Zum Inhalt Zur Navigation

Profil der Anden Logo TU Berlin Logo GFZ Potsdam Logo Uni Potsdam

Subproject F3: Images and seismologic signatures of deformation processes in the Andean convergence zone

 

Research Field: Geophysics, seismics, seismology

 

 

Staff

 

Project-Leader(s)

Prof. Dr. Sergei Shapiro

 

Co-Leader(s)

Dr. Stefan Buske

Dr. Andreas Rietbrock

Prof. Dr. Frank Scherbaum

Dr. Peter Wigger

 

Members

Dr. Erik Saenger

Dipl.-Geophys. Christof Sick

 

Working area of the subproject F3

 

f3_working_area

Fig. 1: Working area of project F3

 

 

Abstract

Deformation and forces at convergent plate margins control the petrophysics and morphology of the subduction channel, fault systems and lithological heterogeneities of convergence zones. Through seismic and seismological imaging and characterization of these features in the Andean Zone, we hope to contribute to the understanding of stress coupling and force transmission across convergent plate boundaries and to the modelling of the material flux in the convergent zones as well as to the modelling of properties of convergent plates. In this way a better understanding of the geophysical signatures of active geodynamic processes at Andean type margins will be achieved. The present seismic and seismological image of the Andean convergence zone is mainly based on the data acquired by the SFB 267 group, DEKORP and the oil industry. The nature and origin of a number of remarkably high reflectivity zones as well as zones of very high seismic attenuation in different tectonic positions is still not yet fully understood. Further, it is not clear which of these anomalies are typical for which specific convergence zone enviroments. Seismic depth slices generally show reflection bands with laterally varying reflection strengths and thicknesses of up to 10 kilometers. We will employ true-amplitude, prestack depth imaging in conjunction with waveform modelling for statistically heterogeneous media in the Central and Southern Andes, in order to investigate the nature of the observed reflectors. Moreover, we propose to study statistical properties of small scale heterogeneities of the lithosphere and their contribution to the seismic attenuation, using the techniques of wave propagation in random media applied to local and regional seismic events. Finally, we want to compare the seismic images with the fine scale seismicity structure in the seismogenic contact zone, determined by high precision earthquake locations using relative relocation techniques. Closely linked to this issue is the question of the existence of a double seismic zone in this region. The comparison of results from the Central and Southern Andes will yield similarities and differences between an accretionary and an erosive continental margin.

 

 

Objectives, methods, work plan, and schedule

Objectives

This sub-project will contribute to a number of key issues by using seismological methods.

 

It is the aim of our sub-project to contribute to the solutions of these problems by means of seismic imaging, the study of scattering-attenuation, and relative hypocenter relocation. We will start analysing data from the southern part of the convergence zone and will compare the results from the Central Andes with those from the Southern Andes. We also aim to find out, on one hand, what is unique to the Southern or Central Andes and, on the other hand, what is common to both these regions. Finally, through comparison with other convergence zones, we hope to be able to identify general features of subduction zones.

 

Seismic Imaging

The structure, thickness and intensity of dominant features such as the Nazca reflector and the QBBS form the basis for petrological, geodynamical and structural geological interpretations from which, for instance, geometrical parameters of the coupling zone and the role of fluids are inferred. Thus the migrated sections on which these interpretations are based must represent a clean and true image of this tectonic subsurface signature.

 

In most of the ANCORP 96 profile, the Nazca reflector shows up with a thickness of 5-10 km. This apparent thickness is directly related to the length of the recorded signal. The complexity and length of this signal is, in turn, related to the internal structure of the reflector (e.g. multiple reflections from laminated structures, etc.). But other features may also affect the signal. It is for instance possible that small-scale heterogeneous structures above the reflector behave like scatterers and contribute to the complexity of the signal. In this case the true thickness of the Nazca reflector would be smaller and the internal structure less complex. In addition, the correct geometrical imaging plays an important role because a 2-D stack of such a strongly crooked profile as ANCORP 96 certainly leads to a smearing of reflection points and therefore to an apparent increase in thickness, especially for dip components perpendicular to the strike direction. The latter has been observed by Yoon (2001) for the QBBS. In our opinion an implementation of the migration in 3-D is absolutely necessary.

