CWP Seminar - Fall, 2006

Date

Speaker

Subject

Notes

08/30

Carlos

PP vs. PS Time Migration Algorithm: Prestack time migration of converted wave data.

In my previous seminars I talked about registration of converted waves using only stacked data. Now I will discuss registration in the context of using prestack time migration of converted wave data. More specifically I will touch in the practical difficulties and limitations of doing prestack time migration of converted wave data compared to the prestack time migration of pure modes. I will show that prestack time migration of converted wave data has more limitations than for pure modes.

09/04

Steve

Imagemodel Software and SUEA2DF_SSN

Imagemodel is a front end modeling code for a modified and debugged SUEA2DF elastic 4^th order finite difference code. Users begin with an image of the geology they wish to shoot (BMP, TIF, PNG). Imagemodel walks the user through the model construction (aliasing, materials, source, BCs, etc.), and outputs all files necessary to automatically shoot and post-process complete 2-D P-SV seismic lines. Talk includes a brief example.

Ivan

Interferometry with SAFOD drill-bit seismic data

After identifying which portion of the SAFOD drill-bit seismic data was reliable for processing, we were able to reconstruct the wavefield from drill-bit noise as if excited at receiver locations in the SAFOD pilot-hole (PH). Because drill-stem accelerometer recordings are not available, deconvolution interferometry is much more successful in synthesizing a wide-band response from the drill-bit records than cross-correlations. By playing with which receiver components are used for interferometry, we can control the orientations of excitation/recording in the output interferometric data. From the results of deconvolution interferometry, our preliminary interpretation identify direct arrivals, free-surface multiples, and possibly P- and PS-waves reflected from the San Andreas fault zone. By combining the PH drill-bit data with the main-hole active shot data, we will verify whether we can use the drill-bit excitation to image the fault zone.

Laxmidhar

Migration velocity analysis for factorized tilted TI media

The tilt of the symmetry axis plays an important role in imaging complex overthrust areas, where TI shale layers are often bent due to tectonic processes. Here, we present some synthetic results for a typical anisotropic thrust sheet (layer) that has a variable symmetry-axis orientation. The sheet is embedded in an isotropic medium and imaged with and without taking the tilt into account in order to analyze the character of improvements of the image.

09/06

Rodrigo

Time-lapse traveltime differences around a compacting reservoir

I derive a 3-D solution for time-lapse traveltime differences occuring around a compacting reservoir. Using Fermat's Principle of least time --- and its equivalent formulation under Hamilton's formalism of the Principle of Least Action --- it is possible to derive very general equations for timeshifts for non-zero offset rays in a inhomogeneous and generally anisotropic, layered media. The result is an important generalization of previous works which tackled only very specific models: isotropic and homogenous medium, focusing mostly on zero-offset rays.

Xiaoxia

Analysis of Azimuthal AVO at Rulison Field, CO (Part I)

I will present seismic attributes of azimuth AVO and moveouts extracted from a few horizons at Rulison Field. Among others, the main issues we attempt to address in this project are as follows. 1. To evaluate the impact of anisotropic spreading correction on azimuthal AVO analysis in a field data; 2. To demonstrate how azimuthal AVO and NMO complement each other in a real case; 3. To study the influence of preprocessing, (e.g. statics correction) on the reliability of azimuthal seismic attributes.

Masatoshi

Relocation of induced micro-seismicity in an oil reservoir during steam injections

At Cold Lake, Albata, Canada, Imperial Oil Ltd. produces heavy oil by injecting high-temperature and high-pressure steam into the oil reservoir at depth of ~450 m. During the productive operations, the stress/strain changes in the structure due to the steam injections, induce micro-seismic events. We detected the seismic event signals from down-hole data recorded by geophones, and relocated the hypocenters by using the double-difference method, which can provide much precise distribution of locations than conventional methods.

09/11

Yuanzhong

The acurate measurement of phase velocity and attenuation of ultrasound in a complex viscoelastic material, dough.

The acurate measurement of phase velocity and attenuation of ultrasound in a complex viscoelastic material, dough. The difficult of the measurement is a) The unknown complex acoustic impedance of dough. b) The time evolution of the material. c) The frequency dependce of transducer impedence was not clear. I will show you how I designed experiments and analyzed the data to get final results.

Greg

Reverse time migration

I will be discussing my upcoming project on reverse time migration (RTM). I will give some background on the development of RTM as well as an "industry perspective" on why RTM is of interest today. This project is still in its formative stages, so I am interested in obtaining feedback, not just presenting material.

Roel

Extracting the Green's function for the diffusion equation

In seismic interferometry one extracts the Green's function by cross-correlating the response of the system to ambient noise. Contrary to common belief, time-reversal invariance is NOT needed for doing this. As an example I show that the Green's function for the diffusion equation can be extracted from the cross-correlation of the response of a diffusive system to ambient noise.

09/13

Xiaoxiang

Introduction to the research progress on the D'' region

My previous master's study mainly focused on the shear wave splitting in the lowermost mantle. In this presentation, I would like to make some introduction mainly about the background seismological research on the D'' region--the major thermo-chemical boundary layer in the lowermost mantle, including some conceptual models of D'' and the primary seismological constraints on D''.

Jianmin

Seismic signal detection based on coded excitations

We tried to apply Coding Theory to weak seismic signal detection. The method could be loosely described as follows, a large number of discrete excitations are released according to a time coding scheme. Simulation result and analysis show that, from the point of view of signal detection, this method is efficient (good capability of resisting disturbance).

Dave

Ultrasonic imaging in holes

Ultrasonic tools for imaging holes come in various sizes that depend naturally on the size of the hole; and they produce images with corresponding resolution. Examples include images from tools with central frequencies ranging from 250 kHz to 25 MHz. The lower frequency is used to image petroleum boreholes. The higher frequency is used in intravascular ultrasonic (IVUS) imaging of human arteries. Problems in processing and interpreting IVUS images have much in common with seismic imaging of the earth's subsurface.

09/18

Derek

An Intuitive Framework for Real-Time Freeform Modeling

I will go over a 3D modeling technique [Botsch and Kobbelt 2004]. That allows the user to create complex model manipulations using a single 9-dof manipulator. I will discuss details of the implementation of this technique and possible applications.

Gabi

Amplitude preservation in Gazdag migration method

A version of Gazdag migration (Gazdag, 1980) in such a way that it uses the true amplitude one-way wave equations instead of the standard ones, in order to realize a true amplitude wave equation migration for zero-offset data.

Dirk

Using CRS parameters for multiple suppression

Multi parameter processing like the Common Reflection Surface (CRS) processing is a major research topic in the Wave Inversion Technology (WIT) consortium. The parameters are determined in a data driven way by a coherency measure and can be used for various applications like CRS stacking, tomography, or aperture optimized Kirchhof migration. I will briefly introduce the concept of multi parameter processing and sketch a methodology how these parameters may be used to describe the kinematic operator of multiples which allows their supression in post stack data. For the removal in pre stack data the virtual real source concept may provide a promising tool. The major advandage of the described concept is it applicability to sparse 3-D data which poses a major challenge to the existing multiple supression techniques.

09/20

Jia

Anisotropic migration

A first touch on anisotropic migration. I review over several papers on anisotropic imaging, which tells something about importance of anisotropic imaging and the difficulties lies in there. Also will be covered is the modeling for anisotropic media, such as finite element modeling method.

Niran

Application of GBM Using Sigsbee2A Data

If we use linearity of filtering and migration, can we get a good image?

Ilya

Critical angle vs. "critical slowness" in anisotropic parameter estimation

Last spring Martin Landro and I introduced a parameter-estimation method for TI and orthorhombic media that we called "critical-angle reflectometry." The method is based on computing the critical angle at the target horizon, which requires knowledge of the overburden velocity field. For layer-cake models, however, it is more practical to operate with the horizontal slowness (ray parameter) of the critical ray. In particular, the "critical slowness" can be combined with the NMO (stacking) velocity to estimate the anellipticity parameters required for time imaging in VTI and azimuthally anisotropic media.

09/25

John M.

Tau-P and Radial deconvolution demultiple

Shallow water multiples can be hard to kill. Methods such as parabolic radon transform (PRT) demultiple often do not work very well, and we need to take advantage of periodicity. This presentation describes the application of deconvolution in Tau-P and Radial domains.

Jyoti

VRS: 2 applications

The most common question I encounter when talking about Virtual Real Source (VRS) is about the source variability. This afternoon, I will show how the variability effect can be minimized. Apart from that, I will also show how the VRS can be used to find source receiver directivity and compute intrinsic attenuation (God willing).

