Edelvays Spassov

Key words: seismic energy, composite fault plane solution, stress field

The purpose of this study is to make an estimate of the stress field in the Australian continent from composite fault plane solutions of the strongest events in the area. The spatial distribution of more than 400 earthquakes with magnitude greater than 4.5 since year 1902 suggests a linear trend, where the greatest number of events is in a relatively narrow belt, crossing the country in the middle from North-West to South-East. The largest amount of elastic energy is released within the same seismic belt.

Using data from ISC bulletins we have calculated composite fault plane solutions for six major earthquake zones in the country. An individual fault plane solution for at least one event in each zone has been computed, too. The space distribution of the focal mechanisms confirms a dominant East-West compression stress in the continent, with different tilt in different areas. The main compression axes are with clear North-South direction in the Tennant Creek area (Northern Australia). The general direction of the plate motion is North - Northeast.

Various numerical models of the stress field in Australia have been suggested (Cloetingh & Wortel, 1986; Coblentz et al, 1995; Zhang et al, 1996) and only in some restricted areas earthquake and in-situ measurements have been used for estimation of the main continental stresses (Denham et al, 1979; Denham et al, 1981; Lambeck et al, 1984, Everingham & Smith, 1979). The reasons for these limitations are the vast territory and the inadequate station coverage. This makes the study of the fault plane solution of the events with magnitude less than 4.0 throughout the country especially difficult. The catalogue of Australian earthquakes is most complete for events with magnitudes 4.5 and above. Normally, only earthquakes with magnitude greater than 5.0 are being recorded on international stations. Such records are needed for an optimum distance-azimuth coverage, to be used for a reliable fault-plane solution.

Although seismicity in Australia is a typical case of intraplate seismic activity, there are few events, which arguably have magnitude greater or equal to ML 7.0 in both South and Western Australia. Basically, every state in the country has relatively well defined seismically active zone with earthquake(s) with magnitude greater than 6.0. This fact gives reasonable ground to believe that at least one composite fault plane solution could be calculated for each zone. These composite focal mechanisms will give more robust and more general characteristic of the stress field than using individual event mechanisms. Obtaining such composite solutions for all know seismic zones will produce the most direct evidence for the main compression stresses in Australia, especially for areas without in-situ stress measurements.

ISC bulletins provide first P-wave arrival data in an easy to access format since about 1968. Starting from this year on we collected all ISC data on Australian events, where there is a reasonable amount of P-arrivals at different distances. These are normally data for events with magnitude greater than 5.0. Finally, a total of 21 Australian earthquakes have been included in the initial database - Table 1. Those events are distributed in groups of three to four per seismic zone. In two important areas - Queensland and South Australia (around city of Adelaide) there have not been recent earthquakes with greater magnitude, so we could not include these two zones in the present study.

For each composite solution first P-wave arrivals have been collected in the distance range between 1 and 95 degrees. Appropriate concentration of earthquakes has been found in five main zones: Southwest (Meckering), Northwest (south of Kimberley), Northern Territory (Tennant Creek area), Central Australia (Marryat Creek) and Southeast Australia (the region between Sydney and Melbourne. It is known from previous studies (Spassov et all, 1997) that in Southeast Australia there are at least three seismic zones with different seismicity and stress patterns. In one of them no strong earthquake has occurred in the last 40 years. We separated the data from southeastern Australia into Northern and Southern zones and performed the calculations separately for each zone. For calculating the fault plane solutions we used a program written by D. Suetsugu (1994), which offers an automatic and manual options for choosing the nodal planes. The program also allows for unlimited number of iteration on the same input data, until the best ratio between consistent/inconsistent data is achieved.

The first P-wave arrivals from three to four events within close distance range are put together and after a various number of iterations a single composite fault plane solutions has been calculated for each seismic zone. Also, for each of the zones, the event with the largest number of first arrivals (this is usually the strongest earthquake in the region) has separately been included in an individual fault plane solution calculation. This should help in estimating the influence of the single events and the robustness of the composite solution.

The distribution of more than 400 earthquakes with magnitude greater than 4.5 since 1902 is given in Figure 1. The time distribution of their number and the elastic energy released is shown in Figure 2. Most of the significant picks in this distribution are related to the strongest events in this time period. The spatial distribution of the number of events (first digits) and the seismic energy (second digits) in cells with size 5 by 5 geographical degrees is presented in Figure 3. An obvious belt with higher number/energy values could be traced crossing the continent in the middle from Northwest to Southeast. The only exception is the Southwest corner of the country, which is also the area with the two strongest known earthquakes in the instrumental history of seismicity in Australia. A median number of 1 to 3 events is typical for nearly half of the cells and could be assumed as a background level for events with ML greater than 4.5 within the continent. Similar average level of energy release per cell is between 0.1 and 10^20 Joules for more than half of the territory. These values distinguished one quarter of the territory to stand out with higher than average seismic activity and approximately one eight of the continental Australia have no activity above ML 4.5 level.

The individual fault plane solutions of two major events in Southeast Australia are shown in Figure 4. These are both thrust fault ruptures, but with distinctive difference of the strike orientation. The composite solution for the whole area (Figure 5 - left) does not correspond much to neither of the individual solutions. The composite solutions for events separated in zone North (Figure 5 - right/top) and South (Figure 5 - right/bottom) have better ratio between consistent/inconstant data and are closer to focal solutions for the individual events.

