Introduction

Geodynamic processes acting mainly in the Arabian and Red Seas result in deforming the Arabian plate in different ways. The most obvious is its NNE relative movement and the presence of regional geological structures, mainly faults and fault systems that run in the N, NE, NW and EW directions. Earthquake activity shows a clear concentration along the boundaries of this plate and its intra-regional fault systems. Recent, historical and pre-historic seismicity data indicate the occurrence of some destructive earthquakes in and around this plate with a noticeable correlation with its major tectonic elements (Barazangi, 1983; Adams and Barazangi, 1984; El-Isa et al., 1984; El-Isa and Mustafa, 1986; Al-Sinawi, 1986; El-Isa and Al-Shanti, 1989; Ambraseys et al., 1994).

The seismicity of Oman in particular has received little attention so far despite the fact that many recent and historic reports about felt earthquakes are available. Considering its location on the southeastern part of the Arabian plate and the geological, tectonic and the limited seismicity information, the presence of a relatively low-to-moderate seismic hazard in this country is quite evident. The rather fast development of Oman in the last two decades and the ambitious planning for large future projects urgently call for the implementation of a comprehensive program that accounts for this hazard.

The establishment of an earthquake monitoring center is therefore an immediate requirement towards the fulfillment of a national program for the assessment and mitigation of earthquake hazard in this country which was recommended by the Cabinet in the year 1995. The first of a multi-phase program was approved by the Cabinet in June 1999 through which ten stations were proposed to be installed in the Sultanate (El-Isa, 1999). Locating all earthquake sources in and around the country and the determination of their levels of activity and other characteristics in a quantitative manner represents the major objective of the earthquake monitoring center ultimately leading to the production of earthquake hazard maps. Such data and maps represent the foundations for planners and structural engineers in selecting the suitable sites and designs for all civil constructions to reduce their vulnerability.

The details of this network, station locations, system description, the central data acquisition system and some preliminary results are presented in this paper.

The Seismological Network

1. The major earthquake sources around Oman are represented by the regional faults of the plate boundary, see Figure 1. These occur at variable distances from the Omani shorelines in the ranges 100 km - 300 km in the Gulf of Oman region, 300 km - 500 km in the Arabian Sea region and at about 150 km or more, farther south. Other secondary earthquake sources are represented by the regional crustal faults that cross the country in definite directions, mainly NE, NW, and N-S. Some of these are known to be associated with the ophiolite occurrences in northern Oman. Though a detailed regional fault map for Oman does not exist, an attempt was made to compile a preliminary one, which was considered in distributing the proposed stations around the regional faults, particularly in the northern part of the country.
2. Most of the population and infrastructure in this country is concentrated in its northern parts while less density is found in the south and very disperse population occurs in its middle part.
3. Station accessibility, hard-rock foundation, low seismic noise level, smooth topography and other local environmental factors were also considered. Sites were selected to be as distant as possible from the shorelines to reduce the seismic noise of the sea waves. The closest to the Gulf of Oman, in the north is about 15 km distant. The southern stations were stationed more than 30 km from the shorelines of the open Arabian Sea.
4. The station distribution should show a reasonable coverage in the N-S and E-W directions, parallel and perpendicular to the plate boundaries in the Arabian Sea and the Gulf of Oman. The spacing between successive stations should typically be in the range of 60 km - 80 km.
5. The rather rigid topography of Oman, particularly in its northern parts and the large extension of the country, more than 1300 km in the N-S direction restricted the selection of communication links to data transmission via satellite at an early stage of the project.
6. In the second phase, some ten stations more will be added. These will be distributed in the northernmost part of the country as well as eastern and middle parts. Special attention should be paid to reduce the gap between the existing northern seven stations and those three in the south.

Table 1. Coordinates and elevation of installed stations.

Station Name	Code	Latitude [ N ]	Longitude [ E ]	Elev [ m ]	Sensor Type
Samad	SMD	23° 03.5333'	58° 02.9500'	1000	3 x SS-1
Jabal Madar	JMD	22° 22.2083'	58° 06.2083'	350	3 x SS-1
Wadi Bani Khalid	WBK	22° 36.5983'	58° 58.1983'	450	3 x SS-1
Araqi, Ibri	ARQ	23° 20.1967'	56° 31.3167'	400	3 x SS-1
Bidbid	BID	23° 31.2667'	58° 07.6150'	200	3 x SS-1
Bisya	BSY	22° 44.6733'	57° 11.9700'	450	3 x SS-1
Hoqain, Rustaq	HOQ	23° 34.9417'	57° 18.6533'	350	3 x SS-1
Wadi Al-Hawf	WHF	17° 55.1400'	53° 46.2350'	550	3 x SS-1
Rabkut	RBK	17° 30.2100'	54° 12.2317'	550	3 x SS-1
Aybut	ABT	17° 21.2267'	53° 17.7983'	650	3 x SS-1

At each station, a piece of land 30 m by 30 m was officially acquired. A 1-m square hole was dug in the rock down to the unexposed hard rock. The depths of these sensor holes varied between 1 m to 5 m. In each sensor hole, special precaution was taken to prevent filtration of moisture. The walls of each pit were built with concrete bricks and coated. At the bottom, a 50 cm thick concrete slab was built as sensor pier. A 2 m by 3 m room of 3 m height was built above the sensor hole. A fiberglass umbrella, 3m x 4m, was firmly constructed 1 m above the roof of the room, which provides excellent thermal insulation. Six radial trenches, 30 cm deep and 30 cm wide, were dug around the room with variable lengths ranging between 10 m and 15 m. These were filled with sand, char-coal and metal matrix in which a 10 mm thick copper cable was buried that extends to a brass rod installed on top of the fiberglass shade for lightning protection. Necessary precautions were also taken against vandalism including the construction of a 2-m high fence around each station, see Figure 3.

