The Mw 9.3 Indonesian Earthquake and the Holocene Great Earthquake Record from the Sumatran Subduction Zone Supported by the US National Science Foundation PI’s Chris Goldfinger, Joseph Stoner Indonesian Colleagues Dr. Yusuf Djadjadihardja, BPPT Dr. Danny Natawidjaja, LIPI, Dr. Haryadi Permana, LIPI Contact Information Oregon State University Marine Geology Active Tectonics Group College of Oceanic and Atmospheric Sciences Ocean Admin Bldg 104 Corvallis, Oregon, 97331 USA [email protected] voice: (541) 737-5214 fax: (541) 737-2064 http://www.activetectonics.coas.oregonstate.edu Scientific Summary The December 26th 2004 Mw 9.3 subduction earthquake off northern Sumatra was the second or third largest ever recorded. While we will learn a great deal by studying this event, quantifying the temporal and spatial patterns of earthquakes remains elusive because our observations often span only one, or several seismic cycles. In the case of northern Sumatra, only smaller events were known from the historical record in the 2004 rupture zone, and there was little memory of great events and associated tsunami from the past. Real-time strain measurements (GPS) can only represent a fraction of one strain cycle, and the historical record and instrumental seismology represent typically one or two cycles. Fundamental issues such as the seismic gap hypothesis, clustering, recurrence patterns, and applicability of slip or time-predicable earthquake models remain difficult to address because we rarely have a long enough time series to discriminate between models. Around the Indian Ocean, the lack of information is extreme, such that over 200,000 people died, partly because there was little or no expectation that an event like December 26th could occur, except among a very few in the scientific community. Even among the tectonic community, the northern Sumatran and Andaman-Nicobar segments of the Sunda Subduction Zone were considered poorly coupled, and thus some inferred, at low risk for great earthquakes. Paleoseismology has the potential to address these questions directly using a longer time span than available to geodesists and seismologists. Recent work in Cascadia and along the Northern San Andreas Fault has shown that developing a Holocene time series of earthquakes is possible, and yields invaluable information about recurrence intervals and patterns. The Cascadia turbidite record of great earthquakes, now 10,000 years long, has revealed clustering, and perhaps a repeating temporal pattern. The Sumatran/Indian margin now offers the opportunity to both address the science of this great earthquake, and provide critically needed information about the past behavior of this fault system. Will a long quiescent period follow this earthquake? No one can say, however there is there is evidence of both recent rupture on this segment, of pairs of earthquakes closely spaced in time in Indonesia, as well as in other subduction zones around the world, thus an assumption a long repeat time following a great earthquake is not well justified. This project will apply submarine paleoseismologic techniques to the turbidite history along the northern Indonesian/Indian convergent margin. The margin exhibits strong parallels to the Cascadia margin, where turbidite paleoseismology has proved successful. We propose a project to collect jumbo piston and Kasten cores along and across the trench axis to test, develop, and adapt paleoseismologic methods we have used in Cascadia and the SAF to the Sunda margin. Ultimately, we expect to construct a great earthquake record for this margin extending back 10,000 yrs or more. We will use AMS 14C dating to date correlated turbidites generated by great earthquakes. We will apply criteria developed in Cascadia and Japan to discriminate such events from those triggered by other mechanisms, if present. We now have learned that we can correlate earthquake generated turbidites between sites using magnetic and density signatures that fingerprint individual events. This method of regional correlation is working well in Cascadia and along the San Andreas, and greatly strengthens the stratigraphic framework that previously has depended mostly on radiocarbon ages.