Paul

Seismic data structure

When we analyze seismic data, we look for coherent structures i.e. reflections off geologic markers changing with space and time. Spatial coherence characterizes reflections and we use this information to infer geologic structure. When we image seismic data, we ignore coherent structures: source and receiver wavefields are imaged irrespective of their time coherence. All we care about are the traveltimes from sources and receivers to image points. Thus, in imaging we disregard the main element that makes seismic data usable. Why do we do that? What are we losing out? Can we change this? In this seminar, we begin to explore these questions.

09/27

NO TALKS due to SEG

SEG

10/02 & 10/04

SEG

NO TALKS SEG

NEW ORLEANS

10/09

Jun

Carlos

PP and PS wavefield separation

I compare three techniques for wavefield separation in the tau-p domain using a simple synthetic dataset. I discuss the advantages and limitations of the methods.

John Stockwell

Asymptotic series

Are convergent series the best approximations to functions? Are more terms better? In general the answer is no to each of these questions. Many of our most useful approximations consist of a few terms of an asymptotic representation of a function of interest. Because asymptotic series do not in general converge, more terms will tend to produce a divergent partial sum. Convergent series representations may not converge quickly enough to be useful as approximations.

10/11

Xiaoxiang

The D'' anisotropy and the differential travel time method

Following my first representation, I introduce the reseach progress on the D'' anisotropy, plus the geometries (VTI and AA), the mechanisms (LPO and SPO) of the D'' anisotropy and the related models. Then I introduce the differential travel time method which I used in the shear wave splitting analysis. This method can easily remove the effect of source-side and receiver-side anisotropy, helping us get the splitting time of ScSV and ScSH through D''. At last, I say something about the dataset.

Rodrigo

P-wave AVO for a compacting reservoir

Pursuing the same line of thought that led me to derive closed-form analytic equations describing timeshifts around a compacting reseroir, I show closed form equations for P-wave AVO attributes. These expressions show how the AVO intercept, gradient and curvature change in terms of applied stress and the stress-sensitivity tensor. The equations confirm our intuition that deviatoric stresses cause angular dependency of the AVO attributes . Moreover, since we used an isotropic stress-sensitivity tensor, the same combinations of elements of this tensor appear in the intercept and gradient terms.

Dirk

Using CRS parameters for multiple suppression

After briefly introducing the concept of multiple parameter processing (MPP) in my last seminar I will describe today, how it can be used to suppress multiples. The first step is the identification of the multiple in the stacked section together with its corresponding wave field attributes as determined by MPP. Leaving these parameters out during the stack results in a primary only post stack section. The MPP parameters determine the kinematic operator of the multiple. To remove the multiple in the pre stack data it has to be folded by the source signature. The “virtual real source” concept may be an interesting approach to achieve this.

10/16

NO TALKS

Fall Break

10/18

Masatoshi

Stochastic approach for obtaining seismic wave paths in random heterogeneous media

For cases where the media include random heterogeneities characterized by statistic functions, we cannot deterministically obtain seismic ray paths. We propose a stochastic approach for obtaining two-point seismic wave paths in a three-dimensional random heterogeneous media, by deriving and solving analytical equations and showing some numerical examples.

Yuanzhong

How do the gas bubbles affect the phase velocity and attenuation in materials.

Last time I presented how the ultrasonic phase velocity and attenuation were measured in a complex viscoelastic material,dough. What infomation could we get from these results? I will show you a model to relate the results to the bubble size and its distribution in the material.

Laxmidhar

Migration velocity analysis for factorized tilted TI media

The effect of tilt in seismic imaging plays a crucial role in complex geological terrains having dipping shale layers of the Canadian Rocky Mountain Foothills and Gulf of Mexico etc. Ignoring the effect of tilt causes serious mispositioning of target events although all other anisotropic parameters are taken into account. Here, we extend the VTI (vertical transverse isotropy) methodology to TTI (tilted transverse isotropy) by taking into consideration symmetry axis orthogonal to the bedding plane of dipping layers. We present some synthetic results for typical subsurface geological models like syncline, salt, and thrust sheets etc. Each model is characterized by at least one dipping layer and the symmetry axis orthogonal to the dipping beds embedded in factorized isotropic media. We have imaged all these models with the help of anisotropic prestack Kirchhoff depth migration by taking into account the effect of tilt and compared that without tilt. The migration velocity analysis is carried out for each model by generating image gathers. Flattening of events in the image gathers show that the reflectors are correctly positioned without misfocusing when tilt is incorporated with parameter estimation of individual layers in a layer-stripping manner.

