DEPARTMENT OF PRIMARY INDUSTRIES AND ENERGY
BUREAU OF MINERAL RESOURCES GEOLOGY AND GEOPHYSICS
BULLETIN 236
Geologicaland geophysicalstudies in the AmadeusBasin, central Australia
R.J. Korsch& J.M. Kennard Editors
Onshore Sedimentary & Petroleum Geology Program
AUSTRALIAN GOVERNMENT PUBLISHING SERVICE CANBERRA 409
Teleseismictravel-time anomalies and deep crustal structure of the northernand southernmargins of the AmadeusBasin
K. Lambeckl
Teleseismictravel-times recorded acrossthe central Australian basins and Musgrave and Arunta Blocks impose signifrcant constraints on crustal and upper mantle structure. Major discontinuities in lateral structure are required, particularly acrossthe Redbank-Ormiston Thrusts in the Arunta Block and the Woodroffe-Mann Thrusts in the Musgrave Block. The deep structure of these tectonic units exhibit considerablesimilarity, and in both instances the thrusts dip at about 45" through to the Moho. Major offsets in Moho depth are produced which have persisted since the time of the last movements on the faults, about 300 Ma ago in the case of the Redbank Thrust and much earlier in the case of the Woodroffe-Mann Thrusts. The teleseismic models are consistent with deep crustal seismic reflection observations across the Redbank Thrust Zone, and they confirm the conclusion drawn from gravity studies that the region as a whole is not in local isostatic equilibrium and that maximum stress- differenceswithin the crust and upper mantle are of the order of 100MPa.
I ResearchSchool of Earth Sciences,Australian National University, PO Box 4,Canbena, A.C.'[.260I, Australia. lntroduction into which sedimentscan be deposited,rather than with the details of how this deposition occurs, although some form major feature Australia's Intracratonic basins a of of these models do specify the overall depositional pat- geology,yet the mechanisms leading to their formation terns (e.g.Beaumont & others, 1987)on the assumption poorly This is not remain understood. entirely a conse- that sediment accumulation follows the crustal warping. quence geophysicalprob- of inadequate exploration and The sedimentary record within the basins will often be ing of the basins and underlying crust, but also a much more complex than is allowed by these models consequenceof the complexity of basin formation mech- becausesedimentation will be partly controlled by local anisms for these intracratonic environments. Many of factors, including the nature of the sedimentsthemselves geological the basins have had very long histories of and localised brittle fracturing of the upper crust. Never- subsidence,rebound and deformations, during which no theless,the sedimentaryrecord, if generalitiesof deposi- single mechanism alone was responsiblefor the basin's tion can be established,provides one of the major initiation and evolution. In some cases, such as the constraints on these models of tectonic crustal evol- Amadeus Basin, these histories have spanned nearly ution. 1000million years. Any successful first-order model for the Amadeus Basin formation mechanisms can be characterