Observatories and Research Facilities for EUropean Seismology
Volume 4, no 1 May 2002 Orfeus Newsletter


The DESERT passive seismic experiment 2000/2001 in the Middle-East

M. Weber1, K. Abu-Ayyash2, A. Abueladas2, H. Al-Amoush1, A. Agnon3, Y. Bartov4, M. Baumann5, Z. Ben-Avraham6, G. Bock1, J. Bribach1, M. Daoud2, R. El-Kelani7, A. Förster1, H.J. Förster5, U. Frieslander8, Z. Garfunkel3, S. Grunewald1, H.J. Götze9, V. Haak1, Ch. Haberland1, A. Hofstetter8, M. Hassouneh2, A. Hördt10, K.-H. Jäckel1, D. Kesten1, R. Kind1, N. Maercklin1, J. Mechie1, A. Mohsen1, M. Neubauer10, R. Oberhänsli5, I. Qabbani2, O. Ritter1, G. Rümpker1, M. Rybakov8, T. Ryberg1, F. Scherbaum5, A. Schulze1, S. Sobolev1, M. Stiller1, J. Stoll11, H. Thoss1, U. Weckmann1, K. Wylegalla1.
1GeoForschungs-Zentrum, Potsdam, Germany
2Natural Resources Authority, Amman, Jordan
3Hebrew University, Jerusalem, Israel
4National Ministry of Infrastructure, Jerusalem, Israel
5University of Potsdam, Potsdam, Germany
6Tel Aviv University, Israel
7An-Najah National University, Nablus, the Palestine Territories
8Geophysical Institute of Israel, Lod, Israel
9Free University of Berlin, Germany
10University of Köln, Germany
11University of Göttingen, Germany.

Introduction - Description of the temporary seismic network - Data retrieval and archiving - Work in progress - Acknowledgements - References

Introduction

The Dead Sea Transform (DST) is a major plate boundary separating the African and Arabian plates. It extends over 1000 km from the Red Sea rift in the south to the Taurus collision zone in the north. Present-day left-lateral motion is 4±2 mm/year which is consistent with the kinematics of the Arabian plate assuming a rotation rate of about 0.4°/Ma around a pole at 31.1°N and 26.7°E relative to Africa (Klinger et al., 2000a). The DST became active about 18-21 Ma ago and since then, it has accommodated about 100 km of left-lateral slip (Garfunkel et al., 1981; Courtillot et al., 1987). In the area between the Dead Sea and Red Sea the DST is marked by the Arava fault (indicated by a dashed line in Figure 1) which may have the potential to produce Mw ~ 7 earthquakes along some of its segments about every 200 years (Klinger et al., 2000b).

The aim of the interdisciplinary and multi-scale Dead Sea Rift Transect (DESERT) project (DESERT Group, 2000) is to shed light on the question of how large shear zones work. DESERT consists of several geophysical sub-projects that are carried out by partners in Germany, Israel, Jordan and Palestine. Principal investigators are Michael Weber in Germany, Zvi Ben-Avraham in Israel, Khalil Abu-Ayyash in Jordan, and Radwan El-Kelani in the Palestine Territories. One of the sub-projects was a large-scale passive seismic experiment which was conducted in Israel, Jordan, and the territory of the Palestinian Authority. Aims of the project are (a) study of crust and mantle structure with the receiver function (RF) method, (b) travel-time tomography, (c) to investigate azimuthal anisotropy in crust and upper mantle from shear wave splitting, and (d) the study of local seismicity. In this note, we give a brief overview on the field experiment and the data archiving procedure.

Description of the temporary seismic network

The temporary seismic network consisted of 29 broadband and 30 short-period seismic stations, operated from the end of April 2000 when the first stations were installed in Jordan until the middle of June 2001 when the last stations were pulled out. The maximum number of operating stations was reached in November 2000. The DESERT seismic network crosses the Dead Sea Transform (DST) between the Dead Sea and the Red Sea (Figure 1). It has an aperture of about 250 km in NW-SE direction and approximately 150 km in SW-NE direction.


Figure 1. Station distribution of the DESERT passive seismic array.

The passive seismic experiment was organized by GeoForschungsZentrum Potsdam (GFZ), Germany. The data loggers and seismometers were provided by the GFZ Geophysical Instrument Pool. The following persons from GFZ participated in the actual fieldwork: Günter Bock, Rainer Kind, Ayman Mohsen, Georg Rümpker, Kurt Wylegalla. Other participants came from the Geophysical Institute of Israel, Lod, Israel (Rami Hofstetter); Natural Resources Authority Amman, Jordan (Abdel-Qader F. Amrat, Walid Abdel-Hafiz, Muhamed Hijazi, Bassam Al-Bis, and Khamis Rizik); An-Najah University, Nablus, Palestinian Authority (Radwan El-Kelani, Ayman Mohsen).

All seismometers were three-component. Mark L4-3D short-period sensors were used. Broad-band seismometers used in the experiment were 12 Guralp 40-T, 8 Guralp CMG-3T, and 9 Streckeisen STS-2. All stations were equipped with Reftek data loggers, and recording was continuous in compressed mode at 50 Hz sample frequency. Depending on the noise conditions, about 20-30 Mbyte of data were accumulated per station per day. The data were stored on disks whose capacity varied between 2 and 4 Gbyte. Service visits to the stations were carried out once every 3-4 months.

