more efficient option, using the simpler 1D convolution model with reflectivity logs extracted along verticals in velocity models, generates poor modeling results when lateral velocity variations are expected. A third and intermediate option is to use the various ray-based approaches available, which are efficient and flexible. However, standard ray methods, such as the normal-incidence point for unmigrated poststack sections or image rays for simulating time-migrated poststack results, cannot deal with complex and detailed targets, and will not reproduce the realistic (3D) resolution effects of seismic imaging. Nevertheless, ray methods can also be used to estimate 3D spatial prestack convolution operators, so-called point-spread functions. These are functions of the survey, velocity model, and wavelet, among others, and therefore they include 3D angle-dependent illumination and resolution effects. Prestack depth migration images are thus rapidly simulated by spatial convolution with detailed 3D reflectivity models, which goes far beyond the limits of 1D convolution modeling. This 3D convolution modeling should allow geologists to better assess their interpretations and draw more definitive conclusions.

Fluid accumulation along the Costa Rica subduction thrust and development of the seismogenic zone

AbstractIn 2011 we acquired an 11 × 55 km, 3‐D seismic reflection volume across the Costa Rica margin, NW of the Osa Peninsula, to accurately image the subduction thrust in 3‐D, to examine fault zone properties, and to infer the hydrogeology that controls fluid accumulation along the thrust. Following processing to remove water column multiples, noise, and acquisition artifacts, we constructed a 3‐D seismic velocity model for Kirchhoff prestack depth migration imaging. Images of the plate boundary thrust show high‐reflection amplitudes underneath the middle to lower slope that we attribute to fluid‐rich, poorly drained portions of the subduction thrust. At ~ 5 km subseafloor, beneath the upper slope, the plate interface abruptly becomes weakly reflective, which we interpret as a transition to a well‐drained subduction thrust. Mineral dehydration during diagenesis may also diminish at 5 km subseafloor to reduce fluid production and contribute to the downdip change from high to low amplitude. There is also a layered fabric and systems of both thrust and normal faults within the overriding plate that form a “plumbing system.” Faults commonly have fault plane reflections and are presumably fluid charged. The faults and layered fabric form three compartmentalized hydrogeologic zones: (1) a shallow NE dipping zone beneath the slope, (2) a steeply SW dipping zone beneath the shelf slope break, and (3) a NE dipping zone beneath the shelf. The more direct pathway in the middle zone drains the subduction thrust more efficiently and contributes to reduced fluid pressure, elevates effective stress, and creates greater potential for unstable coseismic slip.

Seismic imaging of subsurface structure using tomographic migration velocity analysis: a case study of South China Sea data

Seismic migration algorithms are often used to reveal subsurface structures and can be roughly divided into stacking methods, pre-stack time migration methods and pre-stack depth migration methods. In theory, pre-stack depth migration methods can provide the most accurate subsurface images. However, these methods are sensitive to the accuracy of the migration velocity model, and thus the creation of a precise migration velocity model is one of the most important steps in seismic processing. We applied the tomographic travel-time inversion and the method of pre-stack Kirchhoff depth migration-based migration velocity analysis (MVA) to certain South China Sea data, in which we iteratively inverted the precise, high-resolution migration velocity model. We selected travel-time differences from the offset-domain common image gathers of the pre-stack Kirchhoff depth migration and inverted the travel-time differences for velocity updating. After several iterations, both the precise migration velocity model and the corresponding pre-stack depth migration image were obtained. The comparisons with Normal Move-out stacking and pre-stack Kirchhoff time migration confirm that the tomographic MVA and corresponding pre-stack Kirchhoff depth migration provide the best image quality and highest-resolution subsurface images from the South China Sea data.

