SR05 by GEOTECHNOLOGIEN

Their hazard potential is especially high along tectonically active continental margins, such as on the western coast of South America or in Indonesia. More than 90% of all global earthquake activity, and almost all of the world’s highly explosive volcanoes, are concentrated at active continental margins. Passive continental margins such as the western coast of Africa, on the other hand are tectonically inactive – but of economic importance. Since continental margins are among the world’s most important populated areas and economic regions, their significance will increase in the future. As such, they have become a major focus of global research. Therefore the German research programme »Continental Margins – Earth’s Focal Points of Usage and Hazard Potential« was launched in 2004 as part of the R&D Programme GEOTECHNOLOGIEN. Goal of the programme is to improve the basic knowledge of continental margins and to mitigate their hazard potential. The funded projects focus on the following key themes: NAMIBGAS – Eruption of Methane and Hydrogen Sulphide from Namibian Shelf Sediments, SUNDAARC – High Risk Volcanism at the Active Continental Margin of the Sunda Arc, and TIPTEQ – From The Incoming Plate to Megathrust Earthquake Processes (South America). This volume contains a collection of the first research results and experiences of the funded projects. The presentations were given at a status seminar held at the GeoForschungsZentrum (GFZ) Potsdam, in June 2005. They cover a wide range of research opportunities and applications, such as geophysics, geology, mineralogy, volcanology, and biology.

Science Report GEOTECHNOLOGIEN

Continental margins mark the boundaries between continental plates and the oceans. Due to their particular geological situation, these areas are especially rich in raw material deposits – but they are also the sites where extreme natural phenomena such as earthquakes, volcanic eruptions and tidal waves occur. Continental margins therefore harbour a considerable risk potential, particularly as over 60% of the world’s population live within 100 kilometres of the coasts of the planet’s oceans.

Continental Margins – Earth’s Focal Points ...

Continental Margins – Earth’s Focal Points of Usage and Hazard Potential

GEOTECHNOLOGIEN Science Report

Continental Margins – Earth’s Focal Points of Usage and Hazard Potential Status Seminar GeoForschungsZentrum (GFZ) Potsdam 9-10 June 2005

Programme & Abstracts

The GEOTECHNOLOGIEN programme is funded by the Federal Ministry for Education and Research (BMBF) and the German Research Foundation (DFG)

No. 5

ISSN: 1619-7399

No. 5

GEOTECHNOLOGIEN Science Report

Continental Margins – Earth’s Focal Points of Usage and Hazard Potential

Status Seminar GeoForschungsZentrum (GFZ) Potsdam 9-10 June 2005

Programme & Abstracts

No. 5

Impressum

Schriftleitung / Editorship Dr. Alexander Rudloff Dr. Ludwig Stroink © Koordinierungsbüro GEOTECHNOLOGIEN, Potsdam 2005 ISSN 1619-7399 The Editors and the Publisher can not be held responsible for the opinions expressed and the statements made in the articles published, such responsibility resting with the author. Die Deutsche Bibliothek – CIP Einheitsaufnahme GEOTECHNOLOGIEN; Continental Margins – Earth’s Focal Points of Usage and Hazard Potential, Status Seminar GeoForschungsZentrum (GFZ) Potsdam 9-10 June 2005, Programme & Abstracts – Potsdam: Koordinierungsbüro GEOTECHNOLOGIEN, 2005 (GEOTECHNOLOGIEN Science Report No. 5) ISSN 1619-7399 Bezug / Distribution Koordinierungsbüro GEOTECHNOLOGIEN Telegrafenberg A6 14471 Potsdam, Germany Fon +49 (0)331-288 10 71 Fax +49 (0)331-288 10 77 www.geotechnologien.de geotech@gfz-potsdam.de Bildnachweis Titel / Copyright Cover Picture: ESA, MERIS (August 2003)

Preface

Continental margins mark the boundaries between continental plates and the oceans. They are places where extreme natural phenomena such as earthquakes, volcanic eruptions and tidal waves occur. Since over 60% of the world’s population live within 100 kilometres of the coasts, their risk potential is particularly high. At active continental margins, such as on the western coast of South America or in Indonesia, more than 90% of all global earthquake activity, and almost all of the world’s highly explosive volcanoes, are concentrated. Passive continental margins such as the western coast of Africa, on the other hand are tectonically inactive – but of economic importance. Since continental margins are among the world’s most important populated areas and economic regions, their significance will increase in the future. As such, they have become a major focus of global research. Therefore the German research programme »Continental Margins – Earth’s Focal Points of Usage and Hazard Potential« was launched in 2004 as part of the R&D Programme GEOTECHNOLOGIEN.

In a first 3-year period (2004-2007) a financial volume of 5,4 Million EURO will be spend by the Federal Ministry for Education and Research (BMBF). The funded projects focus on the following key themes: - NAMIBGAS – Eruption of Methane and Hydrogen Sulphide from Namibian Shelf Sediments - SUNDAARC – High Risk Volcanism at the Active Continental Margin of the Sunda Arc - TIPTEQ – From The Incoming Plate to Megathrust Earthquake Processes The main objectives of this first status seminar »Continental Margins – Earth’s Focal Points of Usage and Hazard Potential« is to bring together all participants from the different research projects to present their current work and to exchange their ideas. Alexander Rudloff Ludwig Stroink

Table of Contents

Scientific Programme ........................... 1 -

Abstracts of Oral Presentations and Posters (in alphabetical order) ....... 6 - 109 Authorsâ&#x20AC;&#x2122; Index................................. 110 - 111 Notes

Programme of the Status Seminar »Continental Margins – Earth’s Focal Points of Usage and Hazard Potential«, GeoForschungsZentrum (GFZ) Potsdam – 9-10 June 2005 Thursday, 9 June 2005 12:00-13:00 Registration and Poster Mounting (GFZ, Building H, Telegrafenberg) 13:00-13:30 Welcome by GeoForschungsZentrum (GFZ) Potsdam & GEOTECHNOLOGIEN 13:30-15:30 – Session TIPTEQ I (Chair: N.N.) 13:30 Oncken O.: TIPTEQ – Short Project Overview Grevemeyer I., Flueh E.R., Villinger H., Dahm T., Scherwath M., Heesemann M., Hofmann S.D., Ranero C.R., Rietbrock A., Tilmann F., and the TIPTEQ working group: Impact of the Lateral Variability of the Incoming Ocean Plate in South Chile on the Structure of the Marine Forearc and the Generation of Mega Thrust Earthquakes Krawczyk C.M., Araneda M., Asch G., Bataille K., Brasse H., Bribach J., Buske S., Dahm T., Galas R., Götze H.-J., Groß K., Haak V., Haberland C., Hackney R., Hanka W., Hofmann S., Kapinos G., Kind R., Lange D., Lüth S., Mechie J., Meyer U., Micksch U., Rietbrock A., Ritter O., Scherbaum F., Schulze A., Shapiro S., Stiller M., Wigger P.: Geophysical Images of Plate Interface Properties at the Southern Central Chilean Margin – The Onshore Geophysical Components of Project TIPTEQ 15:30 – 16:00 Coffee Break 16:00-18:00 – Session TIPTEQ II (Chair: N.N.) 16:00 Röser G., Behrmann J., Kopf A.: Sedimentary and Geotechnical Characterization of Trenchand Slope Sediments off Southern Chile: Preliminary Results Schilling F.R., Kukowski N., Gottschalk M., Tiwari R., Ramelow J., Knoll M.: Subduction Zone Fluid Processes and Their Impact on the Thermal Structure and Seismogenic Zone Reichel T., Wiedicke M.: Turbiditic Sequences in Sediment Cores from the Continental Margin of Southern Chile as Archive of Past Seismic Activity

Melnick D., Adam J., Anderssohn J., Bolte J., Bookhagen B., Echtler H., Kaufmann C., Klotz J., Kudrass H., Moreno M., Niedermann S., Oncken O., Reichel T., Segl K., Strecker M., Wiedicke-Hombach M.: Active Deformation and Coastal Tectonic Geomorphology Along the South-central Chile Margin 18:00 – 19:30 Short Poster Presentation (1-2 transparencies each) followed by Poster Session approx. 19:30 Dinner/Buffet in the Cantine of GeoForschungsZentrum Potsdam

Friday, 10 June 2005 09:00-11:00 – Session SUNDAARC (Chair: N.N.) 09:00 Reichert Chr., Lühr B.-G.: High Risk Volcanism at the Active Margin of the SUNDAARC Bohm M., Asch G., Fauzi F., Flüh E.R., Brotopuspito K.S., Kopp H., Lühr B.-G., Puspito N.T., Ratdomopurbo A., Rabbel W., Wagner D., and MERAMEX Research Group: The MERAMEX Project - A Seismological Network in Central Java, Indonesia Kopp H., Flueh E., Rabbel W., Wagner D., Wittwer A., and the Meramex Scientists: Subduction Zone Processes in Central Java: Preliminary Results of the MERAMEX Amphibious Project Hess K.-U., Spieler O., Dingwell D.B., Müller S.: DEVACOM – Comparative Experimental Volcanology in Active Convergent Margins Reichert Chr., Klinge K., Faber E., Ibs-von Seht M., Hoffmann-Rothe A., Kniess R.: Krakatau Monitoring (KRAKMON) 11:00 – 11:30 Coffee break 11:30–13:00 – Session NAMIBGAS & DEPAS (Chair: N.N.) 11:30 – 12:30 NAMIBGAS Lass H.U., Siegel H., Endler R., Brüchert V., Schiedek D.: Toxic Gas in the Namibian Coastal Upwelling Ecosystem (NAMIBGAS): An Integrated Study of H2S Origin, Abundance, and Mechanisms of Eruptions in a Large Coastal Upwelling Environment Brüchert V., Currie B., Dübecke A., Endler R., Julies E., Peard K., Zitzmann S.: Spatial and Temporal Distribution of Hydrogen Sulphide and Methane Emission from Namibian Shelf Sediment

12:30 – 13:00 DEPAS Schulze A., Weber M.: Status on the Amphibious Seismological Pool # 13:00 – 14:00 Lunch break 14:00 Final Discussion & Further Planning approx. 15:30 End of Meeting

# no abstract available

POSTER SESSION (CHAIR: N.N.) – THURSDAY AFTERNOON (18:00-19:30) TIPTEQ Heesemann M., Villinger H., Grevemeyer I.: Heat-flux off Coast Chile Measured During RV Sonne Cruise SO 181-1b Scherwath M., Grevemeyer I., Flueh E.R., Ranero C.R., Kaul N., Weinrebe W., ContraresReyes E., Tilmann F., Gossler J., and TIPTEQ Working Group: Seafloor, Sediments, Seismicity and Shallow Structures Offshore Southern Chile - Selected Preliminary Results from TIPTEQ Cruise SO181 Hofmann S.D., Lange D., Haberland C., Rietbrock A., Dahm T., Tilmann F., and the TIPTEQ working groups: TIPTEQ – Passive Seismology: Signal to Noise Characterisation in Southern Chile Lange D., Rietbrock A., Haberland Ch., Bataille K., Hofmann S., Dahm T., Tilmann F., and TIPTEQ Research Group: The Southern TIPTEQ Seismic Network Covering the Chilean Forearc Between 41.5° and 43.5° S - Status Haberland Ch., Rietbrock A., Lange D., Bataille K., Hofmann S., Dahm T., Scherbaum F., and TIPTEQ Research Group: The TIPTEQ Seismological Network 2004/2005 in Southern Chile (Between 37°and 39° S) - Status Micksch U., Krawczyk C.M., Stiller M., Araneda M., Bataille K., Bribach J., Buske S., Groß K., Lüth S., Mechie J., Schulze A., Shapiro S., Wigger P., Ziegenhagen T.: High-resolution, Three-component Reflection Seismic Survey in the Southern Central Chilean Andes at 38° S: First Data from Project TIPTEQ Groß K., Buske S., Wigger P., Araneda M., Bataille K., Bribach J., Krawczyk C.M., Lüth S., Mechie J., Micksch U., Schulze A., Shapiro S., Stiller M., Ziegenhagen T.: Numerical Modelling of a 3-component Reflection Seismic Survey at 38 °S - Comparison with First Field Data from Project TIPTEQ Kapinos G., Brasse H., Ritter O.: Magnetotelluric Image of Fluids in the Plate Interface Hackney R., Götze H.-J., Meyer U.: Topographic, Bathymetric and Gravity Characteristics of Convergent Margins Heberer B., Rahn M., Behrmann J.: How to Decipher Upper Plate Denudation by Looking at Fission Tracks from the Lower Plate Sediments – A Concept for a Study of the Southern Chile Trench

Bolte J., Baes M., Klotz J.: Three-dimensional Finite Element Model of the Andean Subduction Zone Oncken O., Kaufmann C., Segl K., Anderssohn J.: Simulation of Surface Deformation During Subduction Earthquakes # Melnick D., Bookhagen B., Echtler H., Strecker M.R.: Mechanisms Linking Surface Processes, Megathrust Earthquakes and the Tectonic Geomorphology of the Arauco Region, Chile SUNDAARC Hess K.-U., Dingwell D.B.: DEVACOM â&#x20AC;&#x201C; A High-load, High-temperature Deformation Apparatus For Volcanological Studies (see Hess et al.) Mueller S., Spieler O., Kueppers U., Scheu B., Dingwell D.B.: DEVACOM - Density Distribution in Pyroclastic Deposits: A Comparative Study (see Hess et al.)

# no abstract available

The MERAMEX Project – A Seismological Network in Central Java, Indonesia Bohm M. (1), Asch G. (1), Fauzi F. (2), Flüh E.R. (3), Brotopuspito K.S. (4), Kopp H. (3), Lühr B.-G. (1), Puspito N.T. (5), Ratdomopurbo A. (6), Rabbel W. (7), Wagner D. (7) and MERAMEX Research Group (1)

GeoForschungsZentrum Potsdam, Telegrafenberg, Potsdam, Germany, E-Mail: mirjam@gfz-potsdam.de

(2)

Meteorological and Geophysical Agency, Jakarta, Indonesia

(3)

IFM-GEOMAR, Kiel, Germany

(4)

Gadjah Mada University, Yogyakarta, Indonesia

(5)

Institut Teknologi Bandung, Bandung, Indonesia

(6)

Volcanological Technology Research Center, Yogyakarta, Indonesia

(7)

Christian Albrechts Universität, Kiel, Germany

1. Introduction The Indonesian volcanic chain is an expression of the ongoing subduction processes along the Sunda island arc. Combined amphibious seismological investigations at 110°E were performed within the project MERAMEX (MERapi AMphibious EXperiment) to study a volcanic system as part of an active continental margin. The active strato-volcano Merapi in Central Java is one of the most active and hazardous volcanoes world-wide. The aim of the project is to acquire deeper comprehension about the relation of subduction zone processes and volcanologic arc processes. Central Java is part the Sunda subduction zone, where the Indo-Australian plate is being subducted beneath the Eurasian plate with a convergence rate of ~67 mm/yr. Around 110°E a well developed seismicity gap seems to characterise the seismogenic zone along the Sunda subduction zone. Obviously it is accompanied by a separation zone in the oceanic plate, which is manifested in very different crust ages of ~140 Ma east and ~70 Ma west of it, with a corresponding difference in heat flow density. The volcano Merapi is situated in this zone close to its eastern boundary. Furthermore, the seismogenic zone in the extension of the seismic gap under the Java Sea in the north shows an

accumulation of hypocenters in much greater depth. This indicates a correlation between the Merapi and its system of magma ascent and the seismicity gap. 2. Objectives of the Project A major goal of the MERAMEX project is the determination of a velocity model for the range above the Wadati-Benioff zone by the application of high resolution tomography procedures (Vp, Vp/Vs and Qp) in combination with the Receiver Function method. The travel time tomography shall provide a three-dimensional image of the lithospheric structure in the area of the Merapi with a resolution of better than 20 km. Important information about the conduits of magma ascent is expected from attenuation tomography. A detailed image of the seismicity distribution of the WadatiBenioff zone as well as the upper crust shall be derived with exact localization methods. The stress distribution within the downthrusting plate will be acquired from moment tensor inversion. In addition to the passive tomography experiment active refraction seismic investigations were carried out offshore while the airgun shots were registered onshore too. The combined on- and offshore experiment will allow a

Figure 1: Plate-tectonic setting of the MERAMEX experiment: Central Java is part of the Sunda island arc, which is formed by the subduction of the Indo-Australian plate beneath the Eurasian plate. The seismicity distribution along the Sunda arc is characterized by a gap around 110째E south of Java.

complete correlation of the Wadati-Benioff zone, and will reveal forearc structures up to below Central Java. The derived seismic velocities from the refraction seismic investigations can serve as essential a priori information for the tomography analysis. 3. Present Status and Results 3.1. Seismological Network The temporary seismological network was in operation for about 150 days from May to October 2004. It consisted of 112 continuously recording seismic stations covering a region of about 150 x 200 km. The stations were equipped with 99 short-period three-component seismometers (Mark L4-3D or Guralp CMG-40T) and Earth data loggers (EDL) recording with a sampling frequency of 100 Hz. The 13 broadband stations were operated with Guralp seismometers (CMG-3T and CMG-

3ESP) and recorders (SAM). The internal clock of the data loggers were regularly checked against Universal Time (UTC) using the GPS satellite signal. The average station spacing was about 20 km. Two of the seismic stations were installed 60 km north of the main network on two small islands belonging to the Karimunjawa island group in the Java Sea just above the accumulation of hypocenters in 600 km depth. The land based network is completed by 9 ocean bottom hydrophones (OBH) and 5 ocean bottom seismometers (OBS) which were deployed offshore at the sea floor operating for a period of 18 weeks. The station spacing of the ocean bottom instruments was about 40-90 km. 3.2. Data Processing During the experiment a total of 500 Gbyte of seismological data were acquired, which is stored and backed up both on hard disk and DVD. Presently, the compilation of the basis catalogue is in progress. A first triage of the seismological data obtained yields about 5 to 10 local earthquakes and a series of regional and teleseismic events per day. The local events are

Figure 2: The temporary seismological network: Beside the dense distribution of 112 stations on land, 14 OBH/S were deployed offshore above the seismogenic zone. Additionally, 2 stations were set up on Karimunjawa islands above the accumulation of the ~600 km deep earthquakes.

Figure 3: Example of a local earthquake: It took place on August, the 19th with a magnitude (Mb) of 5.3 and could be felt in Yogyakarta and surrounding areas. Seismic signals of the event are observed at all stations.

mainly concentrated in the seismic coupling zone. A very local accumulation of small events takes place in the southeastern part of the network (SE of Yogyakarta) which can be observed only at the AJ-, AK- and BK-stations. Furthermore, two ~600 km deep events located 113.1째E could be recorded on the seismic stations during the operation period. We present the status of the study and first results.

and students who participated in the fieldwork and made it possible to set up and maintain such an enormous seismological network. Instruments were provided by the Geophysical Instrument Pool Potsdam (GIPP), the Christian Albrecht Universit채t, Kiel and the IFM-GEOMAR, Kiel.

Acknowledgements This research was founded by the Federal Ministry of Education and Research (BMBF) and the Deutsche Forschungsgemeinschaft (DFG) within the special programme GEOTECHNOLOGIEN. Essential logistical support of the project was provided by the Volcanological Survey of Indonesia (VSI) Bandung, the Volcanological Technologicy Research Center (BPPTK) Yogyakarta, the Gadjah Mada University (UGM) Yogyakarta and the Meteorological and Geophysical Agency (BMG) Jakarta. We would like to thank alle Indonesian and German colleagues

Three-dimensional Finite Element Model of the Andean Subduction Zone Bolte J. (1), Baes M. (1), Klotz J. (1) (1) GeoForschungsZentrum Potsdam, Telegrafenberg, 14473 Potsdam, Germany, E-Mail: [bolte | baes | klotz]@gfz-potsdam.de

1. Introduction and Principles of Model Set Up To explain the present-day deformation in the Andean subduction zone and gain insight into the plate boundary dynamics we model the subduction zone with finite elements. The threedimensional finite element method allows the handling of very complex geometries (which occur almost everywhere in nature) and rheologically layered structures. Furthermore, with multiphysics modelling we can take many relevant physics into account. The robust design of the model and the solution of the finite element method as well as the postprocessing is done with the commercial highend finite element analysis program ANSYS. The conventional modelling of a subduction zone applies kinematic boundary conditions for the slab whereas our model is based exclusively on the geometry of the tectonic plates including the subduction zone and on gravitational forces, i.e. the model runs only based on ridge-push and slab-pull forces including frictional forces at the plates interface. 2. Objectives and Principles of the Model Analysis The first priority is the design of a suitable and realistic finite element model. On the one hand the model should include available topography data (e.g. the ETOPO5 data set), the ridgepush and slab-pull forces which should be established only by the gravitational force and the body load effects of the mountains. Furthermore, the model has to reproduce the rheologically layered structures of the earth,

i.e. a viscoelastic mantle as well as elastic continental and oceanic crust. These high requirements will be an advantage for further ambitious models. On the other hand this computational approach with the mesh-based finite element method constrains the quality of approximation of geometry. The finite element method needs a certain mesh quality (prevention of too small and too large angles, respectively) but due to the computational costs (and ANSYS license) the number of elements and nodes, respectively, are limited. As we are interested in surface deformation in the area where we have GPS-observations we aim an a priori adaptive mesh design with a finer mesh size in the area of interest and a coarser grid elsewhere. The validation of the mesh generation, especially concerning the mesh size, and the determination of mantle viscosity are first objectives. 3. Conclusions and Outlook The flexible finite elements modelling technique yields a better understanding of the subduction earthquake cycle and can explain the present-day deformation in the Andean Subduction zone. The first results show that the best fit to observation vectors occurs when we incorporate a mantle viscosity of 4 x 10**19 Pa s. For further studies, we will also consider the static and dynamic friction coefficients between the slab and the several layers and the

(temperature dependent) elastic material properties. Interpretation of the results and sensitivity studies to determine the first order impacts are objectives of future work. References Klotz, J., Khazaradze, G., Angermann, D., Reigber, C., Perdomo, R., Cifuentes, O. (2001): Earthquake cycle dominates contemporary crustal deformation in Central and Southern Andes, Earth Planet. Sci. Lett. 193, 437-446. Khazaradze, G., Wang, K., Klotz, J., Hu, Y., He, J. (2002): Prolonged post-seismic deformation of the 1960 great Chile earthquake and implications for mantle rheology, Geophys. Res. Lett., Vol. 29, No. 22, 2050. Khazaradze, G., Klotz, J. (2003): Short and long-term effects of GPS measured crustal deformation rates along the South-Central Andes, J. Geophys. Res., 108(B6), 2289, doi:10.1029/2002JB001879. Xia, Y., Michel, G.W., Reigber, Ch., Klotz, J., Kaufmann, H. (2003): Seismic Unloading and Loading in Northern Central Chile as Observed By D-INSAR and GPS, Int. J. Remote Sensing, Vol. 24, No. 22, 4375-4391. Hu, Y., Wang, K., He, J., Klotz, J., Khazaradze, G., (2004): Three-dimensional viscoelastic finite element model for post-seismic deformation of the great 1960 Chile earthquake, J. Geophys. Res., 109, B12403, doi: 10.1029/ 2004J B003163.

Spatial and temporal distribution of hydrogen sulphide and methane emission from Namibian shelf sediment Brüchert V. (1), Currie B. (2), Dübecke A. (1), Endler R. (3), Julies E. (1), Peard K. (4), Zitzmann S. (1) (1) Max-Planck Institute for marine Microbiology, Celsiusstrasse 1, 28359 Bremen, Germany, E-Mail: vbrucher@mpi-bremen.de (2) Ministry of Fisheries, National Marine Information and Research Center, Swakopmund, Namibia (3) Institut fuer Ostseeforschung Warnemuende, Seestrasse 15, 18119 Rostock, Germany (4) Ministry of Fisheries, Marine Research Station, Luederitz, Namibia

1. Introduction The southwest African shelf between 22°S and 27°S has near-continuous upwelling, high primary production, extreme water column oxygen depletion, and episodically occurring sulphidic bottom waters (e.g., Brüchert et al., 2003). Turquoise discolourations of the nearshore surface waters are a regular phenomenon during the austral summer and spring. These discolourations were traditionally interpreted as coccolithophore blooms, but more recent interpretations suggest that some of these patches reflect the presence of dispersed colloidal sulphur – an oxidation product of hydrogen sulphide (Weeks et al., 2004). During the occurrence of these patches, the water column is severely oxygen-depleted up to the photic zone, with occasional severe mortalities for the living resources (fish and crustaceans) in one of the largest marine ecosystems on earth (Hamukuaya et al., 1998). Understandingly, the processes behind the development of water column hydrogen sulphide have therefore been in the focus of fisheries scientists and marine ecologists. A characteristic feature of the central Namibian shelf is the diatomaceous mud belt, which has organic carbon concentrations as high 20 % by dry weight. This accumulation of organic matter permits high rates of carbon mineralization in the sediment. Since oxygen is already largely

consumed in the bottom waters, bacterial sulphate reduction and associated hydrogen sulphide production become dominant in the sediment (Brüchert et al., 2003). Ultimately, even sulphate availability is limited in these sediments, and methanogenesis starts a few centimetres below the sediment surface, which leads to the accumulation of free methane gas (Emeis et al., 2004). Current hypotheses see the shelf anoxia forced by a combination of remote and local processes (Monteiro et al., subm.). A critical component of this coastal upwelling system is the sudden, localized appearance and disappearance of hydrogen sulphide-containing surface waters. The high temporal and spatial variability of free water column hydrogen sulphide require a variable driving force, which could either be extreme, small-scale variability in water column circulation, or changing physical processes affecting sediment-seawater exchange. Emeis et al. (2004) suggested a close relationship between anoxia and episodically recurring eruptions of biogenic methane from the sediment. Methane gas and hydrogen sulphide in the unconsolidated sediments may be released following a lowering of the hydrostatic pressure either due to an increase in the gas pressure from below, or a decrease in the hydrostatic pressure. For a quantitative analysis of the various potential sources of hydrogen

sulphide – sediment or water column – regional and temporal distribution patterns of fluxes, concentrations, and budgets of hydrogen sulphide and oxygen are required.