 

With the help of waveform modelling in random media, we will first work toward understanding the character of the medium above the reflector as well as the structure of the reflector itself. Can a scattering medium explain the length and shape of the signal or do we have to assume a reflector with strong internal complexity? If the reflector exhibits such a structural complexity, its internal structure must be disclosed, i.e. laminated, uncorrelated. In order to answer these questions we will analyse the crustal earthquake-events present in the ANCORP 96 data because they contain only the influence of the 'scattering' medium and no 'reflective' information about the reflector. If part of the recorded signals is due to scattering above the reflector, we will try to filter out this part. The necessary information about the statistical properties will also be inferred from the 'scattering' part of our group. This should then yield a signal which characterizes only the reflector. Only a few papers exist on this subject, e.g. Martini & Bean (2000), Henstock & Levander (2000).

 

After eliminating the 'scattered' part from the signal, the prestack migration has to be performed. The result will be an image from which it is possible to estimate reflection coefficients for both the Nazca plate and the QBBS. This kind of true-amplitude imaging has become a standard procedure in the oil industry but for deep seismic data it is still fairly new and atypical. Only a few approaches can be found in the literature (Simon, 1998; Zillmer et al., 2000), and these approaches are mostly limited to simple structures and yield significantly varied results. For complex structures, no procedure is yet available. This must be developed in the methodical part of this proposal. Within this procedure the estimate must reflect the structural complexity which means that the final result will be a kind of 'backscattering coefficient' which should allow conclusions about the scattering properties of the reflector (in much the same way as the 'reflection coefficient' allows the estimation of the impedance contrast).

 

F3_1

Fig. 2: Position maps of seismic lines in the Central and Southern Andes. Top: Overview of seismic profiles observed by the SFB 267, BGR and DEKORP. Thin solid lines indicate the onshore refraction profiles. Thick lines in the Pacific Ocean delineate the profiles of the CINCA 95 project (BGR), thick grey lines onshore show the location of the near-vertical incidence profiles PRECORP and ANCORP 96. Bottom: ENAP 1, 2, 4 and 6 are seismic reflection profiles the data of which are available from ENAP, Chile. The seismic refraction profile of the ISSA 2000 project is delineated by a solid grey line. The (estimated) location of the combined offshore/onshore seismic profiles of the SPOC project is indicated by dotted lines.

 

The above mentioned processing procedure will yield an image which should also allow further interpretation regarding the physical nature of the reflections, e.g. their relationship to the existence of melts or fluids, their spatial relation to seismicity and a comparison with seismic signatures observed in other active and fossil subduction zone settings, such as the Canadian Shield, the Urals, Vancouver Island and Japan.

 

The second 'imaging' part of this proposal is related to the application of the MMII approach to ENAP offshore data from the southern Andes and thus is focused on gaining a detailed fine-scale structural image of the interface zone between the upper and lower plates in this region.

We also intend to integrate data from the SPOC-project (SPOC=Subduction Processes Off Chile; this project will be part of the BMBF program GEOTECHNOLOGIEN, in the sub-program MARGINS/Kontinentalränder, location see figure 2). During the RV SONNE cruise 161 off southern Chile in November 2001 (to be carried out by BGR Hannover and GEOMAR Kiel), a number of multi-channel seismic reflection profiles will be carried out. We will augment the data set from these E-W lines with seismic refraction observations collected along aligned transects onshore. (The onshore part of this experiment is a joint BMBF proposal by J. Mechie and P. Wigger.)

Based on this image we want to provide answers to the following two questions:

 

a) How are the earthquake locations related to the plate interface?

