Seismic Prestack Depth Migration GRT

SF-GRT stands for »Statoil Fraunhofer Generalized Radon Transform« and refers to a method for seismic prestack depth migration. The measured seismic data are interpreted as scattered responses from the subsurface, summed over areas of equal travel time (isochrones). Based on this, the inverse process – the reconstruction of subsurface scattering points or reflectors – is formulated using the linear Radon transform.

The result is reflection coefficients of the subsurface as a function of ray incidence or reflection angles. These are determined by weighted summation of the contributions from all illumination directions.

Advantages of Migration in the Angular Domain:

The input data set can be used in any order; binning is not required.
The true-amplitude weighting factor is derived directly from the results of dynamic ray tracing – no additional approximations are required. At the same time, strict illumination compensation is applied.
Multi-valent folding wavefronts are taken into account in a physically accurate manner – this improves image quality in complex subsurface structures.
Diffraction imaging is produced as an additional product without any significant extra effort.

»SF-GRT« – Joint Development With Industry

The »SF-GRT« software tool is a joint development by Fraunhofer ITWM and the Norwegian energy company Equinor (formerly Statoil). It was developed in response to a specific industrial need. The application combines geophysical expertise with the solutions from our »High Performance Computing« department and the industry partner’s deep understanding of requirements. For »High Performance Computing«, »SF-GRT« is thus an outstanding example of successful, practical contract research.

© Fraunhofer ITWM
Migration stack overlaid with a representation of how the subsurface structure varies with angle of incidence.

Commercial Use Of »SF-GRT«

We license »SF-GRT« for use on parallel computers with Infiniband interconnects (high-speed networking between computing nodes). In addition, we have many years of experience processing seismic customer data on our own high-performance computers. Typical projects arise from specialized research questions, where true-amplitude migration results and diffraction images provide valuable insights.

We Offer Prestack Depth Migration Using »SF-GRT« For:

Marine surface data
Marine OBN data
Onshore seismic data

The Results Are:

Reflection Angle Gather
Partial angle stacks
Full stacks with custom processing stages
Raw data migration
Gather processing (demultiplexing, flattening)
Angle stacking, including laterally and vertically optimized muting
Bandwidth expansion, Q-compensation, spectral matching

We use PreStack Pro (SR) software from Sharp Reflections for data processing.

GRT Seismic Prestack Depth Migration — © Fraunhofer ITWM
Layering of the subsurface as a result of seismic migration.

In Addition, We Supply:

Diffraction images (stacked results presented as a sequence with progressively increasing suppression of the reflection components)
Special Features
- Q-phase and Q-amplitude correction along physical beam paths
- Signal-to-noise improvement by downweighting contributions from directions far from the stationary directions
- Support for complex velocity models, including orthorhombic anisotropy with tilted axes of symmetry

Locations of refractors in the subsurface as a result of GRT migration — © Fraunhofer ITWM
Migrated stacking result and – when expanded – how this result depends on the direction of illumination. Diffraction patterns appear here as extended patterns. Reflections exhibit curvature.

Project Experience and Areas of Application

Our portfolio of completed projects includes:

Depth imaging down to approximately 9,000 meters
Coverage areas of up to approximately 2,000 km²
High resolutions for specialized applications and high-resolution fault imaging
HR and UHR data with a vertical resolution of up to approximately 1 m and a horizontal resolution of 3.125 m

Since the physical principles underlying the seismic method remain stable across a wide range of scales, we also use »SF-GRT« for HR and UHR data (high- and ultra-high-resolution), for example in the planning of offshore wind farms.

Tabbed contents

Expand all Close all

Applications of »Simulated Migration«

GRT technology is well-suited for simulated migration. Illumination compensation factors enable efficient calculation of point spread functions (PSFs). These can be calculated on a dense subsurface grid for the acquisition geometry of a given or planned dataset, as well as for a migration velocity model that varies significantly both vertically and laterally. When convolved with the subsurface reflectivity, they yield the expected migration response.

In a current research project, we are developing a methodological framework in collaboration with an industry partner. Its applications range from time-lapse seismic interpretation and seismic history matching to the development of reservoir models, the planning of logging campaigns, diffraction imaging interpretation, and modeling for AI applications.

Please feel free to contact us if you would like to learn more about this research or discuss potential applications of the technology with us.

Geothermal Applications

For geothermal applications, it is recommended to use GRT migration in azimuth sectors. In this method, the migration results depend not only on the angle of reflection but also on the horizontal direction of incidence (azimuth). When data quality is good, this allows conclusions to be drawn about fracture systems in the subsurface rock.

When combined with the diffraction imaging mode, which enhances disturbance patterns, GRT provides valuable information for a comprehensive assessment of a subsurface’s geothermal potential. We are pleased to collaborate with local governments and their service providers in the field of geothermal seismic exploration.

Specialized Technology Components

Scalable Processing on High-Performance Computers (HPC)

»SF-GRT« can be described as a data mapping problem: A 5D input dataset comprising many terabytes (x- and y-coordinates of sources and receivers, as well as recording time) is transformed into a 5D model of the subsurface (x, y, z as a function of two reflection angle components).

With the GPI parallelization library we developed, the input data is read once and stored in the working memory of the compute nodes. During migration, the nodes exchange data at very high network throughput – without any lossy reduction in the volume of input data. Double-buffering overlays data transfer and computation, enabling high scalability even in large computer clusters. Parallelization across nodes, within nodes, and the use of CPU vector units are standard practices in our HPC department.

The Core Geophysical Components of »SF-GRT« Include:

Access and interpolation of phase-rotated, anti-aliased, and Q-compensated data using specially developed sinc operators
Use of tapers to process algorithmically detected inner and outer edges when mapping rays to track coordinates—to minimize migration artifacts

All Contributing Amplitudes Are Available for Each Sub-Point. This Makes It Possible To:

Identification of stationary directions
Targeted attenuation of amplitudes outside the Fresnel zone for noise reduction (The Fresnel zone refers to the area around a reflection point from which seismic signals still contribute significantly to the measured reflection.)
selective suppression of events along stationary directions, enabling high-quality diffraction imaging results

For diffraction imaging, we have developed a stepwise attenuation strategy that allows users to interactively track the transition from a reflection image to a diffraction image by gradually reducing the stationary energy components.

Seismic Prestack Depth Migration Using »SF-GRT« (Generalized Radon Transform)

More Information