2. Objectives of the Project The present study was initiated to obtain quantitative spatial and temporal data on the critical bacterial processes regulating hydrogen sulphide and methane concentrations in the sediment and the water column. Geochemical profile measurements, radiotracer incubations, bacterial counting and molecular ecological analyses of the bacterial community were combined to provide an assessment of the role of bacteria in the development of sulphidic shelf waters. An important objective was to understand the biochemical mechanisms regulating the rates of hydrogen sulphide production by sulphate-reducing bacteria: These rates are dependent on the hydrolytic and fermentative activity of microorganisms, which feed sulphate-reducing bacteria with metabolizable organic compounds. The transformation of organic carbon preceding the oxidation of organic matter by sulphate-reducing bacteria is therefore an important regulator for the rates of hydrogen sulphide production. The combined dataset would then allow addressing the following questions: 1. What is the rate of oxygen consumption in the water column and how does this rate compare to oxidation rates at the sediment-water interface? 2. What is the regional and temporal variation in hydrogen sulphide fluxes in the upwelling zone and what regulates the flux from the sediments? 3. How does the gas distribution in the shelf sediment compare to the flux of hydrogen sulphide from the sediment?

3. Present Status and Results / Methods & Results / Results For the purpose of this study, we have assembled a database from cruises on RV Meteor (Meteor expedition 57-3) and RV Alexander von Humboldt expedition AHAB (legs 3 and 4). In addition, two-month time series data were obtained with the research vessel of the Namibian Ministry of Fisheries RV Welwitchia between May 2001 and May 2004. To extend the database, also earlier data obtained during cruises with RV Meteor (M48-2), RV Poseidon cruise 250-2, RV Petr Kottsov BENEFIT cruise Leg 2, April 1997 were included as well. With this data base, our time period of observation extends from April 1997 until May 2004. No data are available for the year 1998. Altogether, sediment and water column data were acquired from over 130 stations for the area between 19°S and 27°S. Table 1 lists the types of biogeochemical studies, the methods applied, and aims of the individual procedures. Of the nine studies listed, data acquisition for all studies except (6) and (9) has been completed and data have been analysed and evaluated. Studies (6) and (9) have commenced, and are still under investigation. Bacterial sulphate reduction. In all stations on the shelf, more than 70 % of the sulphate reduction takes place in the top 10 cm of sediment indicating a very reactive pool of organic material. However, in the area between 23°S and 24°S, the amount of remaining organic material below 10 cm sediment depth is generally still very large. Sulphate reduction continues below 10 cm depth at low rates until all sulphate is consumed and methanogenesis starts. Incubation of sediment from the methanogenic zone with 14C-HCO3- yielded rates of methanogenesis as high as 50 nmol cm-3 day-1. These high rates of methanogenesis lead to methane saturation between 15 cm and 100 centimeters below the sediment-water interface. Steep opposing gradients of porewater methane and sulphate indicate anaerobic oxi-

Tabel 1: Biogeochemical and molecular biological methods.

where Thiomargarita dominate, a higher net flux is measured, indicating a strong control of the bacterial type on the hydrogen sulfide flux.

dation of methane coupled to sulphate reduction. Active degassing of sediment only occurs in areas, where the methane/sulfate transition zone is less than 15 cm deep in the sediment. Hydrogen sulfide production from anaerobic methane oxidation contributes up to 17 % to the total hydrogen sulfide production. Hydrogen sulfide fluxes across the sedimentwater interface are up to 50 % of the total amount of hydrogen sulfide produced in the sediment. Microsensor measurements have indicated that in regions where Beggiatoa dominates the sulfide-oxidizing community at the sediment-water interface, the net flux of hydrogen sulphide across the sediment-water interface is almost zero, whereas in regions,

Figure 1: Areal distribution map of (a) depth-integrated bacterial sulphate reduction rates, (b) diffusive hydrogen sulphide fluxes across the sediment-water interface, (c) annually averaged bottom water dissolved oxygen concentrations.

Areal distribution An areal distribution map of bacterial sulphate reduction rates indicates that the areas with enhanced bacterial sulphate reduction correspond well with areas of seasonal upwelling cells excluding the area of the Luderitz upwelling cell, where little sediment is deposited on the shelf (Fig. 1). The corresponding maps of hydrogen sulphide fluxes show a much smaller area of net hydrogen sulfide flux to the water column (Fig. 1). In addition, the sulfide fluxes are always a fraction of the bacterial sulphate reduction rates. The highest fluxes are restricted to three areas: (1) Between 25°S and 24°S extending approximately from Sylvia Hill to Meob Bay, (2) from 23°30’S to 22°50’S, i.e, Concepcion Bay to Pelican Point including the area of Walvis Bay, and (3) the area from Cape Cross to Henties Bay. Biogeographic maps of Beggiatoa and Thiomargarita spp. show different different distributions (Fig. 2). Beggiatoa is concentrated in the area north of Palgrave Point and is found north until Cape Frio, whereas Thiomargarita is most abundant in the area south of Palgrave Point and around Walvis Bay. Thiomargarita is also abundant between 24ºS and 25ºS near Sylvia Hill. Both bacterial species are also present between Concepcion Bay and Walvis Bay, but in lower abundances. These maps represent quasi-averages over 7 years of observation. Intra- and interannual fluctuations exist, but are integrated into the maps for the areas where stations were revisited several times. According to the maps, the large sulphur bacteria are only present when the flux of hydrogen sulphide to the sediment surface is at least 0.3 mmol m-2 day-1. In the presence of Beggiatoa, none of the hydrogen sulphide will enter the water column. In the presence of Thiomargarita, up to 50% of the hydrogen sulphide flux may penetrate the sediment surface and be distributed in the bottom water (Brüchert et al., 2003). The different physiological adaptations of these two bacteria

to hydrogen sulphide fluxes and concentrations make them good indicators for differences in the severity of bottom water anoxia and the potential localization of bottom water sulphide. Enhanced methane concentrations in the water

Figure 2: Areal distribution map of (a) Thiomargarita, (b) Beggiatoa, (c) annually averaged bottom water hydrogen sulphide.

column were only found at a very shallow station near Walvis Bay. Despite the observation of sediment craters, the water over the craters did not contain elevated methane concentrations, which suggests that most of the area of the crater interior was not emitting methane. Temporal variation We assessed the intra- and interannual variability at a selected station in shallow depth (28 m), with shallow gas saturation. Between May 2001 and May 2004, concentrations of porewater methane, hydrogen sulphide, and water column oxygen were determined nearly every two months (Fig. 3). The data reveal abruptly changing oxygen levels in the water column, in concert with variations in methane concentration and with fluxes of hydrogen sulphide. There was no apparent periodicity in the observed fluctuations, and except for one brief period oxygen concentrations in the bottom water were below 22 ÂľM. The most conspicuous feature were the short-term, extreme oxygen depletions over the whole water column, with sudden drops in surface concentrations of oxygen to values as low as 67 ÂľM (1.5 ml/l). A relatively stable chemocline was only present during the austral summer 2001/2002 (October 2001 until May 2002). During this period, two lowoxygen events occurred in December 2001 and

in May 2002. Oxygen-rich water intruded after May 2002, but was interrupted by another lowoxygen event in October 2002, which was also terminated abruptly. Subsequently, oxygen levels dropped in the deep water and the chemocline rose gradually throughout the year 2003 reachingthe highest levels in March 2004. Data acquisition stopped in May 2004 with an apparent return to better ventilated conditions. During time periods when a shallow chemocline was present, methane concentrations increased abruptly in the sediment so that the depth of methane saturation rose to a sediment depth as shallow as 4 cm. Sulphide fluxes also increased during these periods. The sulphide fluxes ranged from 0.02 to 11.4 mmol m-2 day -1, a variation of more than 2 orders of magnitude. However, the enhanced sulphide fluxes and methane concentrations did not coincide with the punctuated low-oxygen events. In two cases, the low-oxygen event preceded the period of increased methane and sulphide fluxes. This observation is in contradiction to a sudden methane- and hydrogen sulphide-driven eruption as the cause for the low-oxygen event. It is more likely that the period of good stratification was conducive for the accumulation of bottom water hydrogen sulphide. Nitrate concentrations, which were below the detection limit in March 2002, support this interpretation. There was no bottom water hydrogen sulphide before March 2002, but it may have been missed because of the relatively long sampling gap. Coincidentally, during the summer 2001/2002, there were numerous observations of turquoise near-shore surface water discolourations and hydrogen sulphide smell suggesting that hydrogen sulphide was present (Weeks et al., 2004). It is clear that an even higher sampling frequency would be required in order to capture the true dynamic nature of developing anoxia. 4. Conclusions The current results provide good evidence for significant lateral heterogeneity of hydrogen sulphide production and indicate enhanced production in the areas downstream of the upwel-

Figure 3: Time series plot from SEDLAB Station 1 of (a) water column dissolved oxygen, (b) diffusive sulphide flux, (c) sediment methane concentrations.

ling cells. The contribution of methane-driven hydrogen sulphide emission to bottom water sulphide appears to be minor, when viewed regionally and over year-long time scales. This conclusion derives from the observation that gas-charged sediment only covers a small fraction of the total area affected by bottom water hydrogen sulphide. Secondly, a reservoir problem exists. The volume of hydrogen sulphide deeper than 10 cm in the sediment is limited and cannot be rapidly regenerated following an eruption. Fueling the water column with hydrogen sulphide from deep reservoirs with recurring sediment eruptions as proposed by Weeks et al. (2004) would deplete the deep sediment at a rate faster than hydrogen sulphide can be regenerated given the slow rates of sulphate

reduction at depth. In addition, although disrupted sediments indicate violent gas eruptions in the past, cruise observations indicate that only a small fraction of the disrupted sediment surface actively degassed. At the present moment, the available biogeochemical data would favour a diffusion-controlled mechanism for hydrogen sulphide production in the water column. The rapid production of hydrogen sulphide in the topmost 5 cm of sediment is sufficient to balance the hydrogen sulphide flux across the sediment water interface. Hydrogen sulphide oxidation using intracellular nitrate in large sulphur bacteria usually oxidizes most or all of the hydrogen sulphide below the sediment water interface. The hydrogen sulfide flux to the bottom water is only accelerated during

periods, when bottom water oxygen and nitrate are depleted. Combined with small-scale, rapid overturning events, hydrogen sulphide can be transported to the surface and be detected by satellite as elemental sulphur. These events, however, are restricted to the nearest shore, and do not occur offshore (Ohde, pers. comm.). Temporal and spatial detection of these overturning events must be the focus of the next investigations. Overall, the importance of the sediment for regulating water column anoxia is significant. Using mass budgets of oxygen consumption by sulphide oxidation and water column respiration from published data on primary productivity and particle density in the water column, up to 25 % of the total oxygen consumption in the system is due to benthic uptake. The results from these calculations will be compared and further tested with the integrated biogeochemical model results. Our current results demonstrate the success and usefulness of both spatial and temporal biogeochemical sediment data as important tools in environmental management. An establishment of long-term monitoring time series at several stations on the Namibian shelf will be required to capture the long-term fluctuations in water column sulphide. 5. Acknowledgements We would like to thank the captain and crews and scientific participants from the various cruises to the Namibian shelf. Their support was instrumental for the acquisition of this extensivedata set. Uli Lass, Thomas Ohde, Herbert Siegel, Martin Schmidt, Heide Schulz, Thomas Leipe, Thomas Vogt, Kay-Christian Emeis are thanked for scientific discussions and supportive data. Supplemental funding for this work comes from the Max-Planck Society, the GTZ-funded program BENEFIT, and the DFG Research Center Ocean Margins at the University of Bremen.

6. References Brüchert V, Jørgensen BB, Neumann K, Riechmann D, Schlösser M, Schulz H. Regulation of bacterial sulfate reduction and hydrogen sulfide fluxes in the central Namibian coastal upwelling zone. Geochimica et Cosmochimica Acta 2003; 67: 4505-4518. Emeis K-C, Brüchert V, Currie B, Endler R, Ferdelman TG, Kiessling A, Leipe T, Noli-Peard K, Struck U, Vogt T. Shallow gas in shelf sediments of the Namibian coastal upwelling ecosystem. Continental Shelf Research 2004; 24: 627-642. Hamukuaya H, O'Toole M, Woodhead PMJ. Observations of severe hypoxia and offshore displacements of Cape Hake over the Namibian Shelf in 1994. South African Journal of Marine Science 1998; 19: 41-57. Weeks SJ, Currie B, Bakun A, Peard KR. Hydrogen sulphide in the Atlantic Ocean off southern Africa: implications of a new view based on SeaWIFS satellite imagery. Deep-Sea Research 2004; 51: 153-172. Zabel M, Brüchert V, Schneider RR. The Benguela Upwelling System 2003, Cruise No. 57, 20 January - 13 April 2003, Meteor Berichte. Hamburg: Universität Hamburg; 2004.

Impact of the lateral variability of the incoming ocean plate in South Chile on the structure of the marine forearc and the generation of mega thrust earthquakes Grevemeyer I. (1), Flueh E.R. (1), Villinger H. (2), Dahm T. (3), Scherwath M. (1), Heesemann M. (2), Hofmann S.D. (3), Ranero C.R. (1), Rietbrock A. (4), Tilmann F. (5), and the TIPTEQ working group (1) Leibniz Institut fuer Meereswissenschaften, IFM-GEOMAR, Wischhofstraße 1-3, 24148 Kiel, Germany, E-Mail: igrevemeyer@ifm-geomar.de (2) Fachbereich Geowissenschaften, Universitaet Bremen, Klagenfurter Straße, 28359 Bremen, Germany (3) Institut fuer Geophysik, Universitaet Hamburg, Bundesstraße 55, 20146 Hamburg, Germany (4) Department of Earth and Ocean Sciences, Liverpool University, 4 Brownlow Street, Liverpool L69 3GP, United Kingdom (5) Bullard Laboratories, University of Cambridge, Madingley Road, Cambridge CB3 0EZ, United Kingdom

1. Introduction South Central Chile is among the areas of the world with the highest levels of tectonic erosion, producing several kilometres of subsidence to the south of the city of Valdivia. This fact is perhaps best documented by the Golfo Ancud separating the Isla Chiloe from the mainland. The Gulf has been formed by subsidence of the continent and represents the southward continuation of the Valley Central. Further to the south, however, there is evidence for focused local uplift and a gap in arc magmatism. Such prominent changes in the structure of the forearc and continent are governed by the properties of the subducting ocean plate.

In this context slab-age-dependent effects affect the tectonics and volcanism of the overriding South American plate. Anomalously high regional forearc subsidence caused by tectonic erosion strongly correlates with younging ages of the subducting plate producing up to several kilometres of along-forearc subsidence in the region just north of the subducting ridge. Where the ridge itself is actively subducting, there is a focussed pulse of local forearc uplift and subsidence. Furthermore, a gap in arc magmatism exists just to the south of the region of active ridge subduction, in the region between 47°-49° S, where the subducted slab of age <~10 Ma is underlying the 'normal location' for arc magmatism.

At 46°S the Chile Ridge – an active spreading centre – is currently subducted. The Chile Ridge is offset by several left stepping transform faults, resulting in 50 to 300 km long spreading segments. These fracture zone offsets have created a situation in which neighbouring sections of oceanic lithosphere created at the same spreading axis are subducted under Chile at ages ranging from zero age at 46°S to 25 Mio. years near Valdivia at 40°S.

This area of strong lateral change in the structure and properties of the incoming ocean plate and margin tectonics was affected by the world’s largest earthquake ever recorded instrumentelly: the great Chile earthquake and tsunami of 1960.

The major goal of the GEOTECHNOLOGIEN programme TIPTEQ – from The Incoming Plate to mega Thrust EarthQuake processes - is to

evaluate the controlling factors of great subduction zone earthquakes and to understand the interrelation between earthquake nucleation, tectonic erosion and the properties of the incoming ocean plate. The aims and first results described here are from the Themenbereich A and focus on the ÂťLateral variability of the incoming ocean plate and of the marine forearcÂŤ. During a major sea going programme aboard the German research vessel SONNE geophysical data have been collected from December 6, 2004 to February 24, 2005. Two networks of Ocean Bottom Seismometers offshore Isla Chiloe and Isla Mocha are in operation until October 2005 to record the natural seismicity from the plate interface where mega thrust earthquakes nucleate.

2. Objectives and data collection during RV SONNE cruise SO181 A long-term goal of continental margin research is to study the processes by which oceanic plate subduction drives arc magmatism, continental accretion or erosion, and earthquake processes. One key first-order parameter shaping a subduction zone is the thermal structure (i.e. age) of the downgoing plate. RV SONNE cruise SO181 as part of the TIPTEQ initiative studied a roughly 1000 km long region along the Chile trench surrounding the Chile Triple Junction at 46Â°S. This area represents a natural laboratory to study subduction zone processes at various ages and hence thermal regimes. The TIPTEQ project is a multidisciplinary study that aims at the quantification of 1) Material fluxes (mass and fluids) along and across the forearc area, 2) Thermal structure of the oceanic plate and subduction zone, 3) Rheological behaviour of the subducting sediment along the interplate megathrust, 4) Seismic activity and generation of large subduction-related earthquakes, and

5) Relationship between material fluxes and volcanic arc products, including thermal modelling of the entire subduction zone system. The results to these goals will yield a detailed view of the entire subduction system. A comprehensive data set has been collected along five corridors with geophysical data at key locations during SO181 (Figure 1). Each corridor provides wide-angle seismic data, multichannel seismic data (either existing or new high-resolution data), heat flow measurements, multibeam bathymetry and magnetic field data. The corridors were complemented by mapping with multibeam bathymetry and magnetic field data. Additionally, two shortterm (1 month) and two long-term (9 months) marine seismological networks were deployed to obtain a comprehensive data set in the survey area. Growing evidence indicates that the oceanic lithosphere undergoes profound changes as it flexes at the outer rise and plunges into the trench. Multibeam bathymetry at different margins shows that the ocean plate can be pervasively broken by flexural normal faults. In addition, some large outer rise earthquakes (M=6-8) document that normal faults can in some cases cut to > 20 km depth. The consequence is that water may penetrate along permeable faults to mantle depths so that the oceanic mantle is serpentinised and the ocean lithosphere may be dramatically cooled by water circulation along the faults. Recent modelling and geochemical studies suggest that water circulation and serpentinisation of the ocean mantle may play a major role in the thermal structure of the plates and in magma generation at volcanic arcs. In addition, deserpentinisation may control intraslab earthquakes. Therefore, a major goal was to collect a comprehensive data set from the ocean plate immediately seaward of the trench in order to

understand the input of water into the subduction system. Multibeam bathymetry will show the response of the different oceanic crust segments to flexure and the variations in normal faulting. Magnetic data will provide detailed information on lithospheric age and the nature of tectonic boundaries. Wide-angle seismic data (P and S waves) will map the changes in crustal and/or mantle velocities and Poisson ratios along and across the ocean plate yielding an estimate of the amount of serpentinisation and thus on the amount of water added to the ocean plate. Heat flow measurements will show the changes in thermal structure related to age and to water circulation. The subduction of laterally heterogeneous ocean plate segments influences the evolution of the arc-forearc system. The processes, however, can only be understood by collecting data that yield information on the entire thikkness of the overriding plate and on volumes of different types of rocks along and across the forearc area. The geophysical corridors collected across the ocean plate were extended into the margin to evaluate the material flux variation related to changes in structure of the subducting ocean plate and trench sediment supply. Such data allow to quantitatively estimate material fluxes of accreted sediment and continental framework rock. Detailed velocities from prestack depth migrated seismic reflection profiles will yield information on the location and amount of fluid flow at the front of the margin. Geotechnical testing of the trench sediment, heat flow measurements and thermal modelling of the subduction system constrain the depths of metamorphic reactions and dehydration. Heat flow data also provide the observations necessary to estimate the temperature at the plate boundary at depth. Seismological observations are crucial to relate long term deformation observed to other methods and to the short term response of the structures. Earthquake depths, distribution and focal mechanisms determine the depth of faul-

ting in the ocean lithosphere at the outer rise. Furthermore they inform on the evolution of subducted structures of the ocean plate showing whether structures like fracture zones and normal faults formed at the spreading centre are reactivated. It is of particular interest to find out if the subducted part of the Chile Ridge still is expressed in the seismicity pattern beneath the forearc and volcanic arc. This is crucial for the attempt to locate the subducted ridge and understand the relationship between spreading centre subduction, the gap in arc volcanism and slab window processes. The combination of seismicity, heat flow measurement and data from frictional properties of the subducting sediment helps to understand the controls on the updip and downdip limits of the interplate zone of seismic rupture (seismogenic zone). For example, we can test the hypothesis that the updip limit occurs at the 100150째C isotherm and is related to the illite - smectite phase change, or whether other processes and other temperatures are important. 3. Summary of first results From December 2004 to February 2005 the RV SONNE cruise SO181 took place as part of the TIPTEQ project in southern Chile to acquire various geophysical and geological datasets across the subduction zone between 35째 S and 48째 S. Here, the oceanic Nazca plate, the oceanic Antarctic plate and the continental South American plate join at the Chile Triple Junction, south of which the active spreading centre is subducting. To the north of the triple junction, the Nazca plate is segmented by several fracture zones across our survey area, allowing us to study subduction of differently aged oceanic crust and thus different thermal regimes. In addition, a simple plate motion environment exists with a constant convergence rate between the Nazca and the South American plate, constant spreading rates at the spreading centres, and a homogeneous motion vector, turning this area into a favourable natural laboratory for subduction zone process studies. We have chosen five major east-west oriented corridors as transects for the data

acquisition, with crustal ages ranging from 25 Ma down to 3 Ma, four of the lines north and one line south of the triple junction. All transects were covered with seismic wide-angle refraction and vertical incidence reflection data, and the northern four transects also with heat flow either on or near the transects. Furthermore, two short-term and two longterm seismological networks of ocean bottom seismometers and hydrophones were deployed. The short-term arrays were located on top of the outer rise and recorded for about six weeks, and four short streamer profiles were shot across them, firstly to re-locate the ocean bottom receivers and secondly to obtain structural images. The long-term seismological arrays were deployed towards the end of this cruise to be recovered in October 2005. During the entire survey bathymetric profiling took place and several long magnetic profiles were collected. In total, 260 ocean bottom seismometer and hydrophone stations were deployed, with 30 instruments remaining on the seafloor for the long-term seismological networks, one instrument being lost, and 229 successful recoveries. The data quality overall is good to excellent, with only few components being weak or without data. Three types of airguns were used for shooting the seismic lines; for wide-angle refraction shooting these were two simultaneous bolt guns of 32 l each, and a cluster of 8 G-guns of 8 l each, i.e. nominally 64 l capacity for each wide-angle shot; for shallow reflection shooting into a seismic streamer only, we used two 1.72 l GI-guns. A total number of 45180 shots were fired during this cruise at an overall excellent performance rate. Initial results from onboard data analyses exhibited a high data quality, and superb results are to be expected. For example, the outer rise networks recorded several thousand earthquakes, almost 600 of which have been located during the cruise, and it appears that this pilot study exhibits a clear potential for analysing plate-bending related earthquakes at outer rise areas. Furthermore, the combined 1440 km of seismic wide-angle coverage along the five TIPTEQ corridors illuminate well the different sta-

ges of the subduction. Strong seismic anisotropy of about 7% was detected in the upper mantle, with the fast orientation matching the direction of spreading. Seismic reflection profiling produced clear images of the sedimentary cover. Remarkably, only small differences exist in the sedimentary thicknesses between the lines, despite the difference in ages of the oceanic crust. The sedimentation rate appears high except around 41째 S. The reflection data also confirm the existence of a bottom simulating reflector, indicating gas hydrates, and this reflector will allow heat flow estimates. However, from our in-situ heat flow measurements, we directly measured values in the trench, on the outer bulge, and landwards of the trench. Heat flow values near the deformation front are highest on the youngest crust (150-200 mW/m2) and very low on the oldest crust (20 mW/m2). However, anomalously low heat flow values around 20 mW/m2 in the trench seem to be influenced strongly by sedimentation effects, and extremely low values of 7 mW/m2 in the outer rise area offshore Isla Chiloe correlate with apparent sea-water inflow into the oceanic crust. Due to advection of heat into the subduction zone by the downgoing lithosphere, heat flow values decrease landwards of the deformation front to values of 35-70 mW/m2. High resolution bathymetry data were collected along about 17600 km of ship track, complementing existing multibeam data. These bathymetric images show a well developed sedimentary cover, however no surface-cutting faults from plate bending were identified in the new areas, though high resolution seismic reflection data indicate that basement faults continue in the sedimentary blanket. Finally, the magnetic field measurements successfully detected the magnetic field reversals imprinted in the oceanic crust and allow an accurate age determination of the incoming plate. Profiles along these anomalies, roughly north-south oriented, did not detect possible magnetite from serpentinisation, however anomalies from fracture zones could be measured.

Acknowledgements We thank Captain Kull and Captain Mallon and their crews for excellent support and assistance during the RV SONNE cruises SO1811a, SO181-1b, and SO181-2. Funding is provided by the Bundesministerium f端r Bildung, Wissenschaft, Forschung und Technologie (BMBF) through the GEOTECHNOLOGIEN programme for Continental Margin research and for the SONNE cruise.

Numerical Modelling of a 3-component Reflection Seismic Survey at 38 °S – Comparison with first Field Data from Project TIPTEQ Groß K. (1), Buske S. (1), Wigger P. (1), Araneda M. (2), Bataille K. (3), Bribach J. (4), Krawczyk C.M. (4), Lüth S. (1), Mechie J. (4), Micksch U. (4), Schulze A. (4), Shapiro S. (1), Stiller M. (4), Ziegenhagen T. (4) (1) Freie Universitaet Berlin, Malteser Strasse 74-100, 12249 Berlin, Germany, E-Mail: kolja@geophysik.fu-berlin.de (2) SEGMI, Santiago, Chile (3) Universidad de Concepción, Chile (4) GeoForschungsZentrum Potsdam, Germany

Nearly all earthquakes with high magnitudes (Mw > 8) are generated along convergent margins. A quarter of the worldwide seismic energy in the last century was released in the Chilean part of the Andes, and thereof, most energy by the largest historically recorded earthquake in 1960 (Mw = 9.5). It is important to understand the triggering mechanisms and processes that shape those mega-thrust earthquakes. Therefore we have to understand the structural and petrophysical properties of the seismogenic coupling zone, where those earthquakes are suggested to initiate. That is one of the main aims of project TIPTEQ (from The Incoming Plate to megaThrust EarthQuake processes).