The southern Chilean accretionary continental margin is an active seismic zone. We want to investigate the location of the earthquakes (results of the ISSA 2000 experiment; "old" sub-project C4-Asch) in relation to the structure of the seismogenic zone as inferred from enhanced processing of the seismic profiles. Are the earthquakes located between the upper and lower plates as proposed by Tichelaar and Ruff (1993) or within the descending oceanic crust? The latter suggestion would hint at active faulting processes, possibly due to the deformation and dehydration of the oceanic crust.

 

b) How does the subduction channel appear in seismic profiles?

Up until now it is not known whether the incoming material is subducted down to the depth of the seismogenic zone, as active seismic processing in this region could only image regions to a depth of about 9.5 km. We want to track the shear zone and the subduction channel to depths of approximately 15 km. Future projects submitted for funding to BMBF and GFZ will help to extend this image to even greater depths.

 

Answers to the questions above are important for the modelling of the process of tectonic accretion. With improved images we will gain more insight into the structures of the seismogenic zone and the continental margin. Finally, the comparison with results from Northern Chile should show differences and/or similarities in the seismogenic plate boundaries between the erosive and accretionary margins. This comparison should deliver additional constraints controlling mass transfer modes such as crustal thickening, the role of fluids and surface roughness, etc.

Scattering-attenuation

In addition to the above mentioned necessity to correct the seismic images for scattering, we want to quantify the amount of scattering attenuation in the subsurface of the central Andes. The application of the "epsilon-squared" method for analysing local events should reveal the statistical characteristics of the heterogeneities responsible for the scattering. The analysis of the envelope broadening of local events will provide an alternative approach to describe the heterogeneities in the statistical sense. Also the transport theory will be applied to characterize the coda in seismograms of local events. A detailed knowledge of the statistics of the heterogeneities will be used to estimate the reflection coefficient of deep reflectivity zones like the Nazca reflector and the Quebrada Blanca zone. On the other hand, scattering attenuation will be used to understand the physics of low Q values from seismological tomography (see sub-project G3-Asch). Furthermore, sub-project G3 will observe additional regional earthquakes in the Central Andes, which will be used for the application of the "epsilon-squared" method. A comparable statistical description of heterogeneities from different parts of the convergence zone (forearc, arc, back arc as well as erosive vs. accretionary margin) will provide a fundamentally new signature of deformation processes, leading to a large-scale statistical tomography of the tectonic regions. These investigations will help to address the problem of the role and geometry of fluids and melts versus that of heterogeneities causing scattering in the convergence zone.

 

Relative hypocenter relocation

Up until now the observed seismicity pattern of the Wadati-Benioff-Zone in the Central Andes still lacks the precision required for a detailed comparison with the active seismic reflection images. Relative relocation techniques can provide this precision and will improve the combined interpretation of seismicity and reflectivity structures considerably. This will give direct insights into the roughness/smoothness of the plate interface, an important parameter controlling material flux in the subduction channel, and facilitate the identification of particular reflection coefficient contrasts. Furthermore, the high resolution seismicity signature might answer important questions regarding the plate interface deformation. For example, are the seismic events aligned on planes in the subducted crust (reactivated faults) or randomly distributed? In addition, the internal structure of the Wadati-Benioff-Zone in the Central Andes will be accurately determined to answer the question of the existence of a double seismic zone and to discuss the nature of seismicity (frictional failure versus dehydration embrittlement and their controlling factors). Clear that these questions are of tremendous importance for understanding the nature of force transmission in the coupling zone. Here we also plan an intensive cooperation with subproject G3, where a part of the Wadati-Benioff-Zone will be also imaged with high precision using receiver functions along the ANCORP profile.