A-1

We plan to establish a Holocene event chronology in the Sunda trench floor by sampling with longitudinal and along strike piston core transects, as well as abyssal plain channels if these exist. We will apply both correlation tests of synchronous triggering, as well as sedimentological criteria in a parallel approach to test for earthquake origin and long-term fault behavior. We will directly date planktonic forams using AMS radiocarbon, and will establish the chronostratigraphic record using Geotek magnetics and density, xray and optical imagery, U-channel magnetics, microfossil biostratigraphy, and oxygen isotopes. We will also use ash stratigraphy if we find correlatable horizons in our cores. This project is designed to compliment the extensive geologic and geophysical efforts currently underway or proposed by international colleagues. As it has in Cascadia, we expect that this project will result in a Holocene record of great earthquakes along the Sunda-Java margin. For probabilistic hazard assessment, such a time series will have a significant and immediate impact and improve hazard assessment for Indonesia, and tsunami hazard around the Indian Ocean. The data from this project will be publicly available via publications and our web site. We expect to include one full time graduate student, and 16 other graduate and undergraduate students who will participate in the sea-going phase of the project. The students will be both from the REU program, and from collaborating institutions in Indonesia and Japan. These students will receive training in paleoseismology, sedimentology earthquake geology and seafloor mapping. Cruise Plan We plan a cruise of thirty days to perform the proposed sampling program of the floor of the Sunda-Java Trench and selected lower slope basins. Prior to coring we will complete a multibeam survey of the lower slope and adjacent abyssal plain and will map channel systems and likely core sites. Considerable multibeam data has already been collected by Japanese, French, German, and UK vessels off the northern Sumatra margin, thus we will use these data to site our cores in that area, and collect new multibeam data as needed to augment the existing surveys, and add new data along southern Sumatra and western Java lower continental slopes. By the time this cruise takes place (March-May 2007), we anticipate that a minimum of 50% of the needed multibeam will have been collected and made available from other sources. We are planning for 12 days of survey, but the exact number needed, given the other expeditions in the area, is unknown. Multibeam surveys in water depths between 4500 and 6200m in the Indonesian EEZ will be done with a Simrad EM-12 multibeam system, a swath of bathymetry and backscatter system. We expect to collect ~70 new piston cores, 8-12 jumbo Kasten cores and ~6 box cores at ~ 18-24 sites. Site selection is based on collecting representative samples that monitor spatially separate segments of the margin, while avoiding, as much as possible, influences such as local slumping. Preliminary core sites have been chosen, though final site selection will take place at sea following analysis of survey data. The preliminary ship track and core site locations are shown in Figure 1. At sea we will run all cores through the OSU Geotech MST system, collecting gamma density, p-wave velocity, high res. imagery and point and loop magnetic susceptibility series for each core. Cores will be split and hand logged at sea, and imaged with the Geotek high-resolution line-scan camera system. All coring operations and multibeam surveys will be done along the Java trench, on the adjacent lower continental slope, or on the abyssal plain of the Indian plate in water depths ranging from ~ 4500 m off northern Sumatra, to ~ 6200m off western Java. Multibeam survey of selected areas on the lower continental slope up to water depths of 2000m may be required to define channel systems leading to the trench. At no time do we require any science operation within the Indonesian Territorial Sea. All operations will be within the Indonesian EEZ. The Indonesian Science Agencies LIPI and BPPT have offered staff support, equipment, and logistical support for this project, as well as scientific collaboration. We plan to embark Indonesian scientists and students for the cruise, and jointly pursue the science goals through close collaboration before, during, and after the cruise. The joint collaborative project will benefit Indonesia in terms of improved understanding of the earthquake and tsunami recurrent from Aceh to western Java. In addition, the lnowledhge transfer and technology transfer proposed here will allow similar work to be undertaken in the future with Indonesian agencies. BPPT has asked, and we have agreed to assist the agency in constructing heavy piston coring gear for the Indonesian vessels Baruna Jaya 1-1V. Our primary Indonesian collaborators are:

A-2

Dr. Yusuf Djadjadihardja

Head of Center of Assessment and Application on Natural Resources Inventory Technology , Agency for Assessment of Application of Technology (BPPT) Gedung II BPPT,Lt.19 Jl. MH.Thamrin No.8 Jakarta Pusat 10340 [email protected], [email protected] phone 62-21-316-9700 fax 62-21-316-9720 Dr. Ridwan Djamaluddin