A c. 95 km long near vertical reflection (NVR) seismic profile was shot in Southern Central Chile at approx. 38° S in January 2005 (Fig. 1). The profile runs from the Central Valley (Victoria) to the Pacific Ocean (Quidico) along an east-west trending line, thus crossing the relocated hypocenter of the historic 1960 Valdivia earthquake [Krawczyk and the SPOC Team 2003]. An 18 km long spread made up of 180 evenly spaced three-component geophones was moved 4.5 km in a daily roll-along. Four NVRshots along the active spread each day (see Fig. 1, Fig. 2) yielding an up to 8-fold CDP coverage should deliver a high-resolution image of

Figure 1: Location map showing the reflection seismic profile line.

Figure 2: ESP- and NVRshot configuration (top) with CDP coverage for the ESP component (bottom).

the seismogenic coupling zone between the subducting Nazca Plate and the South American continent.

SPOC Team 2003, L端th et al. 2003). In the future, prestack depth migration (P-P, PS, S-P and S-S) will give a detailed structural image of the subduction zone that will serve as a basis for the geodynamic interpretation. The ESP-component of the project allows for traveltime and amplitude analysis, such as AVO and amplitude ratios. This analysis will give additional information on porosity, fluids and heterogeneity in the system and may be used to update the existing velocity model. Finally the SH-data will be used to analyze lateral and vertical variation of medium parameters. For all parts FD modelling will provide synthetic data for comparison and constraints.

An expanding spread (ESP) experiment component focuses on the down-dip limit (30-50 km depth) of the seismogenic coupling zone, harbouring the hypocenter of the 1960 earthquake. The configuration with an approximately 10-fold CDP coverage, made up of 19 shots with offsets up to 90 km (see Fig. 2), was designed to give a detailed image of this region. A SH-shot configuration with single-fold CDP coverage along the whole profile was carried out as a pilot study to test SH-wave generation by three-hole (Camouflet) shooting in a crustal regime. Using 3-component recordings, Swave images should be obtained to yield an improved picture of the petrophysical contrasts within the subduction zone system.

References Krawczyk, C. and the SPOC Team, 2003. Amphibious seismic survey images plate interface at 1960 Chile earthquake. EOS Trans. Am. Geophys. Union, 84: 301, 304-305.

All parts of the seismic survey are complemented by numerical modelling. Numerical simulations of NVR-, ESP- and SH-shots were performed in advance of the seismic survey. Subsurface wavefields were calculated down to 40 km depth, based on the velocity model of the 2001 SPOC data (Krawczyk and the

L端th, S., Mechie, J., Wigger, P., Flueh, E.R., Krawczyk, C.M., Reichert, C., Stiller, M., Vera, E. & SPOC Research Group, 2003. Subduction Processes Off Chile (SPOC) - Results from the amphibious wide-angle seismic experiment across the Chilean subduction zone. Geophysical Research Abstracts, 4.

The TIPTEQ seismological network 2004/2005 in southern Chile (between 37°and 39° S) – Status Haberland Ch. (1), Rietbrock A. (2), Lange D. (1), Bataille K. (3), Hofmann S. (4), Dahm T. (4), Scherbaum F. (1), and TIPTEQ Research Group (1) University of Potsdam, Institut fuer Geowissenschaften, Postfach 60 15 53, 14415 Potsdam, Germany, E-Mail: haber@geo.uni-potsdam.de (2) University of Liverpool, Dept. of Earth & Ocean Sciences, 4 Brownlow Street, Liverpool L69 3GP, UK (3) Universidad de Concepción, Concepción, Chile (4) University of Hamburg, Institut fuer Geophysik, Bundesstraße 55, 20146 Hamburg, Germany

1. Introduction Since November 2004 the large temporary seismological TIPTEQ network is installed in southern Chile, covering the forearc between 37° and 39° S. The network is part of the international and interdisciplinary research initiative TIPTEQ (From The Incoming Plate To megaThrust EarthQuake Processes) which is financed by the German Ministry for Education and Research (BMBF). The main aims of the project are to determine controlling factors for large earthquakes in the coupling zone of convergent margins and their interrelation with surface deformation. This shall be achieved by obtaining high resolution images of the seismogenic zone and the forearc structure, which will form the base for identifying the processes involved. Our studies focus spatially on the nucleation zone of the Mw=9.5 1960 Chile earthquake, the worldwide largest instrumentally ever recorded earthquake. It ruptured along a 1000 km long trench segment starting at 38° 10'S and propagating to the south. A coseismic displacement of up to 40 m occurred and a local tidal wave of up to 15 m height was generated. The network is installed and maintained by the University of Potsdam, the University of Hamburg (Germany), the Universidad de Concepcion (Chile), and the University of Liverpool (UK).

2. Experiment status Due to the availability of the seismic stations the deployment of the instruments took place in two steps. At first in November/December 2004 70 REFTEK and PDAS datalogger were installed.. Subsequently, in February 2005 Earth Data Loggers (EDL) exchanged these recorders, and additional 50 seismic stations were deployed resulting in a total number of 120 stations. All stations are equipped with Mark L4-3D short period 3 component seismometers. The instruments record continuously with a sample rate of 100 Hz; electric power is supplied by solar panels. The onshore network is complemented by 10 ocean bottom seismometers/hydrophones covering the outer forearc (see Figure 1). Most of the network will be dismantled in July 2005, but some 20 stations will remain operating in the region until October 2005. The station network is specifically designed for high resolution analysis of local seismicity as well as of teleseismic events. The coverage of a broad region with seismic station allows the observation and accurate location of deeper local events (from the coupling zone and at inter-mediate depth) in the study area. Furthermore it will allow the determination of the deeper earth structure by local seismic tomography and teleseismic studies (anisotro-

py, receiver functions). The dense deployment of stations in the centre (less then 7 km in the centre) assures... - a high waveform correlation facilitating the analysis of the whole wave field. - an increased quantity and quality of travel time picks in general and of S waves in particular allowing to calculate the S velocity models and the distribution of attenuation constraining the petrophysical interpretation. - an improved localization accuracy of coupling zone events (better then 2km) yielding focussed seismicity images - the possibility to accurately locate the shallow crustal seismicity. The latter two points will facilitate the joined interpretation of the results with planned or already conducted high-resolution studies

such as reflection seismology, MT measurements, and surface geological studies. Last but not least the dense station spacing and the deployment of ocean bottom seismometers focusing at the frontal forearc region will increase the spatial resolution of the determined velocity structure (tomographic images) compared to previous studies along the continental margin. The observation of the chemical explosions used by the controlled source experiment (Krawczyk et al., this issue) will further constrain our velocity model. The deployment of high quality data loggers with improved time accuracy (continuous GPS) will enable the application of cross-correlation techniques, which will be used for relative relocation of the seismicity. Furthermore it is planned to study the stress field and kinematics of

Figure 1: Station distribution of the network south of Concepcion (Chile). Grey triangles indicate onshore station locations, inverted grey triangles OBS/OBH station location. White triangles depict the position of young volcanos. The position of the seismic traverse between the towns of Quidico and Victoria is shown by the thick line.

Figure 2: Example of a local earthquake beneath the Arauco peninsula (200502-04 09:15:37.300 GMT, 37.926334째S/73.442833째W/14 km depth, mb=5.2 event (NEIC/PDE)). Clearly visible are the P and S phase. The traces (raw data, vertical components) are sorted according to epicentral distance and linear moveout corrected with a velocity of 6.5 km/s.

Figure 3: Example of a teleseismic event recorded at our network. It is a mb=6.0 event (NEICE/PDE) from 27.12.2004, located in an epicentral distance of about 151째 from our network in the Andaman-Sumatra region. The three different branches of the PKP caustic are clearly visible even in the raw data.

the crustal seismicity (e.g. at large strike slip fault systems such as the Lanalhue fault) by the analysis of focal mechanisms (moment tensor inversion). 3. First results Data quality control for the first four months shows that most of the stations were working correctly and have collected a total amount of more than 200 GB. Each day between 2 and 5 local/regional earthquakes and a series of teleseismic events are observed. Figure 2 shows an example of a mb=5.2 event (NEIC/PDE) within the network (4.2.2005). This earthquake beneath the Arauco peninsula was the first of a series of approximately 30 small events within 24 hours in this region. The aftershock sequence of the destructive Andaman-Sumatra earthquake (26.12.2005) provides a unique opportunity to study the earth inner and outer core using PKP and PKKP phases. Due to the dense station separation waveform coherency is very high enabling the detection of small amplitude phases like PKPdf. Figure 3 shows a typical example of a magnitude mb=6.0 event (NEICE/PDE) from 27.12.2004, located in an epicentral distance of about 151Â° from our network. The three different branches of the PKP caustic are clearly visible even in the raw data. Acknowledgements We are grateful to many Chilean landowners, companies, and institutions for support and for allowing us to install the seismic stations on their property. In particular we acknowledge the logistical support of the Universidad de Concepcion. Furthermore we thank all field crews for their excellent work under difficult conditions. The project is financed by the German Ministry of Education and Research (BMBF) through the GEOTECHNOLOGIEN/ KONTINENTRĂ&#x201E;NDER program. Land instruments were provided by the Geophyscal Instrument Pool (GIPP) of the GeoForschungsZentum Potsdam (GFZ).

Topographic, bathymetric and gravity characteristics of convergent margins Hackney R. (1), Götze H.-J. (1), Meyer U. (2) (1) Institut fuer Geowissenschaften, Abt. Geophysik, Christian-Albrechts-Universitaet zu Kiel, 24118 Kiel, Germany, E-Mail: [ rhackney | hajo ]@geophysik.uni-kiel.de (2) Bundesanstalt fuer Geowissenschaften und Rohstoffe, Geozentrum Hannover, Stilleweg 2, 30655 Hannover, Germany, E-Mail: uw.meyer@bgr.de

Introduction The regions associated with the two largest recorded earthquakes have characteristic gravity features that reflect the topographic and bathymetric expression of the subduction process. Examining these gravity features has potential to further our understanding of where rupture will occur during devastating great subduction zone earthquakes and of how the characteristic features of convergent margins were formed.

onshore forearc to the trench, the typical signature is a positive–negative anomaly couple. This represents a combination of the negative gravity associated with the deep bathymetric expression of the trench and a gravitational »edge effect« that represents the juxtaposition of continental and oceanic lithosphere. The position of the dense slab below the forearc also controls the magnitude and extent of the positive gravity adjacent to the trench and slope (Tasárová, 2004; Hackney et al., in prep.).

Subduction zone topography and gravity The topographic expression of active continental margins is well known. A subdued forebulge parallels a deep trench related to the bending of subducting plate under the continent. Beyond the landward trench slope, forearc basins may develop, while onshore a forearc high forms that is possibly separated from the elevated volcanic arc by a longitudinal depression. These intrinsic features are representative of the processes occurring at subduction zones and the degree of coupling between the subducting oceanic plate and overriding continental plate. If the subducting and overriding plate are »well stuck« (i.e. highly coupled), then the forearc topography tends to be deeper because the overriding plate is dragged down with the subducting plate. If the plates are only weakly coupled, then forearc topography would tend to be more elevated.

By examining the correlation between gravity anomalies and seismic moment release during subduction zone earthquakes, Song & Simons (2003) demonstrated that seismic moment release is concentrated in the parts of forearcs characterized by negative gravity. This suggests that negative forearc gravity is an indication of high coupling because, where coupling is high, the forearc is locked against and dragged down significantly with the subducting plate. The resulting subdued forearc topography is reflected in negative gravity. In these regions, large earthquakes occur infrequently and the moment release that occurs when the interface is ruptured is correspondingly high. In contrast, where coupling is lower, the overriding plate slides more freely over the subducting plate, which leads to more elevated forearc topography, more frequent, lower magnitude earthquakes and reduced seismic moment release.

The free-air gravity pattern of active margins reflects this topographic expression. From the

The two largest earthquakes ever recorded and the most extensive plate interface rupture ever

Figure 1: Gravity anomalies and seismicity (circles) in south-central Chile. (a) Satellite-altimetry derived free-air anomalies from the KMS2001 model (Andersen & Knudsen, 1998), (b) free-air anomalies from GRACE satellite data (the GGM02 model from the University of Texas), and (c) residual gravity derived by subtracting the longwavelength anomalies in (b) from the anomalies in (a). Plate boundary data are from Bird (2003). The yellow star marks the epicentre of the 1960 Valdivia earthquake, and the dashed line is an approximate indication of the extent of rupture during this earthquake.

observed occurred off south-central Chile in 1960 (Mw 9.5 Valdivia earthquake) and off northern Sumatra at the end of 2004 (Mw 9.3 Sumatra-Andaman earthquake). The gravity signature of the regions associated with these earthquakes show several similarities and also significant correlations with regions of plate rupture (Figures 1 and 2). Figures 1 and 2 show offshore free-air anomalies derived from satellite altimetry (KMS2001: Andersen & Knudsen, 1998), free-air anomalies from the most recent GRACE model (GGM02; http://www.csr.utexas.edu/grace/gravity/), and residual gravity computed by subtracting the long-wavelength GRACE gravity model from the KMS free-air gravity. Earthquakes from the NEIC with magnitude greater than 5 that have occurred since 1973 are also shown. South-central Chile Off south-central Chile, the focus of the TIPTEQ project, seismicity and residual gravity north and south of about 39°S show distinct

differences (Figure 1). North of 39°S, there is almost no seismicity coinciding with the trench but significant activity is evident in the forearc crust. This seismicity coincides with prominent positive residual gravity. In contrast, south of 39°S, seismicity in the forearc crust is significantly reduced, but prominent under the trench within the upper and middle crust. Residual gravity in this region (Figure 1c) is characterized by a band of positive anomalies paralleling and immediately landward of the trench and a band of near-zero or negative anomalies adjacent to the coastline. Importantly, the dominant positive anomalies north of 39°S terminate roughly at the northern limit of the rupture that occurred during the Valdivia earthquake. This means that negative anomalies in the offshore forearc correspond to rupture and, presumably, to high seismic moment release. Based on the Song & Simons (2003) interpretation, this suggests that plate coupling is high south of 39°S and significantly reduced north of 39°S.

Clues as to the cause of this inferred change in coupling come from the seismicity pattern in this region. Little seismicity occurs in regions of relatively negative anomalies, which suggests that the forearc in this region is strong and accumulating elastic strain. In contrast, the prominent seismicity associated with the forearc in the vicinity of the Arauco Peninsula suggests that strain is releasing and that the forearc in this region is weaker. Together, these observations suggest that the rheology of the forearc is an important control on where extensive rupture and great earthquakes can occur, as suggested previously by McCaffrey (1993). Sumatra and Andaman Islands The characteristic gravity pattern of south-central Chile is not isolated. Song & Simons (2003) have already shown that it is typical of many subduction zones, and similar features are also evident in the region affected by the recent Sumatraâ&#x20AC;&#x201C;Andaman earthquake (Figure 2). The residual gravity signature in this region (Figure 2c) also displays a positive anomaly parallel to and immediately landward of the trench. A

prominent band of negative anomalies also parallels the trench and extends from the source region of the earthquake northward to almost 15Â°N. The extent of the negative anomalies coincides almost exactly with the extent of the rupture zone which terminates near a region of positive forearc gravity. The relationship between seismicity and gravity is not so obvious in this region, probably because seismicity since 1973 represents activity during the build up to a big event, as opposed to south-central Chile where seismicity since 1973 represents activity after a large event. However, one particularly interesting correlation is evident: less seismicity exists in the regions of negative gravity than in the adjacent regions of positive gravity. If regions with negative gravity are regions of high coupling, then negative gravity should also correspond to the parts of the plate interface that are locked during the interseismic part of the earthquake cycle. If the plate interface under negative gravity is locked, then little seismicity would be expected to occur there.

Figure 2: Gravity anomalies and seismicity (circles) for the Sumatran region. Data in (a), (b) and (c) are as in Figure 1. Plate boundary data are from Bird (2003). The yellow star marks the epicentre of the 2004 Sumatraâ&#x20AC;&#x201C;Andaman earthquake, and the white dashed line is an approximate indication of the extent of rupture during this earthquake.

Conclusions The pattern of gravity anomalies at convergent margins, which reflect the topographic and bathymetric expression of the subduction process, appears to be an indicator of the degree of coupling across the subduction interface and may indicate where rupture is likely to occur during great subduction zone earthquakes. Negative forearc gravity seems to be an indicator of high coupling and conditions favourable to extensive rupture during larger and more infrequent earthquakes. In contrast, regions with positive gravity seem to reflect a situation involving lower coupling and the more continual release of the strain accumulated during convergence. Negative anomalies coincide with forearc basins, which could suggest that these basins are a direct consequence of the high coupling, or interseismic lokking, inferred for these regions. As a consequence, the basins could have formed as the result of basal erosion during seismic events that rupture the plate interface under the basins (Wells et al., 2003). Further understanding of plate coupling in subduction zones and the potential for extensive rupture during great earthquakes can be gained through comparisons between observed gravity and gravity predicted from dynamic modeling. High resolution calculations of forearc rigidity would also help to define forearc segmentation and to define the parts of forearcs that are most susceptible to extensive plate interface rupture and correspondingly large earthquakes.

References Andersen, O. B. & Knudsen, P. (1998) Global marine gravity field from the ERS-1 and Geosat geodetic mission altimetry, J. Geophys. Res., 103, 8129–8137. Bird, P. (2003). An updated digital model of plate boundaries, Geochem., Geophys., Geosys., 4 (3), 1027, doi:10.1029/2001GC000252. Hackney, R., Echtler, H., Franz, G., Götze, H.-J., Lohrmann, J., Lucassen, F., Lüth, S., Marchenko, D., Melnick, D., Meyer, U., Schmidt, S., Tasárová, Z., Tassara, A. & Wienecke, S. (in prep.). The segmented overriding plate and coupling at the south-central Chilean margin (36–42°S), Final Volume of the SFB267, »Deformation Processes in the Andes«. McCaffrey, R.M., 1993. On the role of the upper plate in great subduction zone earthquakes. J. Geophys. Res., 98 (B7), 11953–11966. Song, T.A. & Simons, M. (2003). Large trenchparallel gravity variations predict seismogenic behaviour in subduction zones, Science, 301, 630–633. Tasárová, Z. (2004). Gravity data analysis and interdisciplinary 3D modelling of a convergent plate margin (Chile, 36°–42°S), PhD Thesis, Freie Universität Berlin. Wells, R.E., Blakely, R.J., Sugiyama, Y., Scholl, D.W. & Dinterman, P.A. (2003). Basin-centred asperities in great subduction zone earthquakes: A link between slip, subsidence, and subduction erosion, J. Geophys. Res., 108 (B10), 2507, doi:10.1029/2002JB002072.

How to decipher upper plate denudation by looking at fission tracks from the lower plate sediments – a concept for a study of the Southern Chile Trench Heberer B. (1), Rahn M. (2), Behrmann J. (1) (1) University of Freiburg, Albertstr. 23B, 79104 Freiburg, Germany, E-Mail: bianca.heberer@geologie.uni-freiburg.de, jan.behrmann@geologie.uni-freiburg.de (2) Swiss Federal Nuclear Safety Inspectorate, 5232 Villigen-HSK, Switzerland, E-Mail: meinert.rahn@hsk.ch

1 Introduction The Southern Chile Trench, and especially the Chile triple junction, provide an interesting opportunity to study the interlude between plate tectonic forces, the coupling of converging plates and the subduction of a large-scale thermal source. In our proposed study, the fission track method will be applied to examine the denudation history of the Patagonian Andes from 31 to 47°S in its spatial and temporal variation. For such purpose analyses will be carried out on detrital apatites and zircons separated from sandy lithologies sampled along the Chilean trench axis. In a second step, river estuaries that feed the major trench fans and some key lithologies from the hinterland (source areas) shall be sampled and investigated. Fission track dating of detrital apatites and zircons has been successfully used to deduce source evolution in several orogenic settings (e.g. Garver et al., 1999; Bernet & Spiegel 2004). During a preliminary study on marine sediments from the Chile trench George and Hegarty (1995) dated detrital apatites from eight samples taken from ODP 141 cores in the immediate vicinity of the triple junction and successfully distinguished between two cooling events that could be correlated with specific periods of subduction. The younger event (7±2 Ma, late Miocene) could be correlated with the collision of the first and/or second

segment of the Chile ridge in the Golfo de Peñas region. The older stage (25 ± 5 Ma, late Oligocene) was interpreted as an effect of deformation processes along the Liquiñe-Ofqui fault zone due to oblique subduction. In contrast to the sparse fission track data on marine sediments, there is by far a denser distribution of published fission track ages onland (e.g. Thomson et al. 2001; Thomson 2002; Adriasola, 2003; Parada et al., 2000). The variation of exhumation and denudation rates with latitude (e.g. Murdie et al., 1993) can be explained by the varying ages of the ocean floor of the subducting Nazca and Antarctic Plates and, therefore, different mechanical and thermal strucures. Moreover, the special plate tectonic setting of a subducting ridge is expressed by heating and local rapid exhumation, followed by subsidence of the Pacific side of the South American Plate (Behrmann et al., 1992). 2 Objectives The major goal of our study is to characterize the dynamics of the different source areas between 31°S and 47°S latitude, in particular with regard to the younger history of denudation and its possible correlation with different modes of subduction along the Patagonian Andes. Specific periods of enhanced uplift and cooling, revealed by fission track ages of the trench sediments, might for the southernmost

Figure 1: Overview of the sample locations along the Chile trench; the dashed frame indicates the position of Fig. 2, map modified after GMT

part of the profile show a thermally induced denudation due to the subduction of the spreading-ridge between the Nazca and the Pacific Plate. Furthermore, times of stronger uplift of the orogenic hinterland are expected to correlate with periods of increased mechanical coupling of the converging plates. By focussing on detrital material, we anticipate to record the large-scale signal in contrast to more local signals when dating bedrock samples on-land. While the grains of an on-land source rock sample all share a common thermal history, detrital grains may be derived from multiple thermotectonic source areas. With reference to investigations on modern river sediments from the European Alps (Bernet et al. 2001), we envisage that the fission track age distribution from river estuaries as well as from the trench fans and subordinately from the trench itself will give a useful representation of cooling ages in the eroding source at the time of deposition. As an additional objective, we will use our data set to test the generally made assumption of a negligible transport time of the eroded sediments. With increasing distance between source and depositional area, there is a rising probability for intermediate and long-term sediment traps along the pathway. If such trapping did not take place or only at short-term duration, the fission track ages from source area, littoral and trench samples should be the same, and would support the general assumption that the lag time between closure age and depositional age is due to exhumation from closure temperature depth to the bedrock surface. 3 Methods Samples for our investigations come from gravity cores taken all over the TIPTEQ working area during SONNE cruise SO181, from the earlier cruises SO102 and SO156 as well as from core material of ODP Leg 141 (Fig. 1), with the latter mainly covering the area of the triple junction (Behrmann et al., 1992). While the gravity core samplesprobably include only Holocene to uppermost Pleistocene sediments, the material from ODP leg 141 also includes Pliocene sands, and will therefore allow to gain

additional insight into older stages of the denudation history of the hinterland in the near field of the Chile triple junction. Composite samples are derived by combining fine-grained sandy lithologies, for the ODP cores over up to 30 m intervals from the same stratigraphic range over the total depth penetrated at Sites 860B (617.8 mbsf), 861C (353.1 mbsf), 862A (22.1 mbsf), 862A (22.1mbsf), 862B (42.9 mbsf) and 863B (742.9 mbsf) (Behrmann et al. 1992). This procedure is necessary to improve the yield and the statistical validity of the results. Key information is expected from those samples taken from the trench fans (Bio Bio, Imperial, Tolten, Calle Calle, Chacao) during SONNE cruise 181 (Fig. 2). The importance of these sites results from the fact that the source area of any detrital grains in the taken samples should be equal to that of the estuaries that feed those trench fans. Therefore the provenance of these grains can be better restricted than it is possible for the other trench samples. A detailed description of the morphology of the trench fans and of the transport mechanisms in the Southern Chile Trench area is provided by Thornburg and Kulm (1990). In a second step, detrital apatites and zircons from present-day coastal river estuaries, that feed the at this stage already investigated trench fans, will be dated. As mentioned above, this data will show any lag times between the two deposits and indicate possible submarine sediment traps. In case of insufficient on-land information in the source area, we will in a third step sample and date selected locations in the hinterland, especially in those areas with restricted numbers of low temperature bedrock cooling ages. The exact sampling strategy will be strongly dependent on the very first data that are obtained from the trench sands. First results from ongoing separation work show that most trench sands are able to provide sufficient amounts of apatites, while for the zircons the yield is sometimes insufficient.

Figure 2: Sample locations (black dots) from Sonne Cruise 181 within submarine fans along the Southern Chile Trench; N of 33째S latitude volume of trench basin decreases abruptly, S of 42째S latitude trench is unchanneled and filled by sheet turbidites. Modified after Thornburg & Kulm (1990).

4 References Adriasola Muñoz, A., 2003: Low temperature thermal history and denudation along the Liquiñe-Ofqui Fault Zone in the southern Chilean Andes, 41-42°S. Dissertation thesis, University of Bochum, 150. Behrmann, J.H., Lewis, S.D.H., Musgrave, R.J., Arqueros, R., Bangs, N., Boden, P., Brown, K.M., Collombat, H., Didenko, A.N., Didyk, B.M., Forsythe, R., Froelich, P.N., Golovchenko, X., Kurnosov, V.B., Kvenvolden, K.A., LindsleyGriffin, N., Marsaglia, K., Osozawa, S., Prior, D.J., Sawyer, D.S., Scholl, D.C., Spiegler, D., Strand, K., Takahashi, K., Torres, M.E., Vega Faundez, M., Vergara, H.P. & Waseda, A., 1992: Proceedings of the Ocean Drilling Program, Part A: Initial Reports 141: 708. Bernet, M. & Spiegel, C., 2004: Detrital thermochronology; provenance analysis, exhumation, and landscape evolution of mountain belts. Geological Society of America 31, 378. Bernet, M., Zattin, M., Garver, J.I., Brandon, M.T. & Vance, J.A., 2001: Steady-state exhumation of the European Alps. Geology 29: 35-38. Garver, J.I., Soloviev, A.V., Brandon, M.T., Kamp, P.J.J. & Anonymous, 1999: Detrital fission-track thermochronology applied to sedimentary provenance studies. Abstracts with Programs Geological Society of America 31: 373. George, A.D., Hegarty, K.A., Lewis, S.D., Behrmann, J.H., Musgrave, R.J., Arqueros, R., Bangs, N., Boden, P., Brown, K.M., Collombat, H., Didenko, A.N., Didyk, B.M., Forsythe, R., Froelich, P.N., Golovchenko, X., Kurnosov, V.B., Kvenvolden, K.A., Lindsley-Griffin, N., Marsaglia, K., Osozawa, S., Prior, D.J., Sawyer, D.S., Scholl, D.C., Spiegler, D., Strad, K., Takahashi, K., Torres, M.E., Faundez, M.V., Vergara, H.P. & Wasedal, A., 1995: Fission track analysis of detrital apatites from sites 859, 860, and 862, Chile triple junction. Proceedings of the Ocean Drilling Program, Scientific Results 141: 181-190.