 

References

Henstock, T.J. & Levander, A., (2000): Impact of a complex overburden on analysis of bright reflections: A case study from the Mendocino Triple Junction. JGR,105, 21.711-21.726.
Martini, F., Bean, C., (2000): Seismic Imaging below highly heterogeneous layers. Expanded Abstracts, SEG meeting Calgary/Canada.
Simon, M., (1998): AVO analysis by offset-limited prestack migrations of crustal seismic data. Tectonophysics, 286, pp. 243.
Tichelaar, B. W. and L. J. Ruff (1993): Depth of seismic coupling along subduction zones. Journal of Geophysical Research 98, 2017-2037.
Zillmer, M., Müller, G., Stiller, M. (2000): Seismic reflections from the crystalline crust below the Continental Deep Drilling Site KTB - modelling and inference on reflector properties. JGR, submitted.
Yoon, M. (2001): Application of Prestack Depth Migration to a deep seismic data set (ANCORP '96). Master thesis, FU Berlin, in prep.

 

Methods, work plan, and schedule

Seismic Imaging

The migration methods described in this proposal have already been mostly developed, except for the estimate of the 'backscattering coefficient,' which demands extensive numerical modelling of seismic wavefields. The corresponding computer codes have already been developed and are at our disposal. The migration code is already developed and has been extended by Yoon (2001), but requires some further modifications. Waveform modelling will be done with the help of Erik Saenger. The work described here should be performed within a PhD thesis. The ANCORP and PRECORP data are available. Depending on experience, the time required to actually start the coding should be less than two months. After this, the calculation of synthetic seismograms for various scattering models with different medium and reflector complexity will begin and the results will be compared to the data (6-8 months). With the help of the resulting models, the influence of the 'scattering medium' will be removed and a preliminary prestack migration will be performed (4-6 months). Following this, the theory for the estimation of 'backscattering coefficients' will be applied to synthetic seismograms (in collaboration with Stefan Buske) as well as to data from ANCORP 96 and PRECORP, and eventually to other profiles (12 months).

With respect to the application of the MMII approach, we intend to start to reprocess the already available ENAP offshore profiles 730 and 732 between 37°30'S and 42°S. A few specific processing steps have to be applied before applying the MMII (e.g. suppressing of multiples). Then we want to apply MMII to envelope, and envelope normalized sections. Firstly, this avoids destructive interference caused by an imprecise stacking velocity model and has proven to be a reliable tool especially in cases of complicated reflections. Secondly, it avoids the influence of amplitudes in MMII procedures. The same processing procedures will be applied to multi-channel seismic reflection data from the SPOC project in cooperation with the BGR Hannover. The data acquisition is scheduled for November 2001, and the processing can start in January 2002. This work should be performed as part of the second "imaging" PhD thesis.

The results of the seismic imaging, including the determination of the reflection coefficients and the reflector geometries, will serve as direct inputs into the modelling projects F1 and G1. We also expect that our results will contribute to the discussion of density, rigidity and viscosity of the Andean lithosphere (see subprojects F2 and F4).

 

Scattering-attenuation

Theoretical results concerning scattering attenuation have already been translated into algorithms that compute attenuation values with the help of medium statistics. The latter have to be estimated from the results of fluctuation analysis of local event seismograms. For this, a generalization of the "epsilon-squared" method will be derived for spherical waves. In addition, we will use the Rytov approach to describe the envelope broadening. Known results of Sato & Fehler (1998) will be generalized and adopted to the geometry of the Andes. Methodical development will be mainly performed by associated researchers of our group who already have independent (FU) funding.

Numerical simulations will be performed in order to verify the theoretical considerations. Our group is able to simulate the propagation of elastic wave fields with the help of finite-difference codes in 2-D and 3-D. The "epsilon-squared" method generalized for local event seismograms and the analysis of their envelope broadening will be used to estimate the statistics of small heterogeneities. Using extensive numerical modelling we will also attempt to apply the transport theory to estimate the effects of the heterogeneities. The role of scattering in seismic attenuation will be discussed in the context of rheologic properties of the lithosphere. Herecooperation with subprojects G2, G3 and G4 will be important.

This research is intended to be performed in the framework of a PhD thesis.