Figure 6 presents the individual solution (left) and composite solution (right) for earthquakes in Marryat Creek area (Central Australia). The similarity and consistency ratio of the data in both solutions suggest only slight variation of the focal mechanisms in this area. A very different orientation of the trust fault type is revealed in Tennant Creek Area (Northern Territory) - Figure 7. The very high number of the consistent data in the composite solution supports the North-South orientation of the main compression stress in this area. Although the sequence of strong events in this region is in a very narrow space/time window (seven days), the individual solutions still differ slightly from the more robust composite one. In the Northwest of the country (South of Kimberley area) however, the individual and composite solutions are very similar (Figure 8). Almost identical are the individual and the composite solutions for Southwest Australia (South of Perth) - Figure 9. Such a stability of the solutions suggests a steady and robust stress field distribution for the Western, geologicaly older half of the continent and less established stress patterns for the generally younger Eastern part of Australia.

The location of all composite fault plane solutions and the orientation of the main compressional stress are presented in Figure 10. The bigger arrow shows the general direction of the plate motion. It seems that only the compression stress in Tennant Creek area (Northern Territory) could be directly influenced from the friction of the plate moving North. The main stress field in the rest of the country is more likely to be a result of compression perpendicular to the plate margins, caused by the interaction with the neighbouring plates. Composite solutions for Queensland and South Australia would perhaps give an idea whether there is more than a 180 degree rotation of the main compression axes throughout the country or this orientation is steadily East-West with some slight variations on the edges of the continental Australia.

A median number of events and an average background value of seismic energy release have been established on the basis of more than 400 earthquakes with magnitude ML greater than 4.5 throughout Australia. This 'normal' level of strong intraplate seismicity reveals two times larger territory to be with higher seismic activity than the seismically quiet areas.

Compression with predominant East-West orientation is the main stress in continental Australia, and it is in nearly 90 degrees angle with the direction of the plate motion. The focal mechanisms in Tennant Creek area (Northern Territory) suggest a perpendicular compression axes in North-South direction. The composite fault plane solutions for the Western part of Australia show a steady and robust orientation of the main compression and are similar to the orientation of the individual focal mechanisms. The solutions for the Eastern part of the country are less stable with more differences with the individual events even in close neighbouring areas.

Cloeting S & Wortel W. 1986. Stress in the Indo-Australian plate, Tectonophysics, 132, 49 - 67.

Denham D, Alexander L & Worotnicki G. 1979, Stresses in the Australian crust: evidence from earthquakes and in-situ stress measurements, BMR Journal of Australian Geology & Geophysics, 4, 289 - 295.

Denham D, Weekes J & Krayshek C. 1981, Earthquake evidence for compressive stress in the Southeast Australian crust, Journal of the Geological Society of America, 28, 323 - 332.

Everingham I. & Smith B. 1979. Implications of fault plane solutions for Australian earthquakes on 4 July 1977, 6 May 1978 and 25 November 1978. BMR Journal of Australian Geology and Geophysics 4, 297 - 301.

Lambeck K, McQueen H, Stephenson R, Denham D. 1984, The state of stress within the Australian continent, Annales Geophysicae, 2, 6, 723 -742.

Spassov E, Kennett B & Weekes J. 1997, Seismic zoning of SE Australia, Australian Journal of Earth Sciences, v. 44, No 4, 527-534.

Suetsugu D. 1994. Lecture Notes: 'Practice on Source Mechanism'.

Zhang Y, Scheibner E, Ord A & Hobbs B. 1996, Numerical modelling of crustal stresses in the eastern Australian passive margin, Australian Journal of Earth Sciences, 43, 161 - 175.

Figure 1. Space distribution of the earthquakes with magnitude greater than ML 4.5 since 1902 across Australia.

Figure 2. Time distribution of the number of the events and seismic energy realised from the same earthquakes shown in Fig. 1.

Figure 3. Spatial distribution of the number of events (first digits) and seismic energy in 10^20 Joules (second digits) in cells with size 5 by 5 geographical degrees.

Figure 4. Individual fault plane solutions for two events in Southeast Australia - Picton, 1973, ML=5.5 (left) and Wannagatta, 1981, ML=5.0 (right).

Figure 5. Composite fault plane solutions for the earthquakes in SE Australia: one solution for the whole area (left); for zone North only (right top) and for zone South (right bottom).

Figure 6. Individual (left) and composite (right) fault plane solutions for events in Marryat Creek area (Central Australia).

Figure 7. Individual (left) and composite (right) fault plane solutions for events in Tennant Creek area (Northern Australia).

Figure 8. Individual (left) and composite (right) fault plane solutions for events in Kimberley area (Northwestern Australia).

Figure 9. Individual (left) and composite (right) fault plane solutions for events in Meckering area (Southwestern Australia).

Figure 10. The composite solutions across Australia with the direction of the compression stress (small arrows) and the general direction of the plate motion (big arrow).

Table 1. List of the Australian earthquakes included in the composite fault plane solutions.