Murdie, R.E., Prior, D.J., Styles, P., Flint, S.S., Pearce, R.G. & Agar, S.M., 1993: Seismic responses to ridge-transform subduction; Chile triple junction. Geology 21: 1095-1098. Parada, M.A., Lahsen, A. & Palacios, C., 2000: The Miocene plutonic event of the Patagonian Batholith at 44 degrees 30'S; thermochronological and geobarometric evidence for melting of a rapidly exhumed lower crust. in Barbarin, B., Stephens, W.E., Bonin, B., Bouchez, J.L., Clarke, D.B., Cuney, M. und Martin, H., eds., Fourth Hutton symposium on The origin of granites and related rocks. Thomson, S.N., 2002: Late Cenozoic geomorphic and tectonic evolution of the Patagonian Andes between latitudes 42 degrees S and 46 degrees S; an appraisal based on fission-track results from the transpressional intra-arc LiquineOfqui fault zone. Geological Society of America Bulletin 114: 1159-1173. Thomson, S.N., Herve, F. & Stoeckhert, B., 2001: Mesozoic-Cenozoic denudation history of the Patagonian Andes (southern Chile) and its correlation to different subduction processes. Tectonics 20: 693-711. Thornburg, T.M., Kulm, L.D. & Hussong, D.M., 1990: Submarine-fan development in the southern Chile Trench; a dynamic interplay of tectonics and sedimentation. Geological Society of America Bulletin 102: 1658-1680.

Heat-flux off coast Chile measured during RV Sonne cruise SO 181-1b Heesemann M. (1), Villinger H. (1), Grevemeyer I. (2) (1) Universitaet Bremen, Fachbereich Geowissenschaften, Klagenfurther Str., 28359 Bremen, Germany, E-Mail: heesema@uni-bremen.de (2) Leibniz-Institut fuer Meereswissenschaften, IFM-GEOMAR, Wischhofstr. 1-3, 24148 Kiel, Germany, E-Mail: igrevemeyer@ifm-geomar.de

1 Introduction The detachment surfaces of subduction zones are the locations where the largest earthquakes in the world occur. Most regions located at convergent margins have experienced such mega-thrust events in historical times (Hyndman and Wang, 1993). Of the 40 largest subduction thrust fault earthquakes that took place in the 20th century five were along the Chilean margin. This includes the largest earthquake ever been reported (magnitude MW = 9.5) that occurred 1960 in the TIPTEQ study area (Kanamori, 1977; Grevemeyer et al., 2003). Current models of great subduction earthquakes state that the size of the ruptured zone, and therefore the magnitude of the event, is controlled by the thermal structure of the plate boundary. E.g. Oleskevich et al. (1999) postulate that the updip limit of stickslip behavior, where earthquakes can nucleate, coincides with a temperature of 100 to 150°C while the downdip limit is at 350 to 450°C. The key factor of a subduction zones’ thermal structure is the thermal state of the incoming plate which itself depends chiefly on the age of the subducted crust. Since the age of the subducted plate changes considerably and often abrupt along the southern Chilean trench, the length of the rupture causing the 1960 earthquake of ~850 km (Plafker and Savage, 1970) is very remarkable.

2 Objectives of the project The objective of this project is to determine lateral variations in the thermal structure of the subduction zone and its correlation with seismic activity in the area around 37°S to 45.5°S off coast Chile, where the rupture of the 1960 earthquake was located. Firstly, the thermal state of the incoming Nazca plate, which mainly controls the thermal structure of the subduction zone, has to be analyzed in several corridors along the trench. The plates’ age, hydrothermal circulation at the ridge flank, and possibly a reactivation of the circulation caused by flexure of the plate close to the trench are the most important factors determining the thermal state of the subducted plate. Following, the position of the seismogenic zone, thermally defined by its updip and the downdip limits, will be localized and compared to results from seismologic experiments. 3 Methods, Results and Present status Finite element method (FEM) models will assist us to determine the thermal structure of the incoming plate and the subduction zone. In order to constrain these numerical models heat flux measurements supplying boundary conditions on the incoming plate as well as on the continental slope are essential. Since in the past there were only a few heat flux measurements done in the working area, these boundary conditions will be mainly provided by our deployments of violin bow type heat flux pro-

bes (HF-Probes) during SO 181-1b and heat flux estimates from BSR depths. Two different HF-Probes of the violin bow type were deployed during SO 181-1b. The first probe was used for many years and is capable to measure thermal gradients with 11 thermistors that are equally distributed over a sensorstring that is 3 m long. The second probe (figure 1) was developed in the BMBF funded INGGAS project (BMBF grant 03G0564C) in GEOTECHNOLOGIEN in 2001/2002 at the University of Bremen by Dr. H.-H. Gennerich in

the working group of Prof. Dr. H. Villinger. The active length of this probesâ&#x20AC;&#x2122; sensor string, sampled with 22 thermistors, is extended to a length of 6 m. The increased penetration depth is especially useful for measurements at continental margins, where thermal gradients in the first meters of the sediments are often disturbed by bottom water temperature excursions. Besides measuring sediment temperatures, both probes are capable of determining thermal conductivities in-situ. Hartmann and

Figure 1: A. Recovery of the new HF-Probe during SO-181-1b. B. Details of HF-Probe operation.

Figure 2: Main study area: with heat flux stations on seismic corridors south of Valdivia, off coast Isla Chiloe and north of the Chile Triple Junction. Squares show locations of ODP Sites drilled during Leg 141 and 202.

Villinger (2002) describe the techniques applied to estimate undisturbed temperature gradients from the measurements that are altered by frictional heating during the probesâ&#x20AC;&#x2122; penetration. Additionally, they discuss the employed computations of thermal conductivities and heat flux values from HF-Probe data. Bottom simulating reflectors (BSRs) are abundant in most seismic transects that cross the continental slope of the study area. These BSRs mark the lower boundary of the gas hydrate stability zone, and temperatures at BSR depths can be estimated by utilizing the gas hydrate dissociation temperature-pressure function as published by Dickens and Quninby-Hunt (1994). BSR temperatures and depths calibrated with the sea-floor temperatures and heat flux values measured during SO 181-1b, and thermal conductivity data measured in boreholes of ODP Leg 202, provide the possibility to estimate the heat-flux over wide areas of the

continental slope, as demonstrated by Grevemeyer and Villinger (2001) and Yamano et al. (1982). During RV Sonne cruise SO 181-1b 63 successful heat flux measurements at 11 stations (H0401-H0411) were performed. These stations are distributed on seismic transects that image the incoming Nazca plate and continue over the trench on the continental slope (figure 2). All of these transects (SCS0401 -SCS0404) were shot during SO-181-1b, except for the RV-Conrad line 743-corridor #4 where heat flux station H0405 is located. Station H0411 located on the seismics transect SCS0404 offshore Concepcion is not shown, here. It extends a heat flux transect obtained in 2003 aboard the Chilean Navy research vessel Vidal Gormaz (Grevemeyer, pers. comm.). In the current stage of the project, the data of all HF-Probe stations has been processed, checked and turned into heat flux estimates.

Picking of BSR depths is in progress and also, first FEM models are under way. These models will, however, be based on a preliminary geometry of the subduction zone until final geometry models, based on the analysis of seismic experiments in the working area, are available. Heat flux values near the deformation front are highest on the youngest crust (150-200 mW/m2) and very low on the oldest crust (20 mW/m2). However, anomalously low heat flux values around 20 mW/m2 in the trench seem to be influenced strongly by sedimentation effects, and extremely low values of 7 mW/m2 in the outer rise area offshore Isla Chiloe correlate with apparent sea-water inflow into the oceanic crust. Due to advection of heat into the subduction zone by the downgoing lithosphere, heat flux values decrease landwards of the deformation front to values of 35-70 mW/m2. 4 Conclusions The heat flux data obtained during SO 181-1b fills a huge gap in the global heat flux data set, which contains no information in the TIPTEQ working area north of the Chile Triple Junction and south of Concepcion. Knowledge of heat flux in the working area is an essential constrain for the integrative modeling of subduction processes and their correlation with lateral variations in seismicity. Acknowledgements We thank Captain Kull and his crew for excellent support and assistance during the RV SONNE cruise SO181-1b. Funding is provided by the Bundesministerium für Bildung, Wissenschaft, Forschung und Technologie (BMBF) through the GEOTECHNOLOGIEN programme for Continental Margin research and for the SONNE cruise.

References Dickens, G., and M. Quninby-Hunt, Methane hydrate stability in seawater, Geophys. Res. Lett., 21, 2115–2118, 1994. Grevemeyer, I., and H. Villinger, Gas hydrate stability and the assessment of heat flow through continental margins, Geophys. J. Int., 145, 647–660, 2001. Grevemeyer, I., J. L. Diaz-Naveas, C. R. Ranero, H. W. Villinger, and Ocean Drilling Program Leg 202 Scientific Party, Heat flow over the descending Nazca plate in central Chile, 32°S to 41°S: observations from ODP Leg 202 and the occurrence of natural gas hydrates, Earth planet. Sci. Lett., 213, 285–298, 2003. Hartmann, A., and H. Villinger, Inversion of heat flow measurements by expansion of the temperature decay function, Geophys. J. Int., 148, 628–636, 2002. Hyndman, R., and K. Wang, Thermal constrains on the zone of major thrust earthquake failure: The Cascadia Subduction Zone, J. Geophys. Res., 98, 2039–2060, 1993. Kanamori, H., The energy released in great earthquakes, J. Geophys. Res., 82, 2981–2987, 1977. Oleskevich, D. A., R. D. Hyndman, and K. Wang, The updip and downdip limits to great subduction earthquakes: Thermal and structural models of Cascadia, south Alaska, SW Japan, and Chile, J. Geophys. Res., 105, 14,965–14, 991, 1999. Plafker, G., and J. C. Savage, Mechanism of the Chilean earthquakes of May 21 and 22, 1960, Geol. Soc. Am. Bull., 81, 1001–1030, 1970. Yamano, M., S. Uyeda, A. Y., and T. Shipley, Estimates of heat flow derived from gas hydrates, Geology, 10, 339–343, 1982.

DEVACOM: Comparative experimental volcanology in active convergent margins Hess K.-U. (1), Spieler O. (1), Dingwell D.B. (1), Müller S. (1), (1) Ludwig-Maximilians-Universitaet Muenchen, SMPG , Theresienstr. 41/3, 80333 Muenchen, Germany, E-Mail: hess@lmu.de

1. Introduction The aim of this project is to quantify the potential risk of selected highly-explosive volcanoes by improving the knowledge of the eruptive processes and the factors controlling these processes. A series of combined field and laboratory investigations will help to characterize the physicochemical properties of the latest eruptive products. These investigations deliver the basis for an interpretation of the eruptive processes. The comparison of the products of active volcanoes from different geological settings is supposed to reveal general information on the parameters controlling the eruptive behavior. The results will be of particular importance with regards to numerical modeling of eruptive processes, as well as for the interpretation of monitoring data. Especially in terms of future volcanic crisis these new insights are supposed to backup critical decisions by giving them a more reliable and objective basis. To arrange the project as effective as possible, the volcanoes to investigate were chosen in coordination with other research institutions. According to this the project comprise Colima (Mexiko); Mt. Redoubt, Alaska (USA); Bezymianny, Kamtschatka (GUS), Anak Krakatau and Kelut (Indonesia). Since the state of surveillance of these volcanoes is rather good, the results of this project are suitable to be synergetically combined with other monitored or investigated data.

2. Objectives of the Project - Field studies and sampling on Colima, St Augustin, Bezymiany, Anak Krakatau and Kelut. - Experimental analysis of the fragmentation and conduit dynamic / fragmentation due shear flow or vesicle overpressure (Experiments on the High T-Rig and »Fragmentation Bomb«) - Experimental analysis of the rheological properties of the sampled eruptive products - Methodical development of viscosity measurements on crystal bearing melts. (In cooperation with Voggenreiter GmbH.) The combination of field-study and laboratoryanalysis will demonstrate similarities or differences in the physical behaviour of the volcanoes. The given comparison of volcanoes with a high standard of knowledge will allow to integrate the expected results into knowledgebase of the regional scientists. 3. Present Status and Results / Methods & Results / Results Field studies / Sampling During 2004 and 2005 all five field study and sampling campaigns have been successfully carried out and a high number of samples has been shipped for experimental analyses. Volcán de Colima (Mexico) the historical eruptions comprise lava flows and explosive events. The current status is active and the volcano shows frequent (~ 4-6) explosions per day. The access and hence the sampling of the recent pyroclastic flow deposits proved to be more complicated and in parts impossible since

Figure 1: Augustine Island Alaska. The broad density variation of the deposits demonstrates the necessity of the performed measurements as a link of laboratory based experiments and a field related database.

an access path form East to the area called El Playรณn between Volcรกn de Colima and Volcรกn Nevado de Colima was destroyed during the rainy season 2003. The path from Volcรกn Nevado de Colima ends on the rim of the northern caldera wall, high above the plane of the upper La Lumbre Valley. All samples collected from Bread crust Bombs form the recent eruptions and Lava Flows (1880, 1961) [Luhr 2002] had to be carried up the northern slope. The block-and-ash flow deposits of 1999 in the lower valley of La Lumbre [Mora et al 2002] could not be reached since the activity of Volcรกn de Colima was high and the preferred direction of the bomb impacts and recent pyroclastic flows. The ascent path to Cordoban W (1998) block-and-ash flow deposits and overlaid lava flow (1998-1999) and Cordoban C was blocked due instable slopes and heavy rock falls. Measurements were performed on the deposits of the 1999 block-and-ash flow [BAF] deposits in the upper San Antonio valley and Cordoban E. Samples were collected from the E-Cordoban and San Antonio BAF and from the 1999 Lava flow [Kueppers et al.

2005]. Several samples of undated fall deposits were collected for the comparison to experimental pyroclasts. Augustine volcano (Alaska, USA) Analyses were performed on nine locations on the recent pyroclastic flow deposits. The access to the dome was hindered by the steep subvertical crater wall. The variation of the density distribution at the different locations demonstrate the strong transport sorting on the short travel distances (Fig. 1). A comparison of the deposits of Augustine and Bezymianny demonstrate the similar range of densities incorporated in the different eruption mechanisms of both volcanoes. (Fig. 2) Bezymianny (Kamchatka, GUS) Samples were taken from recent, proximal BAF deposits and deposits of the pyroclastic flow that followed the 1956 sector collapse. Distal blast layers were collected 8km to the south and distal ashes of the 1956 eruption on the slopes of Shiveluch. The Samples were measured after drying at the Institute of Volcanology and Seismology; Petropavlovsk-Kamchatsky.

Figure 2: Comparison of density measurements on Augustine and Bezymianny volcanoes. The close match of the distribution demonstrates the narrow density range of the different eruption mechanisms. Note that the distribution of Augustine volcano is modified since the pumice enriched Levees are not included in this graph!

Anak Krakatau (Sunda straight, Indonesia) and Kelut (East Java, Indonesia) The field campaign on both volcanoes were successfully accomplished in March 2005. Density Measurements Investigations of explosive volcanism and the modelling of related processes require profound knowledge of the physico-chemical properties of the rock material involved. Parameters such as the rock’s density and vesicularity highly influence the rheological properties as well as the fragmentation behaviour of the magma. As direct observation is not possible, in-formation on the spatial and temporal variability of the ascending magma’s vesicularity can easiest be achieved via the measurement of a statistically reliable amount of representative samples. In order to relate density/porosity distributions of eruptive products to specific volcanic settings and eruption characteristics, we evaluated and compared the results of five field campaigns in the circum-Pacific area. The densities of pyroclastic-flow and block-andash-flow samples on Colima (Mexico), St. Augustine (Alaska), Bezymianny (Kamchatka), Anak Krakatau and Kelut (Indonesia) will ad

to earlier studies on Unzen (Japan), and Merapi (Indonesia) that were measured directly in the field. Since the field-based density measurement method is based on the Archimedean principle taking a sample’s weight in air and under water, evacuated in a plastic bag to prevent water absorption [Kueppers et al., 2005], the method relies strongly on the climate. The results will support by detailed laboratory-based investigations on rock fragmentation [Spieler et al. 2004] and viscosity. The density measurements were performed in the field as long as climate and regional conditions allowed. Viscosity measurements We have developed a unique high load, high temperature deformation apparatus for studying in situ the non-Newtonian flow behaviour of magmas. The apparatus accommodates samples that are up to 100 mm in diameter and 100 mm long, and can be used to run constant dis-placement rate and constant load experiments. The rig is ideal for volcanological studies be-cause it uses experimental conditions that closely match those found in volcanic processes: temperature (25 to 1300 °C),

stress (0 to > 500 MPa), strain rates (10-6 to 10-2 s), and total strain (0 to 100%). The apparatus still has to be optimised, but we can already present some preliminary results. To study the flow behaviour of a reference melt, we performed a viscosity study at a constant temperature on a »NIST 710a« soda-lime composition. The sample was placed between the two pistons of the apparatus and heated up to the desired temperature (609 °C) above the calorimetric glass transition temperature (550 °C). After allowing the system to reach thermal equilibrium (~ 6 hours) in a parallel plate type experiment, load was applied (10, 50, 100, 200, 250 kN), holding that load for several 10 seconds. For applied loads < 100 kN (which correspond to stresses < 80 MPa), the stress versus strain rate relationship always behaves linearly and the viscosity remains constant with time (Newtonian flow regime). If the applied load was > 200 kN (which corresponds to stresses > 160 MPa), the stress versus strain rate relationship curves and the apparent viscosity decrease with time (non-Newtonian flow regime). If the applied load is raised to

250 kN »hot« cracks are produced and the sample is partly fragmented. Thus we are able to reproduce the work of the Bruckner-Group (Berlin) [Hess & Dingwell 2004a]. The major advance is, that we can use sample sizes in the range of several 105 mm3, which means we are able to measure natural samples that contain large phenocrysts, like Colima material. In addition we measured the viscosities of Lipari island (only minor amounts of crystals) and Yellowstone (Cougar Creek) rhyolites (only minor amounts of phenocrysts) at low (< 100 MPa) and high (> 100 MPa) stresses. These measure-ments shows the compatibility of the parallel plate type experiments with previous micro-penetration studies in the Newtonian flow regime (see Fig. 3). The measurement discrepancy of the aforementioned methods on natural samples from the same location is only 0.13 log Pa s, which is very close to the accuracy of the micopenetration method of +/- 0.06 Pa s. In contrast fibre elongation studies (extensional stress field) on rhyolites at high stresses show a marked Non-Newtonian flow regime. This result is now challenged by

Figure 3: Viscosity measurements with micropenetration and parallel plate method on Lipari island rhyolite (P3RR) on different samples from the same location.

Figure 4: Test cell for viscosity measurements on crystal rich lava samples. Measurements were performed on synthetic samples (SillyPutty TM ) and enriched with glass beads of different diameter to test the influence of the cell-geometry.

our new findings (compressional stress field), where we have found no significant NonNewtonian flow regime until brittle failure of the sample [Hess & Dingwell 2004b]. Viscosity Measuring Cell The design of the »syringe-test cell« for analyses on crystal and vesicle bearing melts has been tested on synthetic materials. We used SillyPutty™ as magma analogue, an viscoelastic polymer. Embedding glass beads in the analogue leads to rheology changes and allowed to analyse the geometrical constraints that are necessary for the proposed measurements on crystal bearing melts (see Fig. 4). The lower dog-hole and the geometry of the lower cell was designed to hinder blockage due crystal agglomeration. If the diameter of the lower opening is to small a sieve like crystalmush will increase the force necessary to push the melt through the nar-rowing. First high temperature cells will be assembled in May to test the stability of the graphite-liner. 4. Conclusions All volcanoes were successfully sampled and a first data base on the density distribution of the eruptive products was established. We have calibrated the new press with synthetic and natural samples and showed that we can discriminate between the Newtonian and Non-Newtonian flow regime. Measurements on samples form Colima are under way. The planing of the Syringe technique was completed and first tests were performed successfully. The fieldwork has been accomplished in march 2005.

Acknowledgements Prof. Nick Varley , Dr. Tina Neal, Dipl. Geol. Oleg Dirksen and Dr. Lothar Schwarzkopf gave major input to the fieldwork and helped to make this study successful. References Hess K-U., Dingwell D.B. (2004a) A High-load, High-temperature Deformation Apparatus For Volcanological Studies, Geoleipzig 2004, Leipzig. Hess K-U., Dingwell D.B. (2004b) A High-load, High-temperature Deformation Apparatus For Volcanological Studies, American Geophysical Union Fall Meeting 2004, San Francisco, USA. Kueppers U, Scheu B, Spieler O, and Dingwell DB (2005) Field-based density measurements as tool to identify pre-eruption dome structure: set-up and first results from Unzen volcano, Japan. JVGR 141/1-2, pp.65-75 Luhr J. F. (2002) Petrology and geochemistry of the 1991 and 1998-1999 lava flows from Volcan de Colima, Mexico: implications for the end of the current eruptive cycle. Journal of Volcanology and Geothermal Research, 117, 1, 169-194. Mora J.C.1; Macas J.L.; Saucedo R.; Orlando A.; Manetti P.; Vaselli O.(2002) Petrology of the 1998-2000 products of Volcan de Colima, Mexico. Journal of Volcanology and Geothermal Research, 117, 1, 195-212. Spieler, O., Kennedy, B., Kueppers, U. Dingwell, D.B., Scheu, B. and Taddeucci, J. (2004) A fragmentation threshold for the initiation and cessation of explosive eruptions. Earth and Planetary Science Letters 226, 139-148.

TIPTEQ – Passive Seismology: Signal to noise characterisation in southern Chile Hofmann S.D. (1), Lange D. (2), Haberland C. (2), Rietbrock A. (3), Dahm T. (1), Tilmann F. (4) and the TIPTEQ working groups (1) Institute of Geophysics, University of Hamburg, Germany, E-Mail: sonja.hofmann@dkrz.de (2) Institute of Geosciences, University of Potsdam, Germany (3) Department of Earth and Ocean Sciences, University of Liverpool, UK (4) Department of Earth Sciences, University of Cambrigde, UK

Introduction The large-scale, multi-discipline, experiment TIPTEQ (from The Incoming Plate to mega-Thrust EarthQuake processes) has been taking place on- and offshore southern Chile since October 2004, acquiring seismic (active and passive), magnetic, geothermal, and gravimetric data. The experiment studies one of the tectonically most active regions of the world, the subduction of the Nazca plate beneath Chile, one of the fastest moving plates on the earth with a subduction rate of about 9 cm/yr. The subduction produces strong uplift as well as close-by depression in Chile, e.g. near the Chile Triple junction and south of the harbour city Valdivia, respectively. Properties of the plate as its temperature, age or thickness, as well as structural features as seamount, fracture zones or forearc erosion control the continental deformation and the vertical movement. Understanding the controlling factors of the subduction zone dynamics and tectonics requires interdisciplinary studies and experiments, where seismology plays an important role. The Mw=9.5, 1960 South Chile event is the largest seismically recorded earthquake until today. It occurred in the region of the TIPTEQ experiment and ruptured about 800 km of the plate boundary between Concepcion and south of Ancud (37 – 45 °S). The co-seismic slip was about 21 m. Coseismic uplift on-shore of up to 6 m was observed. The earthquake

generated a major tsunami leading to substantial damage and life-loss not only in Chile but also far away at the east coast of the Pacific, e.g. in Japan. The Chile event was preceded and possibly triggered by a strong aseismic event or slow earthquake. In general, the occurrence and rupture of mega-thrust-earthquakes is controlled by the structure of the seismogenic zone of the subduction, by stress build-up and stress heterogeneities resulting from fault asperities and barriers. Seismicity and the study of small earthquakes is a major tool to understand the controlling factors of great subduction zone earthquakes. The passive seismological experiment within TIPTEQ aims to collect a unique dataset of seismicity occurring in the seismogenic zone of the 1960 Chile earthquake and along normal forearc faults within the bending subducting plate. Three seismic networks are deployed in the field, and the continuous data monitoring will lead to a unique dataset of teleseismic and regional phases open for different structural investigations. Various studies are planned within TIPTEQ in order to resolve the seismic structure, the local seismicity, the structure of the seismogenic zone, earthquake source properties and the stress field. Thus, the passive seismological experiment contributes a strong part to the general goal of the interdisciplinary project for the understanding of subduction zones and mega-thrust earthquakes causing large damage and possible tsunamis.

Figure 1: The TIPTEQ seismological networks. Dark triangles indicate the two land-network locations, the corresponding ocean bottom stations are shown in light diamonds. The dark diamonds are the outer rise arrays. Little white triangles are the volcanoes; the trench and the Chile ridge, as well as bathymetric data are depicted. Seismicity during the recording period until today is shown in light circles (from Univerity of Santiago and NEIC).