10/23

Greg

Reverse time migration of simple synthetic data

I show some early results of reverse-time migration of data created using a simple 2-layer v(z) + density model. The implementation of RTM in the Madagascar software package is briefly discussed. Since this model was built using a free surface, we see surface-generated multiples in the data. Imaging artifacts are present in the data whether or not we use the density contrast in the time-reversed modeling of the receiver data. As I continue the project, I expect to investigate whether or not we can actually use the multiples in the imaging process.

Steve

Tunnel Modeling: SUEA2DF instabilities and the Saenger “Rotated Grid Algorithm”

Modified SUEA2DF has proven stable for modeling marine data. Applications to tunnels – regions with internal traction release B.C.s – have shown the algorithm to be unstable. I briefly cover why tunnels are being modeled, and various software “auditioned” for this purpose. I then trace the history of the current algorithm and discuss reasons for said instability. I conclude with a summary of work by Saenger et al., which allow for implementation of high contrasts in the medium without explicitly defined internal B.C.s.

Xiaoxia

Analysis of Azimuthal AVO on a RIGHT Dataset at Rulison Field, CO

Last time, we looked at some "good" AVAZ results on an inappropriate dataset. Today, I will show you "artificially clean" results on a right dataset. The points I would like to emphasize are as follows: 1. MASC is important to reliable estimation of azimuthal AVO when its azimuthal variation is nonegligible compared to that of AVO; 2. azimuthal AVO is a robust attribute; 3. contrary to common wisdom, analysis of azimuthal AVO can be more robust than that of NMO ellipse; 4. joint inversion of NMO and AVO might be difficult because of their different vertical resolution.

10/25

Jianmin

Time Coded Impulsive Seismic Technique

Following my presentation on 'Coded Excitation' on the previous seminar, I'll show you some features of the new method 'Time Coded Impulsive Seismic Technique' which I'm studying this period. It will improve the resolution and Signal to Noise Ratio (SNR) of the seismic record compared with that from traditional method.

Derek

An Intuitive Framework for Real-Time Freeform Modeling – Implemented in 2D

I will quickly review the Freeform modeling technique presented in my previous seminar. Then I will show my 2D implementation of the technique and discuss the algorithm I have used. I will conclude by introducing some ideas for enhancement to this 2D implementation.

Roel

Extracting the Green's function for general linear systems

In the previous semiar I showed that the Green's function fr the diffusion equation can be extracted from ambient fluctuations. The fact that the Green's function can be found for acoustic waves, elastic waves, and the diffusion equation, suggests an underlying principle. I will show this underlying by deriving the extraction of the Green's function for a wide class of linear systems.

10/30

Gabi

Amplitude preservation in Gazdag migration method

Continuing...: A version of Gazdag migration (Gazdag, 1980) in such a way that it uses the true amplitude one-way wave equations instead of the standard ones, in order to realize a true amplitude wave equation migration for zero-offset data.

John M.

Velocity modeling insights

I have accumulated some experience in the "art" of velocity modeling for depth imaging. In this talk I will try to share some of this painfully acquired knowledge.

Dave

Local phase correlations

Local cross correlations of time-lapse seismic images yield estimates of local displacements and strains. Displacements parallel to image features are less well-resolved than those perpendicular to image features. To improve resolution, we may compute local phase correlations. These are simply local cross correlations of images to which we have applied local prediction error (spectral whitening) filters.

11/01

Niran

Application of GBM Using Sigsbee2A Data - 2

If we define the frequency limits in the migration code (sumiggbzo) with respect to the data which was filtered in four different frequency band, can we get a better image?

Yuanzhong

How do the gas bubbles affect the phase velocity and attenuation in materials.

Last time I presented how the ultrasonic phase velocity and attenuation were measured in a complex viscoelastic material,dough. What infomation could we get from these results? I will show you a model to relate the results to the bubble size and its distribution in the material.

Suzie

Review of Recent Research Progress

I will review the recent research progress made by our team on the following topics: 1)Coded seismic signal study; 2)Earthquake ground motion and seismic hazard study; 3)Application of GPS Geodesy in Geophysics.