 

Relative hypocenter relocation

Some of the ingredients for the techniques required for precise relative onset time determination and event relocation have already been developed and have been applied to other local earthquake data sets, e.g. in the Swabian Jura and the Gulf of Corinth, by some of the PIs (Scherbaum and Wendler 1986, Rietbrock et al., 1996). Here we want to extend the relocation procedure by using previously determined 3D velocity models instead of only 1D velocity models. In the first step of the analysis, the complete continuous data catalogue has to be reprocessed automatically to construct a complete event catalogue for the entire region. For the PISCO 94 data set this has already been done by Asch (1999). In the next step, the obtained data set will have to be cross checked and, if necessary, onset times will have to be adjusted manually (12 months). Subsequently, waveform similarities will be computed by cross-correlation techniques and relative onset times will be determined. The complete data set will then be inverted simultaneously for relative relocations and station correction terms (Joined Hypocenter Determination) using the known 3D velocity model. The results will also serve as geometrical constraints for the projects F1, G1 and G2.

This work will be done within a PhD thesis.

 

References

Asch, G. (1999): Präzise und schnelle Herdparameterbestimmung - eine Herausforderung an die moderne Seismologie. Habilitationsschrift an der FU Berlin, unveröffentlicht, 118 S.
Sato, H., & Fehler, M. (1998): Seismic wave propagation and scattering in the heterogenous earth, Springer.
Rietbrock, A., Tiberi, C., Scherbaum, F. & Lyon-Caen, H. (1996): Seismic slip on a low angle normal fault in the Gulf of Corinth: Evidence from high-resolution cluster analysis of microearthquakes, Geophys. Res. Lett., 23, 1817-1820.
Scherbaum, F. & Wendler, J. (1986): Cross spectral analysis of Swabian Jura (SW Germany) three-component microearthquakes recordings, J. Geophys., 60, 157-166.
Yoon, M. (2001): Application of Prestack Depth Migration to a deep seismic data set (ANCORP '96). Master thesis, FU Berlin, in prep.

 

 

Collaboration with external research groups

Names of important partners in the host countries

Argentina:

 

Bolivia

 

Chile

 

and in Germany

 

 

Publications

 

Literature

Reviewed publications

Buske, S., Lüth, S., Meyer, H., Patzig, R., Reichert, C., Shapiro, S.,Wigger, P. and Yoon, M. (2002): Broad depth range seismic imaging of the subducted Nazca Slab, North Chile.. - Tectonophysics, : ; . - []

Buske, S., Müller, T.M., Sick, C., Shapiro, S.A. and Yoon, M. (2002): True amplitude migration in the presence of a statistically heterogeneous overburden. - Journal of Seismic Exploration, 10: 31-40.

Haberland, C. & Rietbrock, A. (2001): Attenuation Tomography in the Western Central Andes: A Detailed Insight Into the Structure of a Magmatic Arc.. - J. Geophys. Res., : ; . - []

Hoshiba, M., Rietbrock, A., Scherbaum, F., Nakahara,, H. & Haberland, C. (2001): Scattering attenuation and intrinsic absorption in Northern Chile using full seismogram envelope.. - Seismology, : ; . - []

Müller T.M., and Shapiro S.A. (2000): Most probable seismic pulses in single realizations of 2-D and 3-D random media. - Geophys. J. Int, 144: 83-95. - []

Müller, T.M., Shapiro, S.A., and Sick, C.M.A. (2002): Most probable ballistic waves in random media: a weak-fluctuation approximation and numerical results. - Waves Random Media, 12: 223-245. - []

Shapiro, S.A. (2000): An inversion for fluid transport properties in 3-D hetergeneous rocks using induced microseismicity. - Geophys. Int., 143: 931-936.

Shapiro, S.A., Patzig, R., Asch, G., Giese, P., Wigger, P. Triggering of Antofagasta aftershocks - a diffusion like process?. - GRL, : ; . - []


© 2006 Freie Universität Berlin | Feedback |
Stand: 31.10.2006