Objectives General aims are described in the abstract by Grevemeyer et al. (this issue). Major goals of the passive seismological experiment are to compile a complete high resolution image (less than 1 km) of the whole seismogenic coupling zone and to investigate the influence of the thermal structure of the incoming plate on seismicity in the region. In a 1st investigation period objectives of the passive seismological experiment are: 1. high precision earthquake localisation 2. source mechanism studies and retrieval of other kinematic and dynamic source parameters 3. structural studies, e.g. by receiver function techniques and tomography Seismic networks and the preliminary results of a signal to noise characterisation In total more than 170 (land and ocean bottom) passive seismic stations are recording local and teleseismic earthquakes at present (Fig.1). The network is subdivided into 3 arrays, with two of them consisting of an on- and offshore, the third only of an offshore part. The networks have been designed to allow high precision earthquake locations and source-analyses of subduction zone earthquakes. The northern array (located 37-39 °S/ 72-75 °W) consists of 120 land and 10 ocean bottom stations (OBS), mostly short-period. The average station spacing is about 5 km in the center and the network has an aperture of about 200x250 km. Further south, in the region of Chiloe, a network of 20 land stations and 18 OBSs is installed (42-43 °S/73-75 °W, 200x250 km aperture). Both networks will run continuously for about 9 months in total. These two combined on- and offshore arrays cover the whole distance between the volcanic chain and the trench at 38 °S and 43 °S. Additionally, a dense OBS network (2 patches with 15 stations, approx. aperture each 50x50 km, 43-44S/76W) was deployed

in the outer rise and has been designed to detect and locate micro-earthquakes. Up to now, the data from the outer rise network and the first months of the land networks have been collected. Only a few ocean bottom stations have weak components or no data. After battery problems during the first months at some land stations, the land networks are now running completely. These first datasets are of a very high quality, i.e. local earthquakes have been recorded with good signal to noise ratio. Examples of record sections of both land and ocean bottom data will be shown. Power spectral density calculatuions of noise records from the ocean bottom stations in the outer rise show typical patterns for representative noise records. The microseismic peak at about 0.2 Hz is present in all (near-)ocean data. For our hydrophone records the noise is relatively high (up to 10e4 Pa2/Hz) compared to data from the North Atlantic, which can be expected for the southern Pacific. The microseismic noise (between 0.1 and 5 Hz) is mainly generated by oceanic gravity waves and travel mostly as Rayleigh waves, which can be detected far from their origin. First comparisons in the networks indicate that the microseismic noise peak decreases towards the north of the network. The high frequency noise (above 5 Hz) is likely to be caused by wave breaking, local seismicity or local site resonances of mushy seafloor. Higher frequencies of 10 to 50 Hz are mostly due to shipping or marine mammals. At very low frequencies the high noise level is possibly due to infra-gravity waves that might be recorded in the region. Good detection of teleseismic phases can be assumed for the distinctive noise notch at 0.02 to 0.1 Hz. The noise notch level we observe for hydrophones (about 10e-1 Pa2/Hz) is comparable to the levels in the North Atlantic. First results will be presented.

Acknowledgements The project is funded by the BMBF through the GEOTECHNOLOGIEN programme for Continental Margin research and for the RV SONNE cruise. The land stations are provided by the Geophysical Instrument Pool Potsdam.

Magnetotelluric image of fluids in the plate interface Kapinos G. (1), Brasse H.* (1), Ritter O. (2) (1) Freie Universitaet Berlin, Malteser Strasse 74-100, 12249 Berlin, Germany, *E-Mail: h.brasse@geophysik.fu-berlin.de (2) GeoForschungsZentrum Potsdam, Germany

Previous long-period magnetotelluric studies in the Southern Chilean Andes revealed a modest high conductivity zone (HCZ) in the lower crust beneath the active volcanic arc and an additional HCZ associated with the Lanalhue fault in the forearc, running obliquely to the main morpho-structural units. Another finding is

Figure 1: MT sites at the South Chilean margin. Circles: Sites from campaign in 2000;Stars: TIPTEQ campaign; Squares: Monitoring stations.

based on the anomalous behavior of geo-magnetic induction vectors, which point consistently NE over the whole study area instead of the expected E direction due to the presence of the highly conducting Pacific Ocean. This spectacular effect is (and can only be) explained by anisotropy of the lower crust.

To corroborate these unexpected results, a combined on-/offshore magnetotelluric experiment was carried out in southern hemisphere summer 2004/2005 along a profile running perpen-dicularly to the trench and the volcanic arc from the incoming plate to the Argentinien border between 37.5째S and 39째S. As part of the multi-disciplinary TIPTEQ programme it aims especially at imaging fluids in the interface between the downgoing Nazca and the overriding South American plate. The oceanbottom instruments, which were deployed during RV Sonne leg SO181, incorporate newly developed components (fluxgate magnetometers and short-span telluric devices based on saltwater bridges) from Woods Hole Oceanographic Institution. A station separation of approx. 30-35 km was chosen for the sea-bottom line while the connec-ting onshore profile had a station spacing of 10 km. In a pilot study 3 continuously operating MT stations (together with seismological and GPS sites) have been set up along the profile and on Isla Mocha to monitor the long-time variability of geomagnetic transfer functions, related to seismicity from the seismogenic zone as well as from faults in the forearc and in the volcanic arc. This contribution will deal with the specific problems encountered in such an almost unique sea-land experiment and first results concerning an overview of calculated transfer functions will be presented. A further topic is an extension of the work towards better understanding of the conductivity distribution beneath active Llaima and Villarrica volcanoes, where course networks have been established.

Subduction zone processes in Central Java: Preliminary results of the MERAMEX amphibious project Kopp H. (1), Flueh E. (1), Rabbel, W. (2), Wagner D. (2), Wittwer A. (1), and the Meramex Scientists (1) IFM-GEOMAR, Wischhofstr. 1-3, Kiel, Germany, E-Mail: hkopp@ifm-geomar.de (2) Christian-Albrechts-Universität, Otto-Hahn-Platz 1, Kiel, Germany

1. Introduction Within the scope of the BMBF/DFG special initiative GEOTECHNOLOGIEN-Continental Margins, the joint interdisciplinary SUNDAARC project commenced in 2004. Within SUNDAARC, the subproject MERAMEX focuses on the high-risk volcanism and its tectonic implications on the active Sunda subduction zone. During the course of the project, RV SONNE cruises SO176&SO179 set out to collect geophysical data on the Java margin and the incoming oceanic plate to better understand the processes of fundamental importance related to the mechanics of plate convergence and the development of the Java forearc and volcanism. During the first cruise, SO176, in May 2004, 14 Ocean Bottom Seismic Stations (OBS) were deployed to monitor the natural seismic activity, augmenting a 120-element land array. The second cruise, SO179 in September/ October 2004, was primarily dedicated to the acquisition of seismic profiles. The main aim was to shoot into the land array to allow for some 3-D control of the plate interface. Three profiles, the two dip lines P16 and P18 reaching from close to the coast across the trench onto the oceanic plate as well as profile P19 located about 25 nm off the south coast of Java, were recorded onshore in a temporal seismological network which was installed in a dense grid of about 10-20 km station distance around Merapi volcano in Central Java. Due to the location of the eastern dip line across the Java margin and trench (profile P16), additional receivers (short periodic 95 MARK L4C-3D

seismometers in combination with Earth Data Loggers) had to be installed in East Java in August 2004. Altogether, 75 ocean bottom stations were deployed offshore, in addition to the acquisition of mini-streamer, bathymetric, gravimetric, and magnetic data. Here we report initial results of the refraction modelling (onshore and offshore) and bathymetric investigations, which shed some light on the subduction zone processes. 2. Objectives of the Project One of the key elements within the study area is the so-called seismic »gap« around 110°E, where seismic activity is highly reduced while accumulations of earthquake hypocenters occur to the west and east of this 100 km wide corridor. The study focuses on the relation of subduction zone processes and the arc volcanism as it is manifested in the active Merapi strato volcano. In addition to the seismicity and seismotectonics, the three-dimensional subbottom structure of the forearc and the activity pattern of Merapi represent the main objectives of the project. It has generally been accepted that fluids released during subduction of the oceanic plate trigger partial melting in the mantle wedge. These melts are the source for the active volcanism along the arc. The knowledge of fluid pathways and the distribution of fluids and melts in the forearc is essential for the modeling of deformation processes and for a comprehensive understanding of the relation between subduction and volcanic activity.

In addition to the interrelation between the forearc processes at depth and the onshore volcanism, the study was also geared at unraveling the fundamental tectonic framework of this little investigated margin. Modes of mass transfer at subduction zones vary intensely, including accretionary and erosive styles. Most margins are non-uniform either in alternating phases of accretion and erosion or in supporting accretive and erosive regimes simultaneously. While accretionary systems are comparatively easy to identify due to the material accumulation in a compressive setting, the ‘loss’ of material in erosive systems makes them a more obscure target. One of the main objectives of the geophysical data acquisition during the SONNE cruises was the investigation of the tectonic regime of the central Java margin. 3. Re-evaluation of the tectonic regime of the central Java margin The geological framework of a subduction zone, namely the thickness and the properties of the incoming sediments, the convergence

rate, and the oceanic plate roughness, control whether accretion or subduction erosion will dominate, since these features guide the amount of material necessary for accretion and subsequent growth of a wedge or prism. Off central and eastern Java, these factors would currently clearly favour tectonic erosion over sediment accretion. While the minimal sediment supply (0< trench fill < 1km) and the high convergence rate (7.7 cm/a) would also be sufficient for intermediate type processes, i.e. nonaccretive subduction, the subduction of severe oceanic basement relief causes active erosion of the forearc. Erosion may either occur along the front of the margin or along the base of the forearc wedge causing dismemberment of upper plate material along the shallow part of the plate interface and transfer of material to the downgoing oceanic plate. Frontal erosive processes along this margin segment are best documented by the broad retreat of the Java trench and deformation front in the projection of the oceanic Roo Rise (white dotted line in Figure 1). Between 109°E and 115°E, the

Figure 1: Geotectonic setting along the Sunda margin off Java displaying main bathymetric features and location of seismic profiles.

trench is deflected northward by approximately 50-60 km from its normal curvature trend. The collision of the Roo Rise with the forearc dominates the subduction processes off central and eastern Java. The Roo Rise represents a little investigated oceanic relief feature, which forms an irregularly shaped broad swell dotted with isolated morphological summits (Fig. 1). It continues into the trench and is interacting with the margin, where it causes broad-scale uplift of the entire forearc (Fig. 1). Judging from the large-scale topographic features visible in the bathymetry map, seamount and oceanic basement relief subduction has largely destroyed the previous outer forearc high as its ridge-like framework gave way to isolated topographic units arising from subducted basement relief (1 through 4 in Figure 1). At this stage of the project, it remains unclear what volume of the eroded material has been displaced landward and to what extent the material has been transported beyond the forearc.

are 350 km long and start on the oceanic plate crossing the outer arc high and the forearc basin onto the continental shelf.

4. Current Status, Data Description and Methods Three seismic profiles were shot with an array of four 32-l airguns at a shot interval of 60 s (black lines in Figure 1). One profile (SO17919) was especially shot sub-parallel to the trench near the shoreline to allow for energy registration by the landstations and to attain additional 3D structural information to the other profiles. Data on the two profiles (SO179-16, SO179-18) perpendicular to the trench were recorded by a total of 43 ocean bottom stations (OBH and OBS). These profiles

Onshore registration of marine shots: The land receivers stored the raw data of profiles SO179-P16, P18 and P19 in MiniSEED data format, which subsequently was converted to SEGY and sorted to receiver gathers. Spectral analyses of each receiver gather helped to determine the ideal filter section. A bandpass-butterworth filter with filter edges of 3-12 Hz was applied in addition to a notch filter dependant on the individual receiver gather to eliminate spikes. All traces were individually normalized to the maximum amplitude. Currently, data recorded by the landstations on P16, P18 and P19 and the relevant OBH for the tomography of the wide-angle data are processed and the first onsets are picked. The quality of the data recorded onshore is generally good and the range of the air gun signals is high, as seen on Figure 2. For example, receivers as far as CI3 (130 km offset) and CK1 (115 km offset) recorded signals of P19 and P16. Figure 2a shows a seismic section of station AH3 of profile P19- signals of all shots comprising the profile were recorded.

Figure 2a: Seismic section of station AH3 of profile P19- signals of all shots of the profile are recorded.

Figure 2b: Location map of the receiver stations and airgun profiles P16, 18, and 19. The ray coverage of profile P19 recorded on receiver AH3 is illustrated.

Figure 3: Wide-angle data section on the Java shelf, which registered energy from 300 km distance (SO179-18).

Figure 2b illustrates a location map of the onshore receiver stations, the airgun profiles P16, 18, and 19 and the ray coverage of Profile P19 recorded on receiver AH3. Marine seismic wide-angle and reflection data: The marine refraction data were corrected for position and frequency filtering was applied to adjust for time- and offset-dependent variations. To further improve the temporal resolution of the data, a gated Wiener deconvolution was applied to compress the seismic wavelet. The data quality is generally very good, as seen on various stations, which received signals to a maximum distance of 300 km (Fig. 3). Seismic reflection data were acquired simultaneously to the refraction profiles by a 50 m long four-channel streamer and are incorporated as a priori information into the refraction modelling. A forward modelling technique is performed on profiles SO179-16 and SO17918 to gain a starting model for inversion and

tomographic approaches. Future work will include tomographic inversion of the marine wide-angle data to unravel the structure of the forearc and possible fluid pathways. Incorporating the marine data and the land survey shall yield a composite model of the subduction zone and its linkage to the volcanic source of the Merapi volcano. Acknowledgements The SUNDAARC and MERAMEX projects are supported by the BMBF/DFG special initiative GEOTECHNOLOGIEN-Continental Margins. The geophysical data were acquired using RV SONNE as platform. We are grateful to Cpts. Papenhagen and Mallon and their crews for the professional work and their support at sea.

Geophysical images of plate interface properties at the southern central Chilean margin – The onshore geophysical components of project TIPTEQ Krawczyk C.M. (1), Araneda M. (2), Asch G. (1), Bataille K. (3), Brasse H. (4), Bribach J. (1), Buske S. (4), Dahm T. (5), Galas R. (1), Götze H.-J. (6), Groß K. (4), Haak V. (1), Haberland C. (7), Hackney R. (6), Hanka W. (1), Hofmann S. (5), Kapinos G. (4), Kind R. (1), Lange D. (7), Lüth S. (4), Mechie J. (1), Meyer U. (8), Micksch U. (1), Rietbrock A. (9), Ritter O. (1), Scherbaum F. (7), Schulze A. (1), Shapiro S. (4), Stiller M. (1), Wigger P. (4) [TIPTEQ Research Groups Seismics, Seismology, Magnetotellurics, Gravity] (1) GFZ Potsdam, Telegrafenberg, 14473 Potsdam, Germany, E-Mail: lotte@gfz-potsdam.de (2) SEGMI, Santiago, Chile (3) U de Concepcion, Concepcion, Chile (4) FU Berlin, Malteserstr. 74-100, 12249 Berlin, Germany (5) U Hamburg, Bundesstr. 55, 20146 Hamburg, Germany (6) U Kiel, Otto-Hahn-Platz 1, 24118 Kiel, Germany (7) U Potsdam, Karl-Liebknecht-Str. 24/25, 14476 Golm, Germany (8) BGR, Stilleweg 2, 30655 Hannover, Germany (9) U Liverpool, 4 Brownlow Street, Liverpool L69 3GP, UK

Introduction One of the main goals in subduction zone research is to understand the structural and petrophysical properties of the seismogenic coupling zone, and especially its down-dip end (Fig. 1). Here, mega-thrust earthquakes are suggested to initiate, but the triggering mechanism and processes that shape them are less understood. We will present the first data and preliminary results from the January-July 2005 onshore geophysical experiment components of the project TIPTEQ (from The Incoming Plate to mega-Thrust EarthQuake processes) that concentrated between 37°-40° S (Fig. 2). The surveys were designed to (1) yield a structural image also of physical parameters by s-wave information, (2) reveal the vertical zonation of oceanic crust an map active faults, (3) determine the conductivity structure of slab and fault zones including monitoring of transport processes, and (4) result in a 3-D asperity mapping. This should give finally a high-resolution

image of the seismogenic coupling zone between the subducting Nazca Plate and the South American continent in the area of the 1960 Chile earthquake hypocentre. Seismic and structural image of the plate interface The acquisition of the first complete and continuous high-resolution image of a seismogenic plate interface is a basic requirement of TIPTEQ. Therefore, a controlled source seismic experiment took place in the first weeks of 2005 with the aim to provide structural characteristics and an image of the present state of the plate interface ruptured during the 1960 earthquake. Together with the marine SPOC data, the newly acquired high-resolution 3component reflection seismic land data will yield a reflection seismic section that will cover the entire seismogenic coupling zone from its up-dip to its down-dip end.

In January/February 2005, a 95 km long, nearvertical incidence reflection (NVR) seismic profile was shot in southern Central Chile at c. 38°S (Fig. 2). 21 persons from Chilean and German institutions made the field work, together with a company responsible for drilling and explosives handling. 180 three-component geophones were deployed along an 18 km long spread, moving 4.5 km in a daily rollalong for three weeks. Explosive shots, with a spacing of 1.5 km, allowed an up to 8-fold CDP coverage (see Micksch et al., this volume). The W-E trending line runs across part of the Central Valley (starting due west of Victoria) and continues over the costal cordillera towards the Pacific, thereby passing the relocated hypocenter of the 1960 Valdivia earthquake (Krawczyk & the SPOC Team 2003). Challenges to be met included also a pilot study testing SH-wave generation respectively three-component recordings to yield finally an improved picture of the petrophysical contrasts within the subduction zone system. In addition, an expanding spread (ESP) experiment component focused on the down-dip limit (3050 km depth) of the seismogenic zone. Numerical simulations of NVR-, ESP- and SHshots were performed in advance of the seismic survey, and subsurface models were calculated based on the available refraction seismic velocity model (see Groß et al., this volume). The first data reveal different reflective bands in the crust as well as the downgoing plate along the entire profile. Observation of the wave field of earthquakes and deformation monitoring A temporary seismological network has been installed in southern Chile across the onshore and offshore forearc between 37-39° S (Fig. 2). It aims at a high-resolution analysis of local seismicity and teleseismic events including anisotropy and receiver functions studies. In November/December 2004, the first 70 datalogger (REFTEK and PDAS) were installed onshore. Those were exchanged in February 2005 by recorders from Earth Data (EDLs) and further supplemented by another 50 EDLs. All these 120 stations are equipped with Mark L4-

3D short period 3 component seismometers, recording continuously with a sample rate of 100 Hz. Offshore, 10 ocean bottom seismometers/hydrophones cover the outer forearc. The network covers an area of approximately 200x300 km. The denser station spacing close to the reflection seismic profile widens with increasing distance away from the line. Most of the stations will be active until July 2005. The expected improvement of localization accuracy of coupling zone events (better then 2 km) yielding focussed seismicity images and the possibility to accurately locate the shallow crustal seismicity will facilitate the joined interpretation of the results from high-resolution studies such as reflection seismology, MT measurements, and surface geological studies. So far, 2-5 local/regional earthquakes and a series of teleseismic events are observed each day. Among those are events up to a magnitude of M=5.0 which occurred within the network (see Haberland et al., Lange et al., this volume). Gravity and asperity mapping The free-air gravity pattern of active margins reflects in general the topographic expression of a subdued forebulge paralleling a deep trench. From the onshore forearc to the trench, this is usually a positive-negative anomaly pattern, where the position of the downgoing slab controls the magnitude and extent of the positive gravity adjacent to trench and slope. Both the Chile 1960 and the Sumatra 2004 earthquake occurred in regions that have a characteristic gravity signature and show correlations with the regions of plate rupture. By modelling the free-air anomalies derived from satellite altimetry and the latest GRACE model, gravity anomaly patterns are suggested to be an indicator of the degree of coupling across the subduction interface. In southern Central Chile, in the TIPTEQ area between 33-48° S, seismicity and residual gravity are examined (see Hackney et al., this volume). Little seismicity occurs in regions of relatively negative anomalies, which suggests that the forearc in this region is strong and accumulating elastic strain. In contrast, the promi-

Figure 1: Schematic representation of a subduction zone with lateral extent of geophysical experiment components applied at the Chilean convergent margin

nent seismicity associated with the forearc in the vicinity of the Arauco Peninsula suggests that strain is releasing and that the forearc in this region is weaker. Together, these observations suggest that the rheology of the forearc is an important control on where extensive rupture and great earthquakes can occur. Further studies comparing observed gravity and gravity predicted from dynamic modelling as well as high-resolution calculations of forearc rigidity would contribute in the future to define those parts of forearcs that are most susceptible to future plate interface rupture. Magnetotelluric image of fluids in the plate interface High-conductivity zones (HCZ) are revealed in the southern Chilean Andes by long-period magnetotelluric studies. They are found in the lower crust beneath the active volcanic arc and additionally in the forearc. Furthermore, the observed geomagnetic induction vector behaviour can only be explained by anisotropy of the lower crust. Within the project TIPTEQ, these unexpected findings are further investigated.

Aiming especially at imaging fluids in the interface between the downgoing Nazca and the overriding South American plate, a combined on-/offshore magnetotelluric experiment took place along a profile running perpendicularly to the trench and the volcanic arc from the incoming plate to the Argentinian border between 37.5째S and 39째S (Fig. 2). Between December 2004 and February 2005, 7 oceanbottom instruments, 31 onshore stations and 3 continuously operating MT stations (together with seismological and GPS sites) had been set up. The latter are expected to monitor the long-time variability of geomagnetic transfer functions, related to seismicity from the seismogenic zone as well as from faults in the forearc and in the volcanic arc (see Kapinos et al., this volume). In addition to the scientific questions, this contribution will also deal with the specific problems encountered in such an unique sea-land experiment.

Figure 2: Map of the investigated area with profile and area of station locations of the geophysical experiments considered here. The gravity modelling encompasses the entire region. Gravity after topex.ucsd.edu/ WWW_html/mar_grav.html; red triangles: volcanoes.

Final remark Having almost finished the onshore, shortterm geophysical experiment components of TIPTEQ, we can state that the field work went very well. The first and partial evaluations of the newly acquired data reveal an overall excellent quality. This success also goes back to the long-term preparation of the investigations, and especially a well established cooperation and infrastructure between German and Chilean institutions and personal collaborations, respectively. Reference Krawczyk, C. & the SPOC Team, 2003. Amphibious seismic survey images plate interface at 1960 Chile earthquake. EOS, Trans. Am. Geophys. Union, 84: 301, 304-305.

The southern TIPTEQ seismic network covering the Chilean forearc between 41.5° and 43.5° S - Status Lange D. (1), Rietbrock A. (2), Haberland Ch.* (1), Bataille K. (3), Hofmann S. (4), Dahm T. (4), Tilmann F. (5), and TIPTEQ Research Group (1) University of Potsdam, Institut für Geowissenschaften, Postfach 60 15 53, 14415 Potsdam, Germany, *E-Mail: haber@geo.uni-potsdam.de (2) University of Liverpool, Dept. of Earth & Ocean Sciences, 4 Brownlow Street, Liverpool L69 3GP, United Kingdom (3) Universidad de Concepción, Concepción, Chile (4) University of Hamburg, Institut für Geophysik, Bundesstraße 55, 20146 Hamburg, Germany (5) Bullard Laboratories, University of Cambridge, Madingley Road, Cambridge CB3 0EZ, United Kingdom

1. Introduction A network of 18 land stations and 20 offshore stations was installed on the island of Chiloe, the adjacent continental region around Chaiten, and the offshore forearc between 41.5° and 43.5° S. The network is part of the international and interdisciplinary research initiative TIPTEQ (From The Incoming Plate To megaThrust EarthQuake Processes) which is financed by the German Ministry for Education and Research (BMBF). The project aims to identify characteristics of and controlling factors for large earthquakes in the coupling zone of convergent margins and their interrelation with surface deformation. Furthermore, the project aims to study the structural variability of the oceanic Nazca plate and the forearc. This shall be achieved by the application of reflection and refraction seismics, heat flow measurements, multibeam bathymetry, and magnetic field data (see also Grevemeyer et al., this issue) as well as by analysis of earthquake recordings. The observation of (micro-) earthquake activity will be used to study the influence of the age-dependent thermal structure on the geometry of the seismogenic zone, and will allow the construction of a structural image of the chilean forearc at this latitude.

2. Experiment status In November 2004 a network of 18 seismic land stations was installed at the chilean forearc between 41.5° S and 43.5° S (Figure 1). Additionally, 20 ocean bottom hydro-phones/ seis-mometers (OBH/OBS) were installed at the end of February 2005. The whole network is scheduled to operate until October 2005. All land stations are equipped with digital data loggers running in a continuous mode at 100 Hz sample rate. The stations are equipped with a battery and two solar panels thus being independent from the public power supply. In cycles of approximately three months the stations are visited, checked, and the data recovered. The network is installed and maintained by the University of Potsdam, the University of Hamburg (Germany), the Universidad de Concepción (Chile), and the University of Liverpool (UK). The average station spacing is about 50 km. The network is therefore well suited to observe the whole range of earthquakes from local (micro-) earthquakes to teleseismic events. Depending on the amount of observed seismicity we are planning to locate the local seismicity and to derive structural models of the forearc by tomographic methods and receiver functions. The analysis of the stress field

Figure 1:Station distribution of the network covering the chilean forearc between 41.5° S and 43.5° S (on and around the island of Chiloé and the adjacent continental region). Diamonds (inverted triangles) show onshore (offshore) seismic stations, volcanoes are indicated by white triangles.

(moment tensor inversion) will contribute to the structure and dynamics of the forearc, of the oceanic plate (fracture zones, wedge), and of the seismic coupling zone. 3. First results Data quality control for the first four months shows that most of the stations were working correctly and have collected a total amount of ~60 GB. Although most of the stations are deployed close to farms, and despite the fact that unconsolidated sediments dominate the surface geology of the island of Chiloe, the data quality is acceptable. Already, several stronger local events in the offshore forearc have been recorded (Figure 2). Furthermore we observed a series of teleseismic events. We present the status of the ongoing study and first results.

Acknowledgements We gratefully acknowledge the cooperation of many Chilean landowners, companies, and institutions for support and for allowing us to install the seismic stations on their property. Furthermore we thank all field crews for their excellent work under difficult conditions. The project is financed by the German Ministry for Education and Research (BMBF) through the GEOTECHNOLOGIEN/KONTINENTRÄNDER program. Onshore instruments were provided by the Geophysical Instrument Pool (GIPP) of the GeoForschungsZentum Potsdam (GFZ).

Figure 2: Example of a local earthquake in direct vicinity of the westernmost onshore station (2005-01-20 15:11:31.39 GMT, 42째42.87' S, 74째15.32' W, ~25 km depth). Data (vertical component, bandpass filtered 2-20 Hz) are sorted according to epicentral distance.