System Description

Remote Seismic Stations

Each of the remote seismic stations consists of three short-period Kinemetrics SS-1 sensors, a Quanterra datalogger (Q730BL) and the satellite communication interface (Hughes Network Systems' PES 5000). The datalogger converts the analog seismic signals to digital data at 100 sps. The data are time-stamped by the internal clock, which is phase-lock-looped to the time of a GPS receiver. The communication interface takes care to transfer the continuous IP data packets to the designated central data acquisition system. A reliable 900-W solar power system consisting of 12 Siemens SP-75 and one EXIDE 6-50A15 battery block, provides the RSS with the required power.

Central Data Acquisition System (CDAS)

The central data acquisition system is installed in the EMC inside the SQU campus. It mainly consists of a communication server and the Antelope Software Package. The communication server merges all incoming and outgoing data streams and forwards them to the Local Area Network (LAN) from which they are distributed to the appropriate workstation using a TCP/IP socket connection.

The EMNO uses the Antelope software version 4.3. It consists of two major subsystems, namely the Antelope Real-Time System (ARTS) and the Antelope Seismic Information System (ASIS). ARTS provides full functionality for the seismic network operations and control. This includes real-time data acquisition to non-volatile disk ring-buffer, interactive control of field equipment, system state of health monitoring, and real-time automated data processing including detection, seismic phase picking, event association, location and archiving. It also offers automated distribution of raw data and processed results. ASIS uses the relational database formalism and is based on the CSS v.3.0 schema for information organization. It is supplied with all required tools for the manual review and processing of the seismic data.

The main elements of the CDAS are: A primary data acquisition Sun workstation with 21" monitor (model: SPARC Ultra 10); one post-processing Sun workstation with 21" monitor (model: SPARC Ultra 10); and a DDS-4 tape-backup unit. The fourth element is a TrueTime network timeserver with GPS engine (model: NTS-90). An UPS (model: SmartUPS 2200XL) from APC with an external battery provides extended battery backup. The last element is the 9100 UMOD satellite modem with radio and dish antenna manufactured by Hughes Network Systems.

The UMOD connects to another UMOD in the hub of Omantel at their main station in the city of Al-Amerat. This satellite communication link could be viewed as a lease-line with a fixed bandwidth, currently set up at 72 kbps. This is enough for steady-state communication and can support additional bandwidth to acquire backed-up data when required. The network is completely based on TCP/IP communication where the hub of Omantel acts as a network bridge. This allows for the most reliable data acquisition and control communication back to the remote sites. Figure 4 shows these communication flows and Figure 5 shows the connection types between the different subsystems.

Performance and Preliminary Results

Seismological data are continuously received at the EMC in real-time. These are believed to be of good quality with mostly high signal-to-noise ratio, see Fig. (6). Some stations, however, are noisy during some hours of the day due to environmental causes. The data are automatically processed by Antelope modules for detection, phase picking, seismic event association, location, and ultimately archiving of both data and results in the relational database providing a comprehensive seismic information system. Earthquake parameters are tabulated and epicenters are plotted on local, regional and global maps. At a later stage, these data and results are revised. Additionally, the analyst is manually checking the time-series for very small events that did not yield a network trigger. Special attention is paid to all local earthquakes, which represent the main target of this short-period instrumented network at this stage.

Table 2. Distribution of local earthquakes in continental Oman and its vicinity as recorded on EMNO during its first 10 months of operation from August 2001 to May 2002.

Region	Number of Earthquakes	Largest Magnitude (Ml)	Number of Earthquakes with Ml >= 3
Continental Oman	9	2.1	0
Gulf of Oman	29	3.2	1
Arabian Sea	33	4.2	8
Hormuz Strait	17	5.1	8
Arabian Gulf Region	26	4.3	12
Gulf of Aden	7	5.1	6

Up to the end of May 2002, the network recorded 566 earthquakes, out of which some 147 are local with magnitudes in the range of Ml = 0.5 - 5.1, see Figure 7. Only nine of these occurred on-shore within Oman. Their magnitudes were less than or equal 2.1. A number of 112 earthquakes had their epicenters close to the shorelines of Oman and are distributed in the nearby regions as shown in Table 2. More than ten of these were felt in the northernmost parts of Oman and the nearby United Arab Emirates. All of them occurred in the northern part of the Gulf of Oman, Hormuz Strait region and the southern part of the Arabian Gulf.

Acknowledgements

The first author expresses his sincere thanks and appreciation to the University of Jordan, Amman for permitting him three years leave of absence, which resulted in the fulfillment of this vital project. This was followed by a sabbatical year to enable him continue the project and work on the "seismicity of Oman" and the assessment of its hazard.

References