11/06

Jyoti

PP-wave reflection coefficients in VTI attenuative media

Linearized forms of PP-wave reflection coefficient in media with VTI velocity and attenuation anisotropy can be obtained using perturbation approach. It is clear from these equations that apart from dependence on velocity and density contrasts, teh PP-wave reflection coefficients depend on contrasts in the attenuation coefficients along with symmetry axis direction across the interface.

Jia

12 IWSA

A summary of the 12th IWSA.

Ilya

Cusps on PS wavefronts in TI media

The group velocity of SV-waves in transversely isotropic media may become multivalued, which leads to cusps (triplications) on the wavefront. While these cusps are well understood for "pure" SV reflected and transmitted modes, it is more difficult to describe cuspoidal wavefronts for converted PS-waves. I will show that the existence and size of the PSV-wave cusp is controlled by a simple expression that involves the critical angle and the anisotropy parameter sigma. The shape of the cusp is determined not just by the SV-wave slowness surface, but also by the lateral displacement of the P-to-SV conversion point (i.e., of the "virtual" shear source).

11/08

Rodrigo

Problems in numerical computations of 2d displacements

To model seismic data for my reservoir compaction project I need the computation of the state of stress around the reservoir, i.e., the excess stresses, strains and displacements around a reservoir that underwent pore-pressure drop due to production.In the past week, however, I have been experimenting difficulties in computing the state of stress inside the reservoir itself, leading to wrong displacement computations once the integration of the strains goes through the reservoir. In this seminar I will show you possible reasons for this error and how I intend to go about to correct it.

Xiaoxiang

Shear wave splitting results and possible anisotropy mechanisms in D" beneath the western Pacific

I will present the splitting time results and the anisotropy strength results I got in this region. All splitting time values are between -0.92s~3.21s with the average of 1.1s, and all anisotropy strength values are between -0.45%~1.56% with the average of 0.52%. I concluded that VTI is the main theoretical geometry for the large scale LMA (lowermost mantle anisotropy) observed in this region, and SPO (space-preferred orientation) of horizontal sheets or disks containing slow materials can be a very good mechanism. These results above imply that mantle flow is predominantly horizontal or near-horizontal throughout D" in this region.

Paul

Seismic cross-talk

Cross-talk between seismic events takes many forms. For example, cross-talk can occur between different experiments, different arrivals, different wave types, different wave modes, etc. All lead to artifacts. I review various forms of cross-talk artifacts and suggest solutions based on wavefield structure. This is a continuation of my preceding seminar, with examples illustrating problems and solutions.

11/13

Greg

Reverse-time migration as a multiple scattering problem

At the core of reverse-time migration is the reconstruction of the wavefield in the model by forward propagation of the time-reversed recorded data. In a lab experiment, or a computer experiment using synthetic data, it is possible to completely surround the medium with receivers. Therefore, we know that reconstruction is possible in such a case. We can reconstruct forward and backscattered energy for all orders of scattering. However, surface seismic recording does not fully enclose the medium of interest. Energy is lost not only due to limited recording aperture, but also through radiation into the earth ("out the sides and bottom"). What are the implications for reverse-time migration, especially if we want to make use of multiply scattered energy? This will be a topic for further investigation.

Roel

Basic Research Needs for a Secure Energy Future

In the spring I will be panel leader of a workshop of the Department of Energy for "Basic Research Needs in the Geosciences for a Secure Energy Future". I will describe how DOE uses these workshops like this to define research directions that they will support, and also how these workshops are being used to generate funding for basic research.

John Stockwell

Viruses on the CWP Notebook computer

A nasty worm/virus is circulating via USB memory devices. We will put Norton Antivirus on the notebook computer, but if you have been exchanging data between the notebook and your personal computers, you likely have the virus/worm on your system. All USB memory sticks need to be reformatted prior to being used with the CWP notebook computer.

11/15

Derek

An Intuitive Framework for Real-Time Freeform Modeling – Implemented in 2D (The Fast Way)

By using properties of affine spaces and adding a pre-computation step it is possible to achieve a substantial speedup over the algorithm I discussed during the last seminar. I will present the derivation of this new algorithm, show of quick demonstration of the new algorithm and compare run times between to the two implementations.