Toxic Gas in the Namibian Coastal Upwelling Ecosystem (NAMIBGAS): An integrated study of H2S origin, abundance, and mechanisms of eruptions in a large coastal upwelling environment Lass H.U. (1), Siegel H. (1), Endler R. (1), Br端chert V. (2), Schiedek D. (1) (1) IOW, Seestr. 15, 18119 Rostock, Germany, E-Mail: lass@io-warnemuende.de (2) MPI of Marine Microbiology, Celsiusstrasse 1, 28359 Bremen, Germany

Introduction The Benguela upwelling system belongs to the four main major upwelling regions of the world ocean located at the eastern boundaries of the subtropical oceans. The south easterly trade winds drive perennial upwelling of nutrient enriched water into the euphotic zone which is the basis for the development of one of the most productive marine ecosystem of the world. Remineralisation of the produced organic matter in the water column and in the sediment consumes oxygen which makes low oxygen concentration of the subsurface water and hydrogen sulphide and methane in the sediment on the shelf a characteristic feature of the northern part of the system. Dynamical links of the Benguela system with the tropical and the sub-polar Atlantic provide an oxygen and nutrient flux preventing the system from starvation. Outbreaks of toxic H2S gas are a seasonally (austral summer) recurrent feature in the near-coastal shelf environment offshore Namibia. They have a significant economic and societal relevance because of their effects on biota (fisheries being the third largest source of revenue for Namibia). Until recently they were considered to be of local geographical extent and forced by a combination of high biological productivity in combination with reduced advection of oxygenated ocean water. New evidence from remote sensing and ship-borne

observations suggest a far larger geographical distribution than previously assumed and a significant contribution by eruptions of biogenic gas accumulations in unconsolidated organicrich diatomaceous oozes on the shelf. Objectives of the Project The central goal of NAMIBGAS is to elucidate the roles of oceanic and sedimentary gas sources, the conditions and mechanisms that lead to eruptions, and their relation to weather and oceanographic patterns in the coastal upwelling area and in the adjacent oceanic areas as well as the land. The methods to be applied by the project consortium encompass synoptic mapping of eruption patterns and frequency by remote sensing from satellites, ship-borne and in-situ monitoring of water mass structure and properties at times of eruptions; numerical and analytical modelling of circulation patterns and biomass production/dissimilation by a coupled physical and ecosystem model; acoustic sediment mapping by shallow seismic techniques, side-scan sonar, and visual mapping by remotely operated vehicles; analyses of sediment properties and of biogeochemical processes in the sediments; quantification of sedimentary gas accumulations and of microbial H2S turnover at the sediment-water interface; hydrology of alluvial fans discharging terrestrial runoff below the

Figure 1: Sea surface height (colour coded) and oxygen concentration in 300 m depth (contour lines) in the South Atlantic.

Figure 2: Longshore section of oxygen (hydrogen sulphide as negative oxygen concentration) in the upper panels and nitrate in the lower panel.

Figure 3: Total area and maximal offshore (zonal) extension of turquoise pattern in the coastal area of the Benguela system as a function of the season

marine sediments on the shelf. The concerted effort will result in an informed assessment of a major geohazard, contribute to the knowledge on the causes of natural variability in the coastal Benguela upwelling environment, and lay the ground for an improved sustainable use of natural fisheries resources in Namibia and the environmental risk assessment of mining activities on the shelf off Namibia. Present Status and Results Remote wind and sea surface topography measurements covering the whole South Atlantic provided evidence that the pool of subsurface water with low oxygen and high nutrient concentrations in the Angola gyre is maintained by open ocean upwelling, which mirrors in the depressed ocean surface shown in Figure 1, driven by the spatial structure of the wind field in the Gulf of Guinea. The variations of the coastal wind field at the northern boundary of the Benguela system control the intensity of the transport of tropical water into the northern part of the system at a seasonal and monthly time scale. The local long shore wind in the area of the Benguela drives the cross shore circulation which mixes the open ocean water with the shelf water. This mechanism is relatively weak between the Angola-Benguela front and Lüderitz and very intensive in the Lüderitz upwelling cell resul-

ting in a mud belt maintained by strong organic matter deposition at the shelf north of Lüderitz until Cape Cross. The release of hydrogen sulphide from the sediment into the overlying bottom water is controlled for the most part by the nitrate concentration of the bottom water, see Figure 2. Weak lateral nitrate bottom water flux associated with both reduced poleward undercurrent and cross shore circulation, observed frequently in austral summer and autumn, causes accumulation of hydrogen sulphide in huge volumes of bottom water. Subsequently this sulfidic water is transported by the cross shelf circulation towards the shore, wells up into the surface water and there the hydrogen sulphide becomes oxidised to elemental sulphur causing a turquoise coloration of the water at the coast. The spectral analysis of ocean colour observed by satellites revealed that these sulfidic waters were observed exclusively at the coast with a characteristic offshore scale of 1020 km and long shore scales in the order of 100 km, Figure 3. The simulations by the coupled circulation and ecosystem model suggest that the complicated spatial circulation pattern and its fluctuations in time distribute the sulphidic water over the whole northern Benguela with a focal point between 25°S and 24°S.

Figure 4. Turquoise water in the open ocean off Namibia identified as coccolithophores bloom.

Singular pattern of turquoise water off the coast turned out to be caused by blooms of coccolithophores, however with a slightly changed colour spectrum compared with the coastal sulphidic water, Figure 4. Coccolithophores bloom in waters which are depleted of silicate by preceding diatom blooms developing in fresh upwelled water close to the coast. Conclusions The properties of the low oxygen water in the Benguela system and its extreme appearance, the hydrogen sulphide eruptions, depend on both local and remote forced processes. Both processes are controlled by variations of the wind field in the South Atlantic in space and time on different scales. While the remote forced processes depend on the large scale variations of the wind field the local forced processes depend on small scale properties. The biogeochemical processes of the Benguela system occurring in the water column and in the sediment must be considered as a strongly interacting unit which are coupled via the oxygen and nitrate concentration of the bottom water to the local and remotely forced water mass transport.

Acknowledgements We greatly acknowledge the qualified support of the Captains, officers and crews of R/V Meteor , R/V A.v.Humboldt and R/V Welwitschia during the field work in the Benguela system. Colleagues of the NatMIRC, Namibia, generously supported the ship based measurements and provided local measured data. QuikScat data and images are produced by Remote Sensing Systems and sponsored by the NASA Ocean Vector Winds Science Team.

Active deformation and coastal tectonic geomorphology along the south-central Chile margin Melnick D. (1), Adam J. (2), Anderssohn J. (1), Bolte J. (1), Bookhagen B. (3, 6), Echtler H. (1), Kaufmann C. (1), Klotz J. (1), Kudrass H. (4), Moreno M. (5), Niedermann S. (1), Oncken O. (1), Reichel T. (4), Segl K. (1), Strecker M. (6), Wiedicke-Hombach M. (4) (1) GeoForschungZentrum Potsdam, Germany, E-Mail: melnick@gfz-potsdam.de (2) Dalhousie University, Canada (3) University of California, Santa Barbara, U.S.A. (4) Bundesanstalt fuer Geowissenschaften und Rohstoffe (BGR), Hannover, Germany (5) Universidad de Concepción, Chile (6) Universitaet Potsdam, Germany

The south-central Chile margin is a first-order research area to study processes linking great (M≥8) subduction earthquakes, coastal deformation, surface processes, and margin-parallel seismotectonic segmentation. Large earthquakes in this region that have successively ruptured three main seismotectonic segments in the past: the Valparaiso, Concepción, and Valdivia segments. We focus on the Arauco region, located within the overlap between the Concepción and Valdivia segments. This area corresponds to a geophysical, geomorphic, and structural long-term anomalous part of the margin. We use geomorphic and structural analysis, stratigraphy, different dating techniques (optical stimulated luminescence, in-situ produced cosmogenic radionuclides, and radiocarbon), crustal seismicity, and industry offshore seismic reflection lines. Our results show that crustal-scale seismogenic faults bound the Arauco peninsula and control the evolution of small wavelength (~5-20 km) Quaternary coastal landforms in the Isla Mocha-Arauco-Concepción area. Furthermore, our field observations, structural analysis, space geodetic measurements, and interpretation of seismic reflection lines reveal ongoing shortening since the Late Pliocene and that the active extensional faults observed at the surface represent complex systems of hinge grabens

reflecting bending-moment strain and local gravitational collapse associated to folding and uplift and do not represent the present-day tensional state of the crust. Combined studies relating onshore coseismic uplift recorded in sequences of emergent strandlines and offshore coseismic seismoturbidite from cores optained in mid-slope terraces are tentatively integrated to derive a paleoseismic record of megathrust earthquakes during the Late Holocene. Sand-box analogue modelling shows that basal accretion of underplated trench sediments is a plausible mechanism to explain the long wavelength (~50-100 km) upwrapping features of the forearc region. Material eroded from the Main Cordillera – the area of highest relief where the active volcanic arc is located – during the Quaternary glaciations conforms the 2- to 3-km-thick turbiditic sequence that fills the trench. As suggested by sandbox modelling, a spontaneous increase in sediment flux can profoundly change the mechanics and accretionary mode which is reflected in non-steady state surface processes and geomorphic evolution, emphasizing the importance of climatic-driven processes in the long-term seismotectonic segmentation of the margin.

Mechanisms linking surface processes, megathrust earthquakes and the tectonic geomorphology of the Arauco region, Chile Melnick D. (1,2), Bookhagen B. (2), Echtler H. (1), Strecker M.R. (2) (1) GeoForschungZentrum Potsdam, Telegrafenberg, 14473 Potsdam, Germany, E-Mail: melnick@gfz-potsdam.de (2) Institut fuer Geowissenschaften, Universität Potsdam, 14415 Potsdam, Germany

1. Introduction The south-central coast of Chile is affected by large subduction earthquakes that have successively ruptured three main seismotectonic segments in the past: the Valparaiso, Concepción, and Valdivia segments (e.g., Lomnitz, 1970). We focus on the area of the Arauco Peninsula, located within the overlap between the Concepción and Valdivia segments. This area corresponds to a geophysical, geomorphic, and structural anomaly in terms of gravity anomalies, Coastal Cordillera topography and with of the on-shore forearc and coseismic as well as long-term uplift rates 2. Objectives of the Project Detailed mapping of coeval seismogenic deposits and establishment of their precise chronology. Definition of boundary conditions influencing magnitude and extent of secondary effects through analysis of topographic, morphologic and lithologic characteristics of tectonically active landscapes. Delivery from parameters for the analogue modelling, location of active faults, deformation rates, seismotectonic segmentation of the margin, and characterization of their boundary zones. 3. Methods and Results 3.1 Mocha-Arauco-Concepción area We use geomorphic, stratigraphic, and structural analysis; optical stimulated luminescence, in-situ produced cosmogenic radionuclides, and radiocarbon dating; crustal seismicity and industry offshore seismic reflection lines provi-

ded by ENAP, Chile. Our data shows that crustal-scale seismogenic faults bound the Arauco peninsula and control the evolution of Late Quaternary coastal landforms in the Isla Mocha-Arauco-Concepción area. Furthermore, our field observations and interpretation of seismic reflection lines reveals syntectonic sedimentation documenting ongoing shortening since the Late Pliocene and that the frequently observed extensional faults at the surface represent complex systems of hinge grabens reflecting bending-moment strain and local gravitational collapse associated with folding and uplift and are not representative of the tensional state of the crust. Surface deformation and crustal faulting The highest coseismic uplift during the last two M>8 earthquakes (Darwin, 1851; Plafker and Savage, 1970) occurred in the areas where the highest late Quaternary uplift rates have been reported, at Isla Mocha 7 ± 2 m/ka (from recalibrated data in Nelson and Manley, 1992), and Isla Santa Maria 2.5 ± 0.5 m/ka (Melnick et al., 2004). Moreover, our mapping of offshore seismic lines and field observations shows that reverse faults control the uplift and geomorphological evolution of these islands and of the elongated peninsulas in the area (Melnick et al., 2004). These observations demonstrate that megathrust earthquakes can trigger faults in the upper plate, which in turn exert a control on the distribution of surface deformation during and between such events. Therefore, these faults must be taken into account when

inverting geodetic data into slip on the plate interface thrust. Such upper plate faults may contribute to the inconsistence between slip values obtained for the 1960 event (Barrientos and Ward, 1990) and the slip budget expected from the plate convergence rate and historic earthquake recurrence (Stein et al., 1986). 3.2 Case study - Santa María Island Here we use geomorphic, stratigraphic, and structural analysis combined with radiocarbon dating. The island comprises two units: Holocene lowlands formed by a flight of up to 25 emerged strandlines between the present sea level and 18 m elevation, coseismically raised during subduction earthquakes; and an upper surface formed by Tertiary rocks unconformably overlain by the Santa Maria Formation (Melnick et al., 2004). This unit is formed by 53to 31-ka coastal and 31- to ~27-ka-old eolian deposits with paleosol horizons at elevations of 15 to 58 m, which were dated using radiocarbon (see Figure 1). The transition from marine to continental depositional environments in the Pleistocene unit indicates complete emergence at 31 ± 1.8 ka BP, when sea-level was 82 ± 12 m lower than at present, obtaining a maximum uplift rate of 4.0 ± 0.2 m/ka and mean rate of 2.5 ± 0.5 m/ka (Melnick et al., 2004). The differences between the topography of the base of the near-shore unit, eolian unit, and present-day surface indicate progressive eastward tilting of ~0.1°/ka and synkinematic sedimentation. These surfaces are asymmetric, in the north they are oriented NE and dip SE, while in the south NW/NE. The Holocene strandlines yield a maximum uplift rate of 3 m/kyr, assuming they are preserved since the end of sealevel rise ~6 ka ago, which is confirmed by OSL ages ranging between 2.0 and 3.6 ka BP. Longitudinal profiles of three ephemeral streams, made with a laser theodolite, on the southern part of the upper surface show that knick-points are aligned along NW-oriented axes (Echtler et al., 2004). The latter parallels the tilting axes determined from the present-day surface topography analyzed in a photogrammetric digital elevation model (5-m resolution) and the contour lines of the Tertiary/Pleistocene unconformity.

Local network seismicity from the ISSA 2000 data set (Bohm et al., 2002) clusters 7 km northeast of the island. Across an E-W transect the earthquakes define a continuous W-dipping zone from 2 km depth to the plate interface at 18 km. Focal mechanisms are compatible with a NNE-striking, steep W-dipping reverse fault (Bruhn, 2003). An E-W oriented seismic reflection line across the cluster reveals Wdipping blind reverse faults, fault-propagation folding, and a piggy-back basin documenting eastward propagation of the deformation. This section accounts for ~8% of ongoing shortening since the Pliocene. The structure has a ramp-flat-ramp geometry, probably rooted at the plate interface. Seismic lines SE of the island show N-vergent folding above a WNWoriented reverse fault. Both the WNW- and NNE-oriented fault/fold systems converge at the center of the island. We therefore conclude that these two convergent systems control the asymmetry of the island. The progressive asymmetric tilting deduced from the base of the Pleistocene sequence, Holocene strandlines, and present-day surface can thus be attributed to lateral propagation of two fold systems that converge at the island. The island is cut by three fault zone, which are at the southern, center, and northern parts. The faults have extensional kinematics, as do most of the Quaternary faults in the Concepción-Arauco region. Most of these faults are relatively small features with slip magnitudes in the order of tens-of-meter slip to 150 m at the Morguilla fault. However, our interpretation of seismic reflection lines shows that small normal faults form above growing folds which in turn are propagated by high-angle reverse faults. The normal faults form small graben and hemigraben structures, which we interpret as hinge grabens, accommodating bending-moment strain cause by folding and bending and probably also gravitational collapse due to the lack of supporting frontal mass in the proximity of the continental slope. These observations emphasize that the surface structures do not reflect the stress field of the crust and

Figure 1: (1) Air photo of Holocene emergent strandlines and optically stimulated luminescence ages. The strandlines evidence successive Late Holocene meter-scale coseismic uplift during megathrust earthquakes. The thick lines shows the cliff formed at the end of sea-level rise ~6 ka. (2) Holocene debris flows in the hanging-wall of an active normal fault and calibrated radiocarbon ages. (3) General stratigraphic section of the Santa María Formation and calibrated radiocarbon ages (Melnick et al., 2004). (4.1) Photogrammetric digital elevation model (5-m resolution); (4.2) Contours of the base of the Santa María Formation. Coincident tilting axes indicated; profile along the western coast showing the longitudinal asymmetry of the island and normal faulting. (5) Uplift rates based on radiocarbon ages and the sealevel curve of Siddall et al., (2003), normal faulting causes the differences in rate.

are a local response to complex upper crustal structures which can have a completely different kinematics. 3.3 Seismotectonic segmentation Detail geologic mapping of the Pliocene and Quaternary units (1:50.000 scale) between 39 and 36S allows to define six morphotectonic blocks namely: Valdivia, Budi, Coi-Coi, Arauco, Concepción, Itata. The segmentation of these blocks is based on stratigraphic, structural, and

geomorphic elements. In situ produced cosmogenic radionuclide dating of a fluvial surface overlaying near-shore marine and deltaic sediments in the Arauco block yielded ages of 26Al, 126.0 ± 3.4 ka, and 10Be, 121.5 ± 2.9 ka, from which we calculate an uplift rate of 1.2 ± 0.1 m/ka. Further samples currently in process will constrain the uplift rates of the six morphotectonic blocks. These blocks are limited by faults. Some of them, like the Morguilla fault in the south of the Arauco have a clear

Quaternary topographic expression with forming an up to 150-m-high scarp. In other areas, like on the Coi-Coi block, ductile shear zones in the Paleozoic basement with overprinted brittle deformation indicating reactivation bound the morphotectonic blocks. 3.4 Reactivation of inherited structures The Arauco Peninsula is an integral part of the Nahuelbuta range, which is cored by Paleozoic granites intruding Paleozoic metamorphic basement. NW-striking ductile shear zones from the late Paleozoic represent crustal-scale discontinuities bound the Arauco peninsula (Echtler et al., 2002). These shear zones are overprinted by brittle faulting thus, we infer that they have been reactivated during the Tertiary and Quaternary aiding forearc segmentation and localizing uplift over an area where underthrusting of sediments and related uplift may occur. Acknowledgments We thank ENAP, the Chilean national oil company for providing the seismic profiles and to C. Mpodozis and J.P. Radic (ENAP-Sipetrol) for fruitful discussions. References Barrientos, S., Ward, S. 1990. The 1960 Chile earthquake – Inversion for slip distribution from surface deformation. Geo. Jour. Int., 103, 589-598. Bohm, M., Lüth, S., Echtler, H., Asch, G., Bataille, K., Bruhn, C., Rietbrock, A., Wigger, P. 2002. The Southern Andes between 36° and 40°S latitude: seismicity and average seismic velocities. Tectonophysics, 356, 275-289.

Echtler, H., Bookhagen, B., Melnick, D., and Strecker, M.R., 2004, Quaternary Tectonic Tilting Governed by Rupture Segments Controls Surface Morphology and Drainage Evolution along the South-Central Coast of Chile, Eos. Trans. AGU, 85(47), Fall Meet. Suppl., Abstract T13C-1389. Lomnitz, C. 1970. Major earthquakes and tsunamis in Chile during the period 1535 to 1955. Geol. Rund., 59, 938-960. Melnick, D., Bohm, M., Bookhagen, B., Echtler, H., Krawcyk, C, Manzanares, A., Moreno, M., Strecker, M. 2004. Active Faulting, Surface Deformation and Subduction Earthquakes at Isla Santa Maria, South-Central Chile. Eos Trans. AGU, 85(47), Fall Meet. Suppl., Abstract T13C-1380. Nelson, R., Manley, W. 1992. Holocene coseismic and aseismic uplift of the Isla Mocha, south-central Chile. Quat. Int., 15/16, 61-76. Siddall, M., E. J. Rohling, A. Almogi-Labin, Ch. Hemleben, D. Meischner, I. Schmelzer, D. A. Smeed. 2003. Sea-level fluctuations during the last glacial cycle, Nature, 423, 853-858. Stein, S., Engeln, J.E., DeMets, C., Gordon, R.G., Woods, D., Lundgren, P., Argus, D., Stein, C., Wiens, D.A. 1986. The Nazca-South America convergence rate and the recurrence of the great 1960 Chilean earthquake, Geophy. Res. Lett, 13, 713-716. Plafker, G., Savage, J.C. 1970. Mechanism of the Chilean earthquake of May 21 and 22, 1960. Geol. Soc. Am. Bull., 81, 1001-1030.

Darwin, C.1851.Geological observations on South America. Smith, Elder and Co., London. 768 p. Echtler, H., Glodny, J., Gräfe, K., Rosenau, M., Melnick, D., Seifert, W. y Vietor, T. 2003. Active tectonics controlled by inherited structures in the long-term stationary and non-plateau South-Central Andes. EGU/AGU Joint Assembly, EAE03-A-10902.

High-resolution, Three-component Reflection Seismic Survey in the Southern Central Chilean Andes at 38° S: First Data from Project TIPTEQ Micksch U. (1), Krawczyk C.M. (1), Stiller M. (1), Araneda M. (2), Bataille K. (3), Bribach J. (1), Buske S. (4), Groß K. (4), Lüth S. (4), Mechie J. (1), Schulze A. (1), Shapiro S. (4), Wigger P. (4), Ziegenhagen T. (1) (1) GeoForschungsZentrum Potsdam, Telegrafenberg. 14473 Potsdam, Germany, E-Mail: micksch@gfz-potsdam.de (2) SEGMI, Santiago, Chile (3) Universidad de Concepción, Chile (4) Freie Universitaet Berlin, Germany

Convergent continental margins are the Earth's principal locus of important earthquake hazards. Some 90% of global seismicity and nearly all interplate megathrust earthquakes with magnitudes >8 occur in the seismogenic coupling zone between the converging plates. At the southern Chilean convergent margin the largest instrumentally recorded earthquake occurred in 1960 (Mw = 9.5). It ruptured the margin starting at 38°S at a hypocentral depth of some 30 km below the continental forearc towards the south for approximately 1000 km with a coseismic slip of up to 40 m, up to 2 m vertical displacement and a tsunami up to 15 m high that affected the entire Pacific. The TIPTEQ project (from The Incoming Plate to mega-Thrust EarthQuake processes) studies the processes which generate these large earthquakes. One of the main task is to identify the key properties and processes in the seismogenic zone related to large subduction earthquakes, e.g. to image a complete seismogenic plate interface at the resolution of the expected scale of the associated processes from the updip to below the downdip end and that yield key petrophysical and mechanical properties which may be linked to mechanical and fluid-assisted processes and asperities in the seismogenic zone.

Amongst 13 sub-projects within the TIPTEQ project, the reflection seismic sub-project aims at the imaging and identification of processes in the seismogenic coupling zone of the present state of the ruptured plate interface at the southern Central Chilean margin (Fig. 1). Together with the marine SPOC data, the newly acquired high-resolution 3-component reflection seismic land data will yield a reflection seismic section that will cover the entire seismogenic coupling zone from its up-dip to its down-dip end (see also Krawczyk and the SPOC Team 2003). In January 2005, a c. 95 km long reflection seismic profile was shot in southern Central Chile at c. 38° S (Fig 1). The E-W trending line runs across part of the Central Valley (starting west of Victoria) and continued over the costal cordillera towards the Pacific, thereby passing the relocated hypocenter of the 1960 Valdivia earthquake. 180 three-component geophones were deployed along an 18 km long spread, moving 4.5 km in a daily-roll along for three weeks. Explosive shots, with a spacing of 1.5 km, allowed an up to 8-fold CDP coverage for the NVR part (Near Vertical Reflection) of the experiment (Fig. 2).

Figure 1: Position map of the 3-component reflection seismic experiment (red line).

Figure 2: Acquisition scheme for the Near-Vertical Reflection recordings. Profile km 0 is located at the Pacific coast, km 115 in the Longitudinal Valley.

In addition, an expanding spread experiment component focuses on the down-dip limit (3050 km depth). S-wave source signals will be generated and S-waves obtained with 3-component recordings to yield an improved picture of the petrophysical contrasts within the subduction zone system. Here, we will present the first reflection seismic data and preliminary results from the January 2005 experiment which should deliver a highresolution image of the seismogenic coupling zone between the subducting Nazca Plate and the South American continent. The seismic data show distinct reflections and broad reflectivity bands, e.g. between 3-4 s and 6-7 s TWT at the coast, which are dipping towards the cordillera. The overall data quality can be described as good. The reflectivity pattern shows similarities to the two-fold SPOC 2001 profile and should provide, therefore, a more detailed image of the subsurface. The downgoing plate is clearly visible at c. 8 s TWT near the coast reaching 17 s TWT at the eastern end of the profile. References Krawczyk, C. and the SPOC Team, 2003. Amphibious seismic survey images plate interface at 1960 Chile earthquake. EOS Trans. Am. Geophys. Union, 84: 301, 304-305. L端th, S., Mechie, J., Wigger, P., Flueh, E.R., Krawczyk, C.M., Reichert, C., Stiller, M., Vera, E. & SPOC Research Group, 2003. Subduction Processes Off Chile (SPOC) - Results from the amphibious wide-angle seismic experiment across the Chilean subduction zone. Geophysical Research Abstracts, 4.