Head of Technology Center for Marine Survey Agency for Assessment of Application of Technology (BPPT) Gedung II BPPT,Lt.19 Jl. MH.Thamrin No.8 Jakarta Pusat 10340 [email protected] phone 62-21-316-8800 fax 62-21-310-8849 Responsibilities of Investigators and Collaborative Statement C. Goldfinger, Associate Professor of Oceanography in the College of Oceanic and Atmospheric Sciences, will be the principal investigator and co-direct the proposed program. Goldfinger will be responsible for conducting the field work, the bathymetric analysis of canyon/channel systems, and the development of the paleoseismic record. Co Pi Joseph Stoner will conduct U-channel magnetic work and develop a radiocarbon supported paleomagnetic time line. We also are collaborating (at no cost to the proposal) with Takeshi Nakajima, a leading sedimentologist specializing in turbidite/fan systems and paleoseismology at AIST in Japan, Joel Johnson, and Kerry Sieh and his Tectonic Observatory lab at Caltech. Sieh has pioneered the record of onshore paleoseismology in Sumatra over the last decade, with unprecedented success. Sieh is actively engaged in mapping the uplift and subsidence signature of the recent earthquake on the ground in Aceh Province, and our labs are working collaboratively to expand this ground truth to a regional representation of vertical motion with satellite imagery. We are collaborating with Yusuf Djadjadihardja (BPPT) and Hariyadi Permana (LIPI). BPPT is our host in Indonesia, and OSU and BPPT have a Memorandum of Understanding (MOU) established with agreement on the issues of data sharing, cruise participation, technology transfer, and scientific collaboration and publication. We are also directly collaborating with the UK Survey team that has recently surveyed a portion of the rupture zone with multibeam sonar, and who has plans for further scientific work in 2008.

A-3

2007 preliminary Core Sites, Java-Sunda Trench ID X Y 7 92.94189374940 3.31602995972 8 92.94189374940 4.31586864635 9 93.04187761810 4.14922886524 10 92.80858192460 4.14922886524 11 93.40848513650 3.44934178460 12 93.37515718030 3.28270200350 13 93.57512491760 3.21604609106 14 94.07504426100 2.58281492286 15 94.30833995450 2.51615901042 16 94.17502812960 2.38284718553 17 95.44149046600 1.94958375466 18 95.27485068490 1.94958375466 19 95.34150659740 1.78294397355 20 96.27468937150 0.51648163715 21 98.10772696370 -2.61634624762 22 96.17470550290 1.24969667402 23 96.37467324020 1.14971280535 24 96.94124849600 0.14987411872 25 97.04123236460 -0.05009361860 26 97.54115170790 -0.98327639279 27 97.64113557660 -1.14991617389 28 97.44116783930 -1.18324413012 29 98.64097426320 -2.68300216006 30 98.80761404430 -2.84964194117 31 98.60764630700 -2.91629785361 32 99.37418930010 -3.44954515314 33 99.47417316880 -3.61618493425 34 99.34086134390 -3.64951289047 35 100.14073229300 -4.24941610245 36 100.27404411800 -4.48271179599 37 100.90727528600 -5.48255048262 38 100.94060324200 -5.64919026373 39 100.74063550500 -5.68251821995 40 101.87378601700 -6.94898055635 41 102.54034514100 -7.24893216234 42 102.80696879100 -7.41557194344 43 102.64032901000 -7.44889989966 44 103.64016769600 -7.78217946187 45 103.87346339000 -7.88216333054 46 103.70682360900 -7.94881924298 47 104.80664616400 -8.24877084897 48 105.07326981400 -8.28209880519 49 104.87330207700 -8.38208267385 50 105.40654937600 -8.98198588583 51 105.80648485100 -8.94865792961 52 106.00645258800 -9.08196975449 53 106.73966762500 -9.51523318536 54 107.00629127500 -9.68187296647 55 106.87297945000 -9.68187296647 56 107.63952244300 -9.84851274757 57 108.07278587400 -9.98182457246

A-4

A-5