Steve

Spectral Element and Rotated Finite Difference for High Density Changes

I will review the methodology of the standard 2^nd order time, 4^th order space O(2,4) standard staggered grid finite difference for elastic wavefields. I will follow up with a more detailed discussion of the rotated staggered grid method for heterogeneous media (O(2,2)), and limitations related to 1) dip of the free surface and 2) computing times. I will also discuss spectral element methods, and how all of these techniques will factor into the inversion method I'll used to determine velocity structure/water saturation around a super-heated tunnel in volcanic tuff.

Laxmidhar

Which Migration should we use: Experience from real synthetic data

There are many different migration techniques available, but which migration to be used in practice is a big issue depending upon the type of data set, cost and what we are looking for. Taking into consideration all these aspects, we have used different types of migration algorithms (Stolt, PSPI, Gaussian beam, Gazdag, FD, Kirchhoff etc.) to typical synthetic data set generated over very simple model having thin sand channel and pinch out. Different migration algorithms behave differently for the same model. We have tried to obtain accurate image with conventional processing methods, still resolution is a big issue to image this model. Based on the premise of these migration algorithms and the experience we have gained from all these migration methods, we have applied the most suitable and cost effective migration technique (PSPI) to SEG Salt model. We obtain good image for this to certain extent. It is observed that if velocity model is accurate, the migration output is good, then the issue is how to obtain the most accurate velocity model!

11/20

John M.

Arctic 2D coherent noise example

A test line was acquired that is a very good example of the type of noise one encounters in Arctic seismic data, especially the noise generated by Vibroseis sources on floating ice. In this presentation displays of seismic data illustrate the coherent noise problems and suggest techniques for noise attenuation.

Gabi

Something about angle-decomposition

I will talk about what is most likely to be the main subject of my comps, which is find/implement an efficient way to perform angle-decomposition (reflection angle and azimuth at the imaging point). I am just starting getting familiar with the problem and I will talk about some papers I've been reading. I hope to get some other references, sugestions, opinions...

Masatoshi

Micro-seismicity associated with steam injections and velocity structure above the oil reservoir in Cold Lake

We determined the locations of micro-earthquakes due to steam injections into the oil-reservoir by using new velocity models, so that we can minimize residuals between observed travel times and calculated travel times for each event. High seismicity zone is observed around the depth of 380m, which is about 100 m above the top of the oil reservoir. The results of 1-D traveltime tomography to get velocity perturbations from the initial model, show that there is a low velocity zone corresponding to the high seismicity zone. S-wave velocity should be higher than the initial model by more than 20%, which is extremely large. We apply a geomechanical method to explain why the high seismicity is observed at the region and use seismic interferometry method to see if or not the velocity structures obtained by the relocation approach are reasonable.

11/22

NO TALKS

Thanksgiving

11/27

Jyoti

Reciprocity in VRS

The only requirement for source signature estimation using Virtual Real Source (VRS) is the source location coincide with one of the receiver locations. I show that is reeciprocity is valid in the seismic acquisition, then this requirement is no longer needed. The Green's function between two receiver locations is computed by cross-correlating the two receiver gathers there. If reciprocity is valid, a source gather can replace the receiver gather at the same location, thereby eliminating the requirement of the source location coinciding with a receiver location.

Niran

Gaussian Beam Migration

General definitions of Gaussian Beam and Gaussian Beam Migration (GBM) Method. Why GBM Method is useful than the other three (Finite Difference, Reverse Time and Kirchhoff) migration methods?

Dave

Apparent displacements from sequential shifts of time-lapse images

I estimate a vector field of apparent displacements from a sequence of local cross-correlations and shifts of time-lapse seismic images. I first estimate vertical displacements from local cross-correlations with 3-D Gaussian windows but for only vertical 1-D lags. After correcting one of the images for those vertical shifts, I then estimate inline horizontal components of displacement. After correcting for those horizontal shifts, I repeat this process for crossline horizontal components. I then repeat this sequential estimation and shifting for each vector component of displacement until all shifts are insignificant. This algorithm is simpler, more robust, and more efficient than directly searching for peaks in multi-dimensional local cross-correlations.

11/29

NO TALKS

12/04

Yuanzhong

An application of the seismic interferometry.

I will briefly explain the basic concept of the seismic interferometry. Then I applied this concept to extract the Green's function from the random noise in 2D. We found that we do not need too many sources to get the information we need. Also a numerical model was used to test this theory.

Suzie

12/07

Uwe Albertin (BP)

12/11