Turbiditic Sequences in Sediment Cores from the Continental Margin of Southern Chile as Archive of past Seismic Activity Reichel T. (1), Wiedicke M. (1) (1) Bundesanstalt fuer Geowissenschaften und Rohstoffe (BGR), Stilleweg 2, 30655 Hannover, Germany, E-Mail: [ Thomas.Reichel | Michael.Wiedicke ] @bgr.de, Phone: +49 / 511 / 643-2799, Fax: +49 / 511 / 643-3663

Introduction Sediment cores from RV Sonne Cruise SO1615 have been investigated to study the relationship between local seismic activity and mass movements as documented in the marine sedimentary record. This cruise was part of the research campaign SPOC (Subduction Processes Off Chile). We present data from sedimentological, geochemical and geochronological examinations on several sediment cores to clarify this relationship. The tectonically active continental margin of Chile offers an excellent opportunity to study the context between seismicity and turbiditic deposits. The historical earthquake record of the past 500 years shows that even strong seismic events of magnitude 7 (Richter scale) have been fairly common. Seismic events may lead to remobilisation of sediments triggering downslope mass movements. Several deeply incised canyons channel most of the sediment to the trench, where it is being deposited; the trench sequence, thus, should be capable to provide a regional record of events. Some intra-slope basins also bear a high potential for the preservation of a seismically triggered turbidite record which is of a more local significance. Methods Several sediment cores were taken from different topographic settings along the continental margin and were analyzed for their succession of turbiditic layers (Fig.1, Fig.2). Core SL 50 taken at the trench represents the depositional regime influenced by canalized canyons

and gives evidence for the long-time regional record of sedimentation. Cores SL 49, SL 112, SL 39 and SL 66 represent the situation of different intraslope basins from the middle to the lower slope, which are characterized by local sedimentation processes. Colour reflectance spectroscopy and X-ray analyses of sediment slices revealed a high resolution record of the internal textures and layering of each sediment core and visualized even those turbiditic layers which were difficult to detect macroscopically. Age dating is based on AMS14C on planktonic foraminiferal tests and 210Pb analyses on bulk sediment samples. Results During the Pleistocene sedimentation rates were high and turbidite layers more frequent due to strong erosion in the hinterland associated with glacial activity and a colder and more humid climate. Even sediments in topographically elevated and distant locations within the trench (SL 50) contain numerous turbiditic layers, although age determination shows that these layers are older than 10 ka (Fig. 2). The most interesting results gives sediment core SL 112 from a deep intra-slope basin at the foot of the slope in our working area ToltĂŠn (Fig. 1). Calibrated AMS14C-ages on planktonic foraminifera revealed extremely high sedimentation rates of more then 8 m/ka. The sediment core contains several turbidite layers which were deposited during the past 700 years. Initial correlation of this turbidite

Figure 1: 3D-bathymetric map of the working Area Tolten between 38째- 39째S and 74째-76째W with all locations of the sediment cores investigated in this area. Data was produced during SONNE-cruise 161-3, 4 and 5.

Figure 2: Lithology of sediment core SL 50 and its geochronological frameset. More then 30 turbidites are contained in this core. The results of the color reflectance studies which represent the lithogenic content and the chlorophyll content within the sediment of the top 2m are given on the right side. Shaded areas show turbidites of Pleistocene age. X-ray photographs are printed in the middle.

Figure 3: Historical Earthquake record of Southern Chile and its possible correlation with turbidite depositions in sediment core SL 112. Further AMS14C-ages are in progress and we expect them to underline certain interrelationships of past strong seismic events and triggered turbidites.

sequence with the central Chilean historical record of mega earthquakes reveals a reasonable match. Additional age determinations of core SL 112 are presently in preparation, and we expect to be able to properly link the sediment record with past strong seismic events of the last 500 years (Lomnitz, 1969). 210Pb datings of the youngest turbidite of SL 112 (>200 BP) bear the potential for correlating this mass-wasting event with either the 1960(Mw>9)- or 1835(Mw ~8,3)- event off Valdivia/Concepcion. We think that these findings will prove the validity of our general approach to use the sediment record as an archive for paleoseismicity.

Conclusion Our AMS 14C ages and the preliminary results of the 210Pb datings suggest an average recurrence rate of ~100 a for turbidite deposits in sediment core SL 112. This result correlates well with the recurrence interval for large earthquakes in this region as presented by Nishenko (1985) or Lomnitz (1970) for southern Chile (Fig. 3). We, therefore, conclude that the turbidites of this sequence should be labelled seismoturbidtes. References Lomnitz, C., 1970 Major Earthquakes and tsunamis in Chile during the period 1535 to 1953, Geologische Rundschau, 59, 938-960. Nishenko, S.P., 1985, Seismic potential for large and great inerplate earthquakes along the chilean and southern peruvian margins of South America: A quantitative reappraisal: Journal of Geophysical Research, 90, 3589-3615.

High risk volcanism at the active margin of the SUNDAARC Reichert Chr. (1), Lühr B.-G. (2) (1) Bundesanstalt fuer Geowissenschaften und Rohstoffe (BGR), Stilleweg 2, 30655 Hannover, Germany, E-Mail: christian.reichert@bgr.de (2) GeoForschungsZentrum Potsdam, Telegrafenberg, 14473 Potsdam, Germany, E-Mail: ase@gfz-potsdam.de

Within the frame of the special initiative »GEOTECHNOLOGIES: continental margins« of the German Federal Ministry of Education and Research and the Deutsche Forschungsgemeinschaft the joint research project »SUNDAARC« was launched in January 2004, in cooperation with Indonesian science institutions. The aim is to study the various expressions of volcanism, its relations to the active subduction environment, and to install a real-time multi-parameter (MP) monitoring station. Three different Indonesian volcanoes were seleced: Krakatau, Merapi and Kelut. The interdisciplinary project is subdivided into three subprojects: Krakatau Monitoring (KRAKMON) comprises real-time observations of the volcanic activity parameter like micro-seismicity, of electromagnetic, gas-chemical and thermal parameters as well as of the deformation of the volcanic edifice and meteorological influences. Video monitoring, a sea level gauge and a data center are integrated. Correlation of the different data sets to another in relation to volcanic activity phases is expected to give insight into the controlling mechanisms and to enable the distinction between fluid-mechanical and fragmentation processes. All preparations and the procurement of the numerous equipment, including testing of its parts could be successfully terminated. The installation of the complete system (MP monitoring station and data center) will be realised between May and July 2005. This includes a

test phase and the launch into continuous operation. Merapi Amphibious Experiment (MERAMEX) aims at tomographic studies (vp and Qp) on the Merapi conduit system and its surroundings by active and passive seismic surveys simultaneously along with nearly 120 land stations and 14 Ocean Bottom Sensors providing high-resolution 3-dimensional models from the surface down to the plate interface in order to image the complete pathways of fluids and melts. The enormous field work with its coincident onshore and offshore activities was successfully accomplished in 2004 comprising two cruises of Research Vessel SONNE. The landstations recorded data sets of 150 days duration providing about 7 suitable local seismic events per day. Data evaluation is under way. Initial inspection shows that there are strong changes in subsurface parameters at the location of the volcano arc stretching down to the subduction plane. First velocity models have been derived from the active seismology data. Development of Highly Explosive Volcanoes at Active Continental Margins (DEVACOM) will analyse samples of erupted material from selected volcanoes under in-situ eruption conditions in order to mineralogically and petrophysically model the active processes. The reliability of the models will be examined by comparison of different volcano types. Visual video monitoring will provide direct

correlation to the data sets observed at the Krakatau MP station. The required samples from all target volcanoes including Augustine, Bezymanny, and Colima have been collected until spring 2005. Laboratory experiments will start soon. Besides a great number of Indonesian partner institutions, the German project partners are: BGR, GFZ, IfM-GEOMAR, University of Kiel and University of Munich. The joint interpretation and overall results shall contribute to a better understanding of the processes in the volcano interior and shall lead to a better hazard assessment in the study area including the establishment of improved early warning systems. This aspect gained particular importance after the disastrous flood catastrophe on boxing day 2004. It was internationally agreed that volcano early warning will be included in a global system which will be launched, soon.

Acknowledgements The funding of all three sub-projects by BMBF is greatly acknowledged.

Krakatau Monitoring (KRAKMON) Reichert Chr. (1), Klinge K. (1), Faber E. (1), Ibs-von Seht M. (1), Hoffmann-Rothe A. (1), Kniess R. (1) (1) Bundesanstalt fuer Geowissenschaften und Rohstoffe (BGR), Stilleweg 2, 30655 Hannover, Germany, E-Mail: christian.reichert@bgr.de

Introduction Active volcanoes are complex dynamic systems, in which interactions of different geological, geophysical and geochemical processes regularly can lead to eruptions that endanger human lives and infrastructures. The disastrous eruption of Krakatau in 1883, e.g., took approximately 36,000 lives. In the KRAKMON project, representing an interdisciplinary approach within the combination project SUNDAARC, new methods for the observation, analysis and interpretation of multi parameter data shall be developed and combined to a consistent model of the geological and dynamic processes inside the Krakatau volcano (Indonesia). This will be based on the experiences and results with the multi parameter monitoring station (BGR) at the volcano Galeras (Colombia) and the previous monitoring project at the volcano Merapi (Indonesia).

The results from this project will enhance indepth physical understanding of the dynamic processes inside volcanoes as well as the development of the measurement techniques and interpretation methods for activity monitoring and eruption prognosis. This will lead to a decrease of volcanic damage hazards, globally through improvement of methods as well as in respect of increased safety in the Krakatau surroundings. In particular for this region, hazard warning has recently gained an increased importance meaning due to the giant flood catastrophy of boxing day, 2004. As stated also in the official declarations of the UNESCOIOC special meetings in Paris (March 2005) and Mauritius (April 2005) volcano early warning systems will be incorporated in the planned Global Multi-Hazard Warning System. The Krakatau Monitoring System will be part of this system.

The KRAKMON project, which is scheduled for a period of three and a half years, includes:

The Indonesian partner institutions are: Directorate of Geology and Mineral Resources (DJGSM), Directorate of Volcanology and Geological Hazard Mitigation (DVGHM), Bureau of Meteorology and Geophysics (BMG), Indonesian Agency of Sciences (LIPI), Agency for the Assessment and Application of Technology (BPPT), Institute of Technology Bandung (ITB), Marine Geological Institute (MGI), Center for Volcano Research and Technology Development (BPPTK), Geodetic Survey (BAKOSURTANAL), Dept. Mines & Energy.

Installation and operation of a multi parameter station on the Krakatau complex, Detection and analysis of the volcano-induced multi parameter signals of the Krakatau, Interdisciplinary synthesis and source modelling of the volcano signals (DEVACOM) and comparison with Merapi (MERAMEX), Quantification of the volcanic activity of the Krakatau by definition of an activity chart. After a project period of three years, the monitoring of the Krakatau shall be continued by the Indonesian partners as a permanent task on the base of the obtained experiences.

Objectives of the Project In the first phase of the project a multi parameter (MP) station with digital telemetry to a data and interpretation center in the Krakatau Observatory at Pasauran will be installed on

Figure 1: Location of the Multi-parameter Volcano Monitoring Components and the telemetry stations on the Krakatau Complex.

the Krakatau complex. This station consists of different sensors recording several geophysical and geochemical parameters of the volcanic behaviour simultaneously and continuously. Among the observed phenomena are the volcanic seismicity (volcanic tremor and the flow noise of the magma-gas fluid called tremor), the chemical composition as well as the thermodynamic parameters and chemistry of fumarolic gases, variations in outer shape and the electromagnetic fields within the volcano building. From the temporal variations of these parameters conclusions can be drawn about the activity state of the volcano and about the dynamic processes inside the volcano. The continuous recordings of the MP station are supplemented by regular measurements of the temperature distribution in the crater area with an infrared camera, video monitoring, a weather station and a sea level gauge. Interpretation and signal analysis The interpretation and signal analysis tasks of the MP data comprise detection, identification and classification of volcano-induced signals in the data streams of the MP station as well as the parameterisation of the signals. Furthermore, the correlation of the respective signal parameters and the activity of the Krakatau is determined. The signatures and parameters of the MP signals and their distribution and correlation functions are the most important initial data for modelling the dynamic processes controlling the activity inside the volcano. Modelling The empiric correlation between the statistical characteristics of the MP signals and the visual activity as well as an identification of precursor signals derived thereof must be observed over many eruption cycles in order to be used as a significant base for eruption prediction. This time interval can be shortened considerably by physical modelling of the internal sources and stimulation processes of the volcano signals. In the Galeras project different methods for the modelling of sources as hydraulic resonators resp. oscillators have been developed which can be used as a first approximation for an

interpretation of the observations at the Krakatau. A subtask of this modelling is the clarification of elementary transport mechanisms for mass, impulse and mechanical respectively thermal energies and the transition mechanisms into an eruptive state. Monitoring The most important result of the monitoring of volcanoes is a quantification of the volcanic activity on a basis of multi parameter signals and their physical modelling through the definition of an improved multi-stage activity chart with a prognosis of the eruption probability for the upper activity levels, based on the fourlevel Indonesian system. For the Krakatau, this system is anticipated to be realised within a relatively short period because of its permanent activity and because of the installation of the MP station in the near field of processes controlling the activity in the upper part of the volcano building. Interdisciplinarity The two main targets of the project, modelling of the volcano signals and of the dynamic process within the volcano controlling its activity, as well as the quantification of the activity through an activity chart can only be realised by interdisciplinary collaboration of different physical and geoscientific subjects. Besides the physical subjects fluid mechanic, thermodynamics and numeric simulation from the field of geosciences three projects respectively areas have to be incorporated: Within the combined SUNDAARC project, close cooperation with the sub-project The development of highly explosive volcanoes on active continental margins (DEVACOM) is organized. In this project the elementary processes of the magma fragmentation as well as the parameters of the produced pyroclastics for different eruption mechanisms are simulated in laboratory under in-situ conditions or observed on site. A correlation of the laboratory results with the signals of the multi parameter station for different eruption patterns are important ancillary conditions for the physical

modelling of the dynamic processes in the Krakatau. The changes in the monitoring parameters will be compared with the properties of ejected material produced in an appropriate temporal context in order to detect and analyse existing correlations. Furthermore, the acoustic signal properties of depth-dependent micro-tremors will be compared with the parameters specified in the laboratory for the fragmentation processes of the ejected material (depending on P and T). The variations of the signal properties of the magma samples during the laboratory experiments depending on viscous flow to brittle fragmentation play an important part as well as the determination of the influence of (sea)water on the volcanicmagmatic processes. Attempts will be made to separate these processes from those of fluiddynamic processes. The results make for an improved definition and modelling of the involved physical/chemical processes, which in turn lead to more reliable parameters from the MP monitor signals for the improvement of early warning. The multi parameter station can be regarded as a window in the uppermost part of the volcano building. The processes in this domain close to the surface, which largely control the activities observed at the surface, are the final link of a long chain of transport mechanisms and reactions, which reach down to the deep earthâ&#x20AC;&#x2122;s crust and into the upper earthâ&#x20AC;&#x2122;s mantle. In order to understand, on the one hand, the relationship of the Krakatau with the geotectonic structure of Sunda Arc and on the other hand the injection of magma into the lower part of the volcanic root, the local seismic stations on the Krakatau complex are complemented by a net of regional stations on the islands of Java and Sumatra (set up and maintained by BMG, Indonesia). The seismotectonic methods and experiences from the Merapi project of GFZ shall be used for mapping the tectonic earthquakes in the Sunda Strait.

The physical reasons for this behaviour are not so much due to the different geometries and dimensions of the magma-producing chambers and channels inside the volcanoes, but more due to the 2-phase properties of the magma-gas fluids. The ratio sound speed/flow speed which is mostly responsible for stability and patterns of flows can assume extreme variations in the 2-phase magma-gas system compared to relatively small changes in the concentration of free gases. For this reason, a comparing examination of different volcanoes is a central issue of physical volcanology. For the project at the Krakatau, especially the results of Galeras project by BGR and of the Merapi project by GFZ can be used as references, as in both projects multi parameter methods are used.

Present Status and Results Within the first year detailed planning and comprehensive preparations were carried out, including an on-site inspection of Krakatau islands and surroundings in June 2004, where appropriate locations for the sensor stations and the telemetry could be determined. The numerous parts of the equipment were procured and intensely tested. The installation of the complete system (MP monitoring station and data center) will be realised between May and July 2005. This includes a test phase and the launch into continuous operation. The configuration of the MP monitoring station is presented in Figure 1. Figure 2 gives an overview on the wider area along with the location of temporary seismic stations. Insofar, no scientific results have been obtained, yet.

Acknowledgements The funding of this project by BMBF grant no. 03G0578A is greatly acknowledged.

In contrast to earthquakes, volcanoes show distinct individual patterns of eruption mechanisms and of volcano- and precursor-signals.

Figure 2: Location of the Krakatau Complex in the Sunda Strait. Red triangles indicate locations of partly temporary seismic stations on the islands, and on Java and Sumatra.

References Faber, E., Moran, C., Poggenburg, J., Garzon, G., Teschner, M. (2003): Gas Monitoring at the Galeras Volcano (Colombia), J. Volc. Geotherm. Res., 125, 13-23. Seidl, D., M. Hellweg, D. G贸mez, R. Torres, B. Pauly, S. Pauly und H. Rademacher (1997): What's better than a Broadband Seismic Station? A New System at Galeras Volcano, Colombia. EOS, v.78, no. 46, Supp., p. F441. Seidl, D., M. Hellweg, D.M. G贸mez und R.A. Torres (1998): Are tornillos caused by free vibrations of gas cavities? EOS, v.79, no. 45, Supp., p. F620. Seidl, D., Hellweg, M.: Parametrization of Multichromatic Tornillo Signals Observed at Galeras Volcano (Colombia), J. Volc. Geotherm. Res., 125, 171-189 Seidl, D., Hellweg, M., Calvache, M., Gomez, D., Ortega, A., Torres, R., B枚ker, F., Buttkus, B., Faber, E., Greinwald, S.: The MultiparameterStation at Galeras Volcano (Colombia) (2003): Concept and Realization, J. Volc. Geotherm. Res., 125, 1-12.

Sedimentary and geotechnical characterization of trench- and slope sediments off Southern Chile: preliminary results RĂśser G. (1), Behrmann J. (1), Kopf A. (2) (1) University of Freiburg, Albertstrasse 23B, 79104 Freiburg, Germany, E-Mail: georg.roeser@geologie.uni-freiburg.de (2) University of Bremen, Klagenfurter Strasse, GebĂ¤ude Geo, 28359 Bremen, Germany

1 Introduction PETROREC forms part of the TIPTEQ-Project and will gather a fundamental dataset on the petrogenesis, origin and physical properties of sediments that were deposited at the trench and the slope off the southern Chilean coast during the last 5 Ma and that are now being subducted beneath the Southern American Continent. This work is important for the understanding of the nature of seismogenic behavior of rocks in shallow areas of subduction. Material for the investigations comes from gravity cores collected from the whole TIPTEQ working area during R/V SONNE Cruise SO181, from earlier R/V SONNE Cruises SO102 (Hebbeln et al., 1995) and SO156 (Hebbeln et al., 2001) as well as from core material of ODP Leg 141 (Behrmann et al., 1992). 2 Geological setting The most prominent geologicalstructure in the working area is the Chile Triple Junction, defining the Nazca-, Antarctic and South American Plate kinematic setting. The recent spreading rate for the Chile Ridge is about 6 cm/year, the subduction rate for the Nazca Plate is 8-9 cm/year and 2-3 cm/year for the Antarctic Plate. Since the initial collision of the ridge with the Southern Andes trench 15 Ma ago, the Triple junction migrated about 1000 km to the north. The spreading ridge strikes almost parallel to the trench, resulting in highly oblique ridge-trench collision (Behrmann et al., 1992). On-land geology is not mapped in detail, yet reasonably well known at a regional level

(Servicio Nacional de Geologia Y Mineria, 1980). Areas in the vicinity of the Triple junction have been subject of field work (Forsythe and Nelson, 1985; Prior et al., 1990). The Area near the Triple Junction is dominated by four principal lithologies: 1. pre-Late Jurassic metamorphic rocks, forming the pre-Andean South American Basement, 2. the largely Mesozoic-aged Patagonian batholith, 3. Mesozoic and Cenozoic volcanic rocks associated with the Patagonian batholith, and 4. Neogene sedimentary and igneous rocks (Servicio Nacional de Geologia Y Mineria, 1980). Worth mentioning, although areally limited, rock types include an unusual suite of young (Pliocene-Pleistocene) granodioritic plutons in and around the Golfo Tres Montes within 20 km of the trench axis and about 150 km seaward of the main axis of the Quarternary Andean volcanic arc, and a tilted but apparently coherent Plioceneaged ophiolite sequence on the Taitao Peninsula (Forsythe et al., 1986). 3 Objectives The specific objectives of this part of TIPTEQ are the following: - To characterize the composition and the dynamics of the delivery area, in par-ticular with regard to the younger history of uplift and exhumation of the rocks in the overriding South American Plate;

Figure 1: Bathymetric map of the TIPTEQ working area with gravity core stations (from IFM Geomar Report Nr. 2, March 2005: FS Sonne Cruise Report SO181).

- To show, how mineralogical composition and microstructure of the trench and slope sediments affect mechanical behavior; - To obtain information about geotechnical properties of rocks that are to be subducted, in order to derive conclusions about the mechanics of subducted sediment packages. 4 Methods The first objective will be achieved by petrographical provenance analyses and fission track dating of relatively coarsegrained lithologies. More information on this part of the planned studies can be found in Heberer (this volume). To show how mineralogical composition and microstructure might affect the mechanical behavior of the sediment, the specimens will be investigated under the SEM using BSE and SE techniques. Special attention will be paid to micro chemical and structural analyses of grain cements. This will lead to a better understanding of the frictional and other strength parameters determined from geotechnical testing. Geotechnical testing forms the third part of PETROTEC. The experiments will simu-late the initial conditions of the subduction in the uppermost part of the dĂŠcollement, where the unimpaired sediments of the downgoing plate are deformed. The main questions in this context are: - Which parameters control the localization of movement plate boundaries? - What controls the onset of seismogenic behavior in the deeper part of these sediment bearing fracture zones? - Why is the coupling between the plates in different areas of the subduction zone so different? Are the controlling factors the low friction coefficients of some min-erals, high pore water pressures, or is it a combination of both? - Why do earthquakes with high magnitudes occur in convergent plate boundaries despite possibly very low resolved shear stresses? The geotechnical experiments will cover triaxial testing, direct shear testing and ring shear testing.

5 Present status and preliminary results 5.1 Sampling on SONNE Cruise SO 181 Leg 1a+b Most of the material for the geotechnical experiments as well as for the fission track dating and provenance analyses was sampled during R/V SONNE Cruise SO181 1a+b in December 2004 and January 2005. At a total of 18 stations 16 cores could be recovered successfully. The exact locations of all stations can be found in Figure 1. All cores were cut into 1 m segments immediately after recovery and stored in the refrigerated core storage aboard R/V SONNE at 4 Â°C. In addition, material from the core catcher was preserved in sample bags in the freezer (-32 Â°C) for shore-based analyses such as pore water chemistry (not part of PETROTEC). After the initial description aboard R/V SONNE, the recovered sediments can be roughly divided into three categories: 1. hemipelagic muds, 2. clayey to silty and sandy deep-sea trench deposits, and 3. deep-sea trench fan deposits, mostly silty or sandy. Sediment penetration depended largely on grain size, sediment strength and the topography of the seafloor. For example was it impossible to recover any material from the last two planned stations above the large normal fault scarp in the Chile Triple Junction area identified during ODP Leg 141 (Behrmann et al., 1994). The surface there was obviously too hard for gravity core penetration, and the steep topography might have been an additional problem. The longest cores of up to seven meters could be recovered in hemipelagic muds. As soon as silt and sand fractions were more abundant, penetration decreased to about 2.5 to 3.5 m. A total of about 70 m of sediment were recovered and shipped in a refrigerated container to Bremen. 5.2 Lithology, MSCT and XRF A thorough description of the lithology along with MSCT and XRF-scanning and a first sample session will take place in early May

Figure 2: Stress-Strain curve of a specimen from ODP Leg 141.

2005 at the University of Bremen. The data will be presented in the talk at Potsdam. 5.3 Geotechnical Experiments 5.3.1 Triaxial Tests Triaxial testing is performed at University of Freiburg, using a ELE »Digital Tritest 50« apparatus. Some preliminary tests were performed with material of SO156, SO102 and ODP141. The tests are performed after BS1377: part 8: 1990 Standard, consoli-dated and undrained. In addition to sediment strength properties, the permeabilities are determined on the basis of consolidation data. Fig. 2 shows the stressstrain curve of a specimen from ODP Leg 141, representative for most of the tests so far performed. The sample is from a depth of 19.3 m below seafloor. The range where Hooke’s law applies can be seen clearly, as well as the point of failure and the residual shear resistance. Saturation pressure was 750 kPa and consolidation pressure was 1500 kPa. The rate of axial displacement was 0,0248 mm/min and the calculated Young’s Modulus is 22.8 kPa. The Young’s Modulus’ of the tests so far performed lie between 17 kPa and 63 kPa.

The permeabilities in the performed tests are between 1.0953*10-8 and 8.7*10-10 m/s. The average permeability is x*10-9 m/s which corresponds to a marine clay to silt (Freeze & Cherry, 1979). The main part of Triaxial testing will be performed in the near future with material of R/V SONNE Cruise SO181. 5.3.2 Ring Shear Tests Theses tests are performed at the University of Bremen. In these tests, saturated sediments are sheared to high strains. This simulates large disalignment at faults. Because of the relatively low axial stress (<2MPa) on the specimens, only shear strengths and frictional behavior of areas at the toe of the accretionary wedge can be simulated. 5.3.3 Direct Shear Tests The Direct Shear Apparatus was developed by Achim Kopf at the SCRIPPS Oceanographic Institute (La Jolla, USA). The rate of shearing is variable over 6 or-ders of magnitude so that beside regular shear tests, so-called shearhold-shear tests will be performed. These tests simulate aseismic creep and slip alternatingly.

The variation of pore pressure as function of stress, state of consolidation and shear rate is recorded permanently. These tests will probably be performed in September 2005 at the SCRIPPS Oceano-graphic Institute in La Jolla, California. 6 References Behrmann, J.H., Lewis, S.D., Musgrave, R., Bangs, N., Bodén, p., Brown, K., Collombat, H., Didenko, A.N., Didyk, B.M., Froelich, P.N., Golovchenko, X., For-sythe, R., Kurnosov, V., Lindsay-Grifin, N., Marsaglia, K., Osozawa, S., Prior, D., Sawyer, D., Scholl, D., Spiegler, D., Strand, K., Takahashi, K., Torres, M., VegaFaundez, M., Vergara, H. & Waseda, A. (1992): »Chile Triple Junction«, Proc. ODP, Init. Rept. (Pt. A), 141, p1-708. Behrmann, J.H., Lewis, S.D., Cande, S. and ODP Leg 141 Scientific Party (1994): »Tectonics and geology of spreading ridge subduction at the Chile Triple Junction; a synthesis of results from Leg 141 of the Ocean Drilling Program.« Geol. Rundschau, 83: p832-852. Flüh, E. & Grevemeyer, I. (Editors) (2005): »FS Sonne Cruise Report SO181«. IFM Geomar Report Nr. 2 March 2005: p102. Forsythe, R.D. & Nelson, E. (1985): »Geological manifestations of ridge collision: evidence from the Golfo de Penas – Taito Basin«, Southern Chile. Tectonics, 4: p477-495. Forsythe, R.D., Nelson, E.P., Carr, M.J., Kaeding, M.E., Herve, M., Mpodozis, C., Soffia, J.M. & Harambour, S. (1986): Pliocene near trench magmatism in southern Chile: a possible manifestation of ridge collision. Geology, 14: p23-27. Freeze, R.A. & Cherry, J.A. (1979): »Groundwater« Prentice Hall, Englewood Cliffs; p.29. Hebbeln, D., Wefer, G. et al. (1995): »Cruise Report of R/V SONNE Cruise 102, Valparaiso – Valparaiso, 9.5.-28.6.1995.« Berichte, Fachbereich Geowissenschaften, Universität Bremen, 68: 126 pp.