Investigation of deep-water Gulf of Mexico subsalt imaging using anisotropic model, data set and RTM — Tempest

Performing accurate depth-imaging is an essential part of deep-water Gulf of Mexico exploration and development. Over the years, depth-imaging technology has provided reliable seismic images below complicated salt bodies, and has been implemented in workflows for both prospect generation as well as reservoir development. These workflows include time domain preprocessing using various multiple elimination techniques, anisotropic model building, and depth-imaging using anisotropic reverse time migration (RTM). However, the accuracy of the depth-migrated volumes is basically unknown because they are tested only in the locations where a well is drilled. In order to learn about the accuracy of anisotropic deep water Gulf of Mexico model building, and depth-imaging tools which are used for processing and imaging of field acquired data, we created a 3D vertical transverse isotropic (VTI) anisotropic earth model and a 3D seismic data set representing subsalt Gulf of Mexico geology. The model and data set are referred to as the Tempest data set, the original being created several years ago. The recent model and data set were created incorporating upgraded technology to reflect recent developments in data acquisition, model building and depth-imaging. Our paper presents the new Tempest anisotropic model, data set, and RTM prestack depth-migration (PSDM) results. The Tempest RTM PSDM is being used to learn about the differences between the exact geological model and the RTM PSDM image, helping in the interpretation of real RTM prestack depth-migrated data.

From time to depth with CRS attributes

A processing workflow was introduced for reflection seismic data that is based entirely on common-reflection-surface (CRS) stacking attributes. This workflow comprises the CRS stack, multiple attenuation, velocity model building, prestack data enhancement, trace interpolation, and data regularization. Like other methods, its limitation is the underlying hyperbolic assumption. The CRS workflow provides an alternative processing path in case conventional common midpoint (CMP) processing is unsatisfactory. Particularly for data with poor signal-to-noise ratio and low-fold acquisition, the CRS workflow is advantageous. The data regularization feature and the ability of prestack data enhancement provide quality control in velocity model building and improve prestack depth-migrated images.

Impact of modelling shallow channels on 3D prestack depth migration, Elgin-Franklin fields, UKCS

A case study from the Elgin-Franklin fields is used to illustrate the influence of seabed low velocity anomalies on imaging deep targets below 5000 m depth. In the uppermost 400 m of sediment, glacial channels are present and generate locally low velocity anomalies resulting in pull-up/push-down effects observed on seismic reflection data. The original 3D prestack depth migration did not take into account these shallow velocity anomalies, but used a smooth interval velocity in the uppermost layer of sediment. In the southern area, the results were locally better than with prestack time migration, but in the northern area they were worse, or at best comparable. A new velocity model with a channel layer was built using well velocities for the initial model. Tomographic and migration iterations were used to refine all velocities, including the channel velocities, and for introducing anisotropy. Imaging of the pre-Cretaceous section was clearly an improvement over prestack time migration. For correct imaging, local shallow velocity variations must be identified and taken into account in the velocity model. Even though the overburden appears to be rather flat, prestack depth migration imaging performs better than prestack time migration overall.

Seismic imaging of the Naga thrust using multiscale waveform inversion

This case study images the structural features related to the Naga thrust fault (northeast India) using a combination of multiscale waveform inversion and prestack depth migration (PSDM). The waveform model and the PSDM image complement each other: the former provides a physical-property map (P-wave velocity model) and the latter provides a structural image. The velocity model encompassing the starting model for waveform inversion is constructed using joint inversion of first and reflected traveltimes. The [Formula: see text] data are inverted consecutively in [Formula: see text] bandwidths to yield the final waveform model, which in turn is used for PSDM. The PSDM image and the waveform model are consistent with the lithological interpretation of an inline exploratory well. When interpreted jointly, the PSDM image and the waveform model reveal the presence of a conjugate fault system in the Naga thrust and fold belt.

The Tempest Project—Addressing challenges in deepwater Gulf of Mexico depth imaging through geologic models and numerical simulation

Performing depth imaging is an essential part of deepwater Gulf of Mexico (GOM) exploration. Over the years, depth-imaging technology has provided the most reliable seismic images below salt and has been implemented in the workflows of the prospect generation process. But how accurate are these images? Since model building for depth imaging is partially an interpretative process, and depth imaging involves resolving seismic propagation through complicated geologic features, it is easy for the resulting prestack depth-migrated images to include imaging and positioning errors.