Seismic stations were powered by one or two 12V batteries of 60 Ah capacity each that were recharged by solar panels of 50-60 Watt capacity, or by an electrical charging unit in cases where 220V mains power was available at the site. A regulator was used to switch off the data logger if the voltage fell below 11.8V. This prevents drainage of batteries in cases where recharging of the batteries failed. The data logger switches on again automatically as soon as the battery voltage reaches 12.6V. For safety reasons stations were set up mainly at police stations, existing sites for the national seismograph networks, schools, government offices, and water reservoirs. The amount of vandalism and losses by theft after more than 1 year of field operation were consequently relatively minor. Unfortunately, as a result of the deteriorating political situation, the stations in Gaza and Hebron could not be maintained after August 2000.

Data retrieval and archiving

For station visits a total of 15 spare disks were available. The actual replacement of disks and checking of other components of the stations was completed within 15-20 minutes. After servicing the stations, the Reftek raw data were saved on magnetic tapes using a portable Linux system. We used the Passcal routine refdump in cases where the accumulated file size was below 2 Gbyte. For file sizes exceeding 2 Gbyte, we used the Unix file copying routine dd by appropriately choosing the starting and end records with the count and skip options. Two safety copies were prepared of each station.

Back in the labarotory at GFZ Potsdam, the data were read from the tapes, quality checked and converted to 24-hour day files in Miniseed format using the extr_file routine written by W. Hanka. The 24-hour files are stored as zipped tar files in the GEOFON data archive. Full Seed volumes can be extracted via breqfast requests from the archive.

At this time, the data are for the exclusive usage of the DESERT group. It is anticipated that access to the data will be open three years after the end of the field experiment, i.e. June 2004. Data requests can then be submitted via the GEOFON web page at GFZ Potsdam.

Work in progress

Work on receiver function analysis, travel-time tomography and local seismicity is in progress. Seismogram examples illustrating local seismicity and teleseismic receiver function analysis are shown in Figures 2 and 3. Preliminary results on azimuthal anisotropy can be summarized as follows (from an abstract submitted to the EGS April 2002 Nice meeting). We have analyzed the splitting of S, SKS and SKKS waves for both temporary stations and permanent broadband stations in the area. The results reveal consistent directions of the fast S wave velocity approximately parallel to the DST. Delay times between fast and slow split waves range from 1.0 s to 2.0 s and show a characteristic lateral variation probably related to the DST. Delay times are high, up to 2.0 s, over the DST itself, while they tend to be smaller (about 1.0 s) at greater distances from the DST. Our results are consistent with sub-horizontal asthenospheric flow parallel to the DST over the whole area investigated, and enhancement of seismic anisotropy in the sub-crustal lithosphere by olivine alignment resulting from shear deformation along the DST.


Figure 2. Record example of a local earthquake (ML = 2.2) on May 5, 2000, 21:41:36.1 UTC, in the Arava valley. Focal depth was 18 ± 2 km, i.e. at the base of the upper crust. Interpretation of P- and S-arrival in a Wadati diagram revealed a ratio of 1.74 for the VP/VS ratio.


Figure 3. Example of receiver function processing for the Sulawesi MW = 7.4 earthquake of May 4, 2000, 04:21 UTC origin time and recorded at a temporary broadband station of the DESERT passive seismological experiment located in Jordan at 89° epicentral distance and back azimuth 92°. The components are indicated on the left-hand side of the panel. Amplitudes of N, E, Q and T are enlarged threefold relative to Z or L. L, Q, T is a ray-based coordinate system with the P-wave mainly on the L-component, and SV and SH on the Q- and T-component, respectively. Traces a-c are the raw data; d-f bandpass-filtered in the frequency band 0.02-0.2 Hz, this step reduces the high-frequency parts of the signal; g-i traces after rotation of d-f to the LQT coordinate system; j-l LQT traces after deconvolution with the P wave of the L-component seismogram (trace g). The source-equalized receiver function is given by the Q-component seismogram trace k. Some phases in trace k have been marked: 1 = Ps phase from the Moho beneath the station; 2 = Ppps multiple between Moho and surface; 3 = Ppss multiple between Moho and surface. The strong negative phase directly arriving after 1 is probably a P-S conversion from a discontinuity in the mantle where velocity decreases with increasing depth.

Acknowledgements

We thank the Deutsche Forschungsgemeinschaft, the Minerva Foundation, and GFZ Potsdam for financial support.

References

  • DESERT Group. Multinational geoscientific research effort kicks off in the Middle East, Eos, Transactions, AGU, 81, No. 50, pages 609, 616-617, 2000.
  • Bock, G., A. Hofstetter, A. Mohsen and DESERT Group. Seismic Anisotropy beneath the Dead Sea Transfrom from Observations of Shear Wave Splitting, Abstract, EGS Meeting 21-26 April 2002, Nice, France.
  • Courtillot, V., R. Armijo and P. Tapponier. The Sinai triple junction revisited, Tectonophysics, 141, 151-168, 1987.
  • Garfunkel, Z., I. Zak and R. Freund. Active faulting in the Dead Sea Rift, Tectonophysics, 80, 1-26, 1981.
  • Klinger, Y., J.P. Avouac, N. Abou Karaki, L. Dorbath, D. Bourles and J.L. Reyss. Slip rate on the Dead Sea transform fault in northern Araba valley (Jordan), Geophys. J. Int., 142, 755-768, 2000a.
  • Klinger, Y., J.P. Avouac, L. Dorbath, N. Abou Karaki and N. Tisnerat. Seismic behaviour of the Dead Sea fault along Araba valley, Jordan, Geophys. J. Int., 142, 769-782, 2000b.


page 3
Copyright © 2002. Orfeus. All rights reserved.