Hebbeln, D and cruise Participants (2001): »PUCK: Report and preliminary results of R/V SONNE Cruise SO156, Valparaiso – Talcahuano, 29.3-14.5.2001.« Berichte, Fachbereich Geowissenschaften, Universität Bremen, 182: 195 pp. Heberer, B, Rahn, M., Behrmann, J. (this volume): »PETROTEC - How to decipher upper plate denudation by looking at fission tracks from the lower plate sediments – a concept for a study of the Southern Chile Trench.« Prior, D.J., Agar, S.M., Flint, S., Hervé, f., Murdie, R. & Styles, S. (1990): »Seismicity and neotectonics associated with subduction of an oceanic ridge transform system in southern Chile.« Terra Abstracts, 3: p378-379 Servicio Nacional de Geologia Y Mineria (1980): »Geological Map of Chile«. Inst. Investig. Geol., Inscr.52527, 6 sheets (scale 1:1,000,000).

Seafloor, Sediments, Seismicity and Shallow Structures Offshore Southern Chile – Selected Preliminary Results from TIPTEQ Cruise SO181 Scherwath M. (1), Grevemeyer I. (1), Flueh E.R. (1), Ranero C.R. (1), Kaul N. (2), Weinrebe W. (1), Contrares-Reyes E. (1), Tilmann F. (3), Gossler J. (4), and TIPTEQ Working Group (1) Leibniz-Institut fuer Meereswissenschaften, IFM-GEOMAR, Wischhofstr. 1-3, 24148 Kiel, Germany, E-Mail: mscherwath@ifm-geomar.de (2) Fachbereich Geowissenschaften, Universität Bremen, Klagenfurter Str., 28359 Bremen, Germany (3) Bullard Laboratories, University of Cambridge, Adingley Road, Cambridge CB3 0EZ, United Kingdom (4) K.U.M. Umwelt und Meerestechnik Kiel GmbH, Wischhofstr. 1-3, 24148 Kiel, Germany

1. Introduction Age-dependent structure and deformation of subduction zones is best studied in regimes of laterally heterogeneous oceanic plates near active spreading centre subduction. This is because here the subducting oceanic plate is youngest and hence hottest. Differences in the thermal structure are large between oceanic plate ages ranging from 0-20 Ma and almost absent in older plates (Kirby et al., 1996). A favourite locality for determining this age-dependence of structure and deformation is Southern Chile. Here, the oceanic Nazca plate, the oceanic Antarctic plate and the continental South American plate join at the Chile Triple Junction at 46.5°S. Within eight degrees to the north the age of the subducting Nazca plate changes from 0 to 25 Ma across several fracture zones (Tebbens et al., 1997). This segmentation enables data acquisition along corridors of distinct ages, i.e. thermal states, within a relatively small area. Furthermore, a simple plate motion environment exists with a constant convergence rate between the Nazca and the South American plate, constant spreading rates at the spreading centres, and a homogeneous plate motion vector parallel to the fracture zones and perpendicular to the trench, thus establishing a favourable natural laboratory for subduction zone process studies. In

100

addition, Southern Chile is the site of the world's largest historic earthquake, the Mw 9.5 event of 1960. The multi-disciplinary, multi-institutional project TIPTEQ (from The Incoming Plate to megaThrust EarthQuake processes) is investigating this age dependence in Southern Chile through various geophysical and geological techniques. Our contribution here focuses on first results from the RV SONNE cruise SO181 that took place from 6 December 2004 to 24 February 2005. Slab-age dependent effects on the tectonics and volcanism of the overriding continental plate are anomalously high regional forearc subsidence due to tectonic erosion (e.g. Behrmann et al., 1994) producing up to several kilometres of along-forearc subsidence north of the subducting ridge. This is driven by the steepness of the subduction thrust which, when older, cooler and thus denser, subducts at a steeper angle in the north. Here, the incoming plate also forms an bulge-like outer rise, whereas close to the margin the outer rise is small or absent (Cande and Leslie, 1986). Similarly, in older regimes the ocean depth and top of the oceanic basement lie at greater depths than in hotter, younger regions, and the sediments are thicker when the plate is older.

Figure 1: Basemap of TIPTEQ cruise SO181, indicating the major corridors of data acquisition, outer rise seismic networks (STA1, STA2), high resolution reflection seismic (SCS 01-04), and high resolution bathymetric coverage; land topography from TOPEX.

101

2. Objectives, Data and Techniques TIPTEQ principally aims at studying the relationship of the seismogenic zone, subduction zone processes, tectonic subduction erosion, continental accretion and arc magmatism. For the quantification of cause and effect, the required techniques offshore comprised high resolution imaging of lithospheric structure, predominantly from seismic velocity, seismicity, bathymetry, sedimentation, and thermal and resistivity distribution. In addition, newly collected magnetic data allow accurate age determination. This will be complemented by onshore-offshore data acquisition comprising structural imaging from seismic and magnetotelluric methods, ground probing, seismological investigations, and finally land GPS studies examine the coseismic, possibly segmented land deformation. To achieve these goals for the marine part, TIPTEQ acquired various complementary data sets along five major corridors. These data sets were seismic wide-angle and vertical incidence data, heat-flow measurements, multibeam bathymetry and magnetic data; the latter two data sets were also collected pervasively throughout the survey area. In addition, two short-term (six weeks) and two long-term (nine months) seismological networks were deployed to study microseismicity at the outer rise and regional seismicity across the shelf. Extensive bathymetric swath mapping was employed not only to produce high resolution images of the seafloor itself, but also to identify and trace outcropping subsurface faults, as could be shown for pervasive normal faulting at outer rises (Masson, 1991; von Huene et al., 1999, 2000). Previous cruises into the study area also collected multibeam bathymetry data (e.g. CTJ, Bourgois et al., 2000; SPOC, Kopp et al., 2004) which are supplementing our new data. High resolution seismic streamer data, collected along all TIPTEQ corridors, show sedimentary and basement structures, important to identify possible causes for the bathymetric features (such as faults), but also to correlate heat flow, seismicity, and deep imaging data

102

from seismic wide-angle data. These wideangle data illuminate the lithosphere to highlight large scale structures such as depths and shapes of the oceanic layers, in particular the depth and dip of the subducting slab. Estimated P- and S-wave velocities show anomalous, deformed regions and yield Poisson ratios. The velocity structure is also important to locate earthquakes from the seismological networks. Accurate localisations exhibit in particular how deep the seismicity occurs and whether outer rise faulting penetrates into the mantle.

3. Preliminary Results Initial data processing concentrated on basic data enhancement to obtain preliminary data quality assessments. Overall the data quality appears good to excellent. Many of the multibeam bathymetry data were cleaned from noise already during the cruise and are now merged with the existing data from previous cruises (Figure 1). The continental slope is now mapped completely, and large gaps on the oceanic plate between the extensive SPOC and CTJ cruises could be filled. The data clearly show spreading ridges where the sedimentary cover is thin, and normal faults from plate bending, the latter, though, largely in the north. It appears that the sedimentation overall is relatively high except around 41째S. This result is confirmed by the high resolution seismic reflection data. Focusing on TIPTEQ Corridors 2 and 3, where the outer rise seismological short term arrays were installed, both seismic sections in Figure 2 show roughly 2-km thick sediments at the trench, despite the difference in ages. The oceanic crust of Corridor 2 at the trench is 14.5 Ma old, and 6.5 Ma at Corridor 3, i.e. the thermal difference between these lines is relatively large. A greater difference between the lines exists in the structure; there clearly exists a bulge at the outer rise of Corridor 2, whereas the seafloor deepens only gradually towards the trench on Corridor 3 (Figure 2, disregarding the small scale spreading ridges here).

Figure 2: Vertical incidence seismic reflection lines from TIPTEQ Corridors 2 (line SCS01) and 3 (line SCS02). Distances are from Deformaion Front (DF).

The microseismicity is also different between the two corridors. During the six weeks deployment of the arrays, the northern array above the older crust registred almost 2000 earthquakes, which is roughly half the number of events registred in the southern network. Onboard analysis during cruise SO181 located 139 events in the northern network and 449 events in the southern network. Figure 3 shows the epicentres of these earthquakes. It appears that the northern network location was slightly outside the main microseismic activity. It is possible that more events would have been registred if the network would have been closer to the major zone of microseismicity. Alternatively, however, it may be that when the bulge has already formed as on Corridor 2, the microseismicity is naturally reduced, so the larger number of events in the younger, southern area are from the early stages of bulging. Further analysis is required to go beyond the hypotheses.

4. Conclusions New data from Southern Chile yield clear evidence of age-dependent lithospheric structure and deformation. Here, we summarise initial results of the shallow structures and outer rise microseismicity focused on two TIPTEQ corridors. Where the oceanic crust at the trench is 6.5 Ma old, bathymetric and highresolution seismic reflection data show a gently dipping ocean floor with no bulge at the outer rise, and microseismic activity is high in this region. At 14.5 Ma old oceanic crust an outer rise bulge exists and only half as many events as at the southern seismic network were registred here. However, many more data are available and await analsys, and excellent data quality promises excellent results to be expected from the TIPTEQ initiative.

103

Figure 3: Microseismic activity recorded at TIPTEQ outer rise networks.

104

Acknowledgments We gratefully acknowledge Captains Kull and Mallon, the officers and crew of RV SONNE for a highly successful data acquisition, Ms I.G. Arroyo for compiling the microseismic data base, and the Bundesministerium fuer Bildung, Wissenschaft, Forschung und Technologie (BMBF) for funding this project. References Behrmann, J.H., Lewis, S.D., Cande, S., and ODP Leg 141 Scientific Party, 1994. Tectonics and geology of spreading ridge subduction at the Chile Triple Junction; a synthesis of results from Leg 141 of the Ocean Drilling Program. Geol. Rundschau, 83, 832-852.

Masson, D.G., 1991. Fault patterns at outer trench walls. Mar. Geophys. Res., 13, 209-225. Tebbens, S.F., Cande, S.C., Kovacs, L., Parra, J.C., LaBrecque, J.L., and Vergara, H., 1997, The Chile ridge: A tectonic framework. J. Geophys. Res., 102, 12035-12059. von Huene, R., Weinrebe, W., Heeren, F., 1999. Subduction erosion along the north Chile margin. J. Geodynamics, 27, 345-358. von Huene, R., Ranero, C.R., Weinrebe, W., Hinz, K., 2000. Quaternary convergent margin tectonics of Costa Rica, segmentation of the Cocos Plate, and Central American volcanism. Tectonics, 19, 314-334.

Bourgois, J., Guivel, C., Lagabrielle, Y., Calmus, T., Boulegue, J., and Daux, V., 2000. Glacialinterglacial trench supply variation, spreadingridge subduction, and feedback controls on the Andean margin development at the Chile triple junction area (45-48째S). J. Geophys. Res., 105(B4), 8355-8386. Cande, S.C., and Leslie, R.B., 1986. Late Cenozoic tectonics of the Chile trench. J. Geophys. Res., 91, 471-496. Kirby, S. Engdahl, E.R., and Denlinger, R., 1996, Intermediate-depth intraslab earthquakes and arc volcanism as physical expressions of crustal and uppermost mantle metamorphism and subducting slabs. In: Subduction: Top to Bottom, eds. Bebout, G.E., Scholl, D., Kirby, S.H., and Platt, J.P., Geophys. Monograph 96, Am. Geophys. U. Kopp, H., Flueh, E.R., Papenberg, C., Klaeschen, D., and SPOC Scientists, 2004. Seismic investigation of the O'Higgens Seamount Group and Juan Fernandez Ridge: Aseismic ridge emplacement and lithosphere hydration, Tectonics 2(23), TC2009, 10.10029/2003TC001590.

105

Subduction zone fluid processes and their impact on the thermal structure and seismogenic zone Schilling F. R. (1), Kukowski N.* (1), Gottschalk M. (1), Tiwari R. (1), Ramelow J. (1), Knoll M. (1) (1) GeoForschungsZentrum Potsdam, 14473 Potsdam, Germany, *E-Mail: nina@gfz-potsdam.de

Of major importance within the TIPTEQ project I is to better understand the influence of fluids on the elastic, mechanical, and thermal properties of the constituents of a subduction zone and in particular its seismogenic zone. As fluid and heat transport processes are coupled in a complicated way with further fluid sources arising from metamorphcs reactions, only the use of different approaches is an adequate strategy to constrain the mechanisms and processes underlying fluid exchange and fluid pressure evolution: Thermodynamic modelling is employed to estimate the fluid and heat budget as a function of the age of the incoming plate and fluids released in different parts of a subduction zone from the slab and subducted sediments. All these parameters show significant fluctuations in the Chilean subduction zone. The approach and its results which serve as input and evaluation conditions for the numerical simulation, are described in more details below. To study in-situ the variation of elastic properties during the dehydration of subducting materials, laboratory experiments in a well defined model system (diaspore-corundum) with a high water release and a strong variation in elastic properties between hydrated and dehydrated mineral phases are undertaken. The goal of coupled numerical simulations is to model the transient thermal state as well as the magnitude and direction of fluid motion and estimate the impact of these variables on fluid pressure evolution which will assist to estimate earth quake conditions.

106

In this contribution we focus on the influence of fluids on the thermal regime within a subduction zone (Fig.1). Fluids entering a subduction zone with the oceanic crust and overlying sediment input will experience different fates: whereas some are expelled as early as with incorporation in the accretionary wedge, another portion will be subducted with the sediments and released within the subduction channel. There fluids also may be brought back to shallower depth without having been transported to greater depths. The portion of fluids is released below the seismogenic zone and may either undergo return flow or contribute to »steem cooking«, i.e. heat up the overlying continent. The remaining fluids may remain in the subducting slab and reach the deep mantle. It has been long recognized that water is an efficient heat carrier in porous media, due to its low viscosity and high heat capacity. However, the effect of fluids on the thermal regime in a subduction zone has been rather ignored in previous models. We show the potential of fluids as a controlling parameter for subduction zone processes, as they – depending on their amount and velocity of migration – have the high capability to significantly vary the thermal structure of a subduction zone. Our concept as presented here is based on the assumption of a thermal steady state. This has the advantage to model the distribution of heat flow density independently from special mechanisms and processes, a concept which is similar to assume equilibrium thermodynamics.

Figure 1: Possible fates of fluids entering a subduction zone. Part of the subducted fluid will transport heat from the mantle wedge through the magmatic arc and backarc. Outline of the Andean subduction zone modified after Scheuber and Gonzales 1999.

Steam Cooking Hypothesis A part of the high amount of fluids which is subducted with the down-going slab is released beneath the mantle wedge. The ascending water is heated when migrating through the hot mantle wedge. Due to its high heat capacity, water is an efficient heat carrier. The fluid transports heat into the crust, leading to an enhanced surface heat flow density. A small variation in the water budget can change »steam cooking« portion of heat flow from close to zero to more than 100 mW/m_. Therefore this parameter needs to be known for the coupled hydro-thermal simulations. The results of a parameter sensitivity analysis on the base of the “steam cooking” hypothesis (Fig.2) lead to a number of consequences and possible explanations: A change in fthe luid-budget may trigger the thermal and therefore seismic evolution of a subduction zone due to an additional heat flux. Advective heat transfer by fluids leads to increased heat fluxes and changing thermal properties in the seismogenic zone. Variations in the water budget may explain strong varia-

tions geological, petrologic, and geophysical parameters, as e.g. observed along the Andean subduction zone. Conclusion and Outlook The parameter sensitivity analysis shows that we need to take into account the thermal effects due to energy transfer through H2O to correctly address the thermal state of a subduction zone. Crucial is the size of fluid sources by metamorphic reactions as addressed in our laboratory experiments and fluids triggering the thermal budget of the mantle wedge. This will lead to an additional cooling of the mantle wedge, enhanced heating of the crust, cooling/heating due to latent heat during dehydration, hydration, fluid induced melting, or crystallization. Lastly, as large interplate earthquakes often nucleate at the downdip end of the seismogenic zone our studies will provide further, quantitative constraints to better understand the thermal and fluid pressure state of the seismogenic zone.

107

Figure 2: Additional heat flow density coming from »steam cooking« is estimated assuming a steady state as a function of the width of »Steam Cooking« and the subduction rate respectively. The larger square (red rectangle in Fig.1 ) represents a zone of 100 km width with a fluid portion of 20 %, a 100 Ma old plate which is subducted with 10 cm/a, and a temperature of the mantle wedge of 900 °C (conservative estimates). One of the above mentioned parameters is varied in each diagram for the purpose of a sensitivity analysis.

108

References Scheuber, E. & Gonzalez, G. (1999): Tectonics of the Jurassic-early Cretaceous magmatic arc of the north Chilean Coastal Cordillera (22째-26째S): a story of coupling and decoupling in the subduction zone. - Tectonics, 18 (5): 895-910. Schilling F.R., Kukowski N., Gottschalk M., Tiwari R., Ramelow J., Knoll M. (2004) Steam Cooking of the Andes, EOS transaction 85, V13A-1454, AGU Meeting San-Francisco (USA)

109

Author’s Index

A Adam J. . . . . . . . . . . . . . . . . . . . . . 72 Anderssohn J. . . . . . . . . . . . . . . . . . 72 Araneda M. . . . . . . . . . . . . . 24, 60, 78 Asch G. . . . . . . . . . . . . . . . . . . . . 6, 60 B Baes M. . . . . . . . . . . . . . . . . . . . . . . 10 Bataille K. . . . . . . . . 24, 26, 60, 64, 78 Behrmann J.. . . . . . . . . . . . . . . . 34, 94 Bohm M. . . . . . . . . . . . . . . . . . . . . . . 6 Bolte J.. . . . . . . . . . . . . . . . . . . . 72, 10 Bookhagen B. . . . . . . . . . . . . . . 72, 74 Brasse H. . . . . . . . . . . . . . . . . . . 54, 60 Bribach J.. . . . . . . . . . . . . . . 24, 60, 78 Brotopuspito K.S. . . . . . . . . . . . . . . . . 6 Brüchert V.. . . . . . . . . . . . . . . . . 12, 68 Buske S.. . . . . . . . . . . . . . . . 24, 60, 78 C Contrares-Reyes E. . . . . . . . . . . . . . 100 Currie B. . . . . . . . . . . . . . . . . . . . . . 12 D Dahm T.. . . . . . . . . . 18, 26, 50, 60, 64 Dingwell D.B.. . . . . . . . . . . . . . . . . . 44 Dübecke A. . . . . . . . . . . . . . . . . . . . 12 E Echtler H.. . . . . . . . . . . . . . . . . . 72, 74 Endler R. . . . . . . . . . . . . . . . . . . 12, 68 F Faber E. . . . . . . . . . . . . . . . . . . . . . . 88 Fauzi F.. . . . . . . . . . . . . . . . . . . . . . . . 6 Flüh E.R. . . . . . . . . . . . . 6, 18, 56, 100

110

G Galas R. . . . . . . . . . . . . . . . . . . . . . . 60 Gossler J. . . . . . . . . . . . . . . . . . . . . 100 Götze H.-J.. . . . . . . . . . . . . . . . . 30, 60 Gottschalk M. . . . . . . . . . . . . . . . . 106 Grevemeyer I. . . . . . . . . . . 18, 40, 100 Groß K. . . . . . . . . . . . . . . . . 24, 60, 78 H Haak V. . . . . . . . . . . . . . . . . . . . . . . 60 Haberland C. . . . . . . . . . 26, 50, 60, 64 Hackney R.. . . . . . . . . . . . . . . . . 30, 60 Hanka W.. . . . . . . . . . . . . . . . . . . . . 60 Heberer B. . . . . . . . . . . . . . . . . . . . . 34 Heesemann M. . . . . . . . . . . . . . 18, 40 Hess K.-U. . . . . . . . . . . . . . . . . . . . . 44 Hoffmann S.D. . . . . 18, 26, 50, 60, 64 Hoffmann-Rothe A. . . . . . . . . . . . . . 88 I Ibs-von Seht M. . . . . . . . . . . . . . . . . 88 J Julies E. . . . . . . . . . . . . . . . . . . . . . . 12 K Kapinos G.. . . . . . . . . . . . . . . . . 54, 60 Kaufmann C. . . . . . . . . . . . . . . . . . . 72 Kaul N. . . . . . . . . . . . . . . . . . . . . . 100 Kind R.. . . . . . . . . . . . . . . . . . . . . . . 60 Klinge K. . . . . . . . . . . . . . . . . . . . . . 88 Klotz J.. . . . . . . . . . . . . . . . . . . . 10, 72 Kniess R. . . . . . . . . . . . . . . . . . . . . . 88 Knoll M.. . . . . . . . . . . . . . . . . . . . . 106 Kopf A. . . . . . . . . . . . . . . . . . . . . . . 94 Kopp H. . . . . . . . . . . . . . . . . . . . . 6, 56 Krawczyk C.M. . . . . . . . . . . 24, 60, 78 Kudrass H. . . . . . . . . . . . . . . . . . . . . 72 Kukowski N. . . . . . . . . . . . . . . . . . 106

Author’s Index

L Lange D. . . . . . . . . . . . . 26, 50, 60, 64 Lass H.U. . . . . . . . . . . . . . . . . . . . . . 68 Lühr B.-G. . . . . . . . . . . . . . . . . . . 6, 86 Lüth S.. . . . . . . . . . . . . . . . . 24, 60, 78 M Mechie J. . . . . . . . . . . . . . . . 24, 60, 78 Melnick D. . . . . . . . . . . . . . . . . . 72, 74 MERAMEX Research Group . . . . . . . . 6 MERAMEX Scientists . . . . . . . . . . . . 56 Meyer U. . . . . . . . . . . . . . . . . . . 30, 60 Micksch U.. . . . . . . . . . . . . . 24, 60, 78 Moreno M. . . . . . . . . . . . . . . . . . . . 72 Müller S. . . . . . . . . . . . . . . . . . . . . . 44

S Scherbaum F. . . . . . . . . . . . . . . . 26, 60 Scherwath M. . . . . . . . . . . . . . 18, 100 Schiedek D. . . . . . . . . . . . . . . . . . . . 68 Schilling F.R. . . . . . . . . . . . . . . . . . . 106 Schulze A. . . . . . . . . . . . . . . 24, 60, 78 Segl K. . . . . . . . . . . . . . . . . . . . . . . . 72 Shapiro S. . . . . . . . . . . . . . . 24, 60, 78 Siegel H. . . . . . . . . . . . . . . . . . . . . . 68 Spieler O.. . . . . . . . . . . . . . . . . . . . . 44 Stiller M. . . . . . . . . . . . . . . . 24, 60, 78 Strecker M.R. . . . . . . . . . . . . . . . 72, 74

N Niedermann S. . . . . . . . . . . . . . . . . . 72

T Tilmann F. . . . . . . . . . . 18, 50, 64, 100 TIPTEQ Research Group . . . . . . . 26, 64 TIPTEQ Working Group . . . 18, 50, 100 Tiwari R. . . . . . . . . . . . . . . . . . . . . 106

O Oncken O. . . . . . . . . . . . . . . . . . . . . 72

V Villinger H.. . . . . . . . . . . . . . . . . 18, 40

P Peard K. . . . . . . . . . . . . . . . . . . . . . . 12 Puspito N.T. . . . . . . . . . . . . . . . . . . . . 6

W Wagner D. . . . . . . . . . . . . . . . . . . 6, 56 Weinrebe W. . . . . . . . . . . . . . . . . . 100 Wiedicke M. . . . . . . . . . . . . . . . 72, 82 Wigger P. . . . . . . . . . . . . . . . 24, 60, 78 Wittwer A.. . . . . . . . . . . . . . . . . . . . 56

R Rabbel W. . . . . . . . . . . . . . . . . . . 6, 56 Rahn M.. . . . . . . . . . . . . . . . . . . . . . 34 Ramelow J. . . . . . . . . . . . . . . . . . . 106 Ranero C.R. . . . . . . . . . . . . . . . 18, 100 Ratdomopurbo A. . . . . . . . . . . . . . . . 6 Reichel T. . . . . . . . . . . . . . . . . . . 72, 82 Reichert C.. . . . . . . . . . . . . . . . . 86, 88 Rietbrock A. . . . . . . 18, 26, 50, 60, 64 Ritter O.. . . . . . . . . . . . . . . . . . . 54, 60 Röser G.. . . . . . . . . . . . . . . . . . . . . . 94

Z Ziegenhagen T. . . . . . . . . . . . . . 24, 78 Zitzmann S. . . . . . . . . . . . . . . . . . . . 12

111

GEOTECHNOLOGIEN Science Report’s – Already published

No. 1 Gas Hydrates in the Geosystem – Status Seminar, GEOMAR Research Centre Kiel, 6-7 May 2002, Programme & Abstracts, 151 pages. No. 2

Information Systems in Earth Management – Kick-Off-Meeting, University of Hannover, 19 February 2003, Projects, 65 pages.

No. 3

Observation of the System Earth from Space – Status Seminar, BLVA Munich, 12-13 June 2003, Programme & Abstracts, 199 pages.

No. 4

Information Systems in Earth Management – Status Seminar, RWTH Aachen University, 23-24 March 2004, Programme & Abstracts, 100 pages.

112

Notes

Science Report GEOTECHNOLOGIEN

Continental Margins – Earth’s Focal Points ...

Continental Margins – Earth’s Focal Points of Usage and Hazard Potential

GEOTECHNOLOGIEN Science Report

Continental Margins – Earth’s Focal Points of Usage and Hazard Potential Status Seminar GeoForschungsZentrum (GFZ) Potsdam 9-10 June 2005

Programme & Abstracts

The GEOTECHNOLOGIEN programme is funded by the Federal Ministry for Education and Research (BMBF) and the German Research Foundation (DFG)

No. 5

ISSN: 1619-7399

No. 5