Unified imaging of multichannel seismic data: Combining traveltime inversion and prestack depth migration

Imaging 2D multichannel land seismic data can be accomplished effectively by a combination of traveltime inversion and prestack depth migration (PSDM), referred to as unified imaging. Unified imaging begins by inverting the direct-arrival times to estimate a velocity model that is used in static corrections and stacking velocity analysis. The interval velocity model (from stacking velocities) is used for PSDM. The stacked data and the PSDM image are interpreted for common horizons, and the corresponding wide-aperture reflections are identified in the shot gathers. Using the interval velocity model, the stack interpretations are inverted as zero-offset reflections to constrain the corresponding interfaces in depth; the interval velocity model remains stationary. We define a coefficient of congruence [Formula: see text] that measures the discrepancy between horizons from the PSDM image andtheir counterparts from the zero-offset inversion. A value of unity for [Formula: see text] implies that the interpreted and inverted horizons are consistent to within the interpretational uncertainties, and the unified imaging is said to have converged. For [Formula: see text] greater than unity, the interval velocity model and the horizon depths are updated by jointly inverting the direct arrivals with the zero-offset and wide-aperture reflections. The updated interval velocity model is used again for both PSDM and a zero-offset inversion. Interpretations of the new PSDM image are the updated horizon depths. The unified imaging is applied to seismic data from the Naga Thrust and Fold Belt in India. Wide-aperture and zero-offset data from three geologically significant horizons are used. Three runs of joint inversion and PSDM are required in a cyclic manner for [Formula: see text] to converge to unity. A joint interpretation of the final velocity model and depth image reveals the presence of a triangle zone that could be promising for exploration.

Resolution and illumination analyses in PSDM: A ray-based approach

Prestack depth migration (PSDM) should be the ultimate goal of seismic processing, producing angle-dependant depth images of the subsurface reflectivity. But the expected quality of PSDM images is constrained by many factors. Understanding all of these factors is necessary to improve depth imaging of geologic structures. In all PSDM approaches, e.g., Kirchhoff or wave-equation, migration always includes compensating for wave propagation in the overburden (back propagation, downward continuation, etc.), before focusing back the reflected/diffracted energy at each considered location in depth (imaging). Ideally, we would like to retrieve the reflectivity of the ground as detailed as possible to invert for the elastic parameters. But the waves perceive the reflectivity through “thick glasses,” seeing blurred structures, and not necessarily all of them, depending on the illumination. Only a filtered version of the true reflectivity is therefore retrieved. Being able to estimate these filters, the so-called re...

2-D traveltime and waveform inversion for improved seismic imaging: Naga Thrust and Fold Belt, India

SUMMARY Exploration along the Naga Thrust and Fold Belt in the Assam province of Northeast India encounters geological as well as logistic challenges. Drilling for hydrocarbons, traditionally guided by surface manifestations of the Naga thrust fault, faces additional challenges in the northeast where the thrust fault gradually deepens leaving subtle surface expressions. In such an area, multichannel 2-D seismic data were collected along a line perpendicular to the trend of the thrust belt. The data have a moderate signal-to-noise ratio and suffer from ground roll and other acquisition-related noise. In addition to data quality, the complex geology of the thrust belt limits the ability of conventional seismic processing to yield a reliable velocity model which in turn leads to poor subsurface image. In this paper, we demonstrate the application of traveltime and waveform inversion as supplements to conventional seismic imaging and interpretation processes. Both traveltime and waveform inversion utilize the first arrivals that are typically discarded during conventional seismic processing. As a first step, a smooth velocity model with long wavelength characteristics of the subsurface is estimated through inversion of the first-arrival traveltimes. This velocity model is then used to obtain a Kirchhoff pre-stack depthmigrated image which in turn is used for the interpretation of the fault. Waveform inversion is applied to the central part of the seismic line to a depth of ∼1 km where the quality of the migrated image is poor. Waveform inversion is performed in the frequency domain over a series of iterations, proceeding from low to high frequency (11‐19 Hz) using the velocity model from traveltime inversion as the starting model. In the end, the pre-stack depth-migrated image and the waveform inversion model are jointly interpreted. This study demonstrates that a combination of traveltime and waveform inversion with Kirchhoff pre-stack depth migration is a promising approach for the interpretation of geological structures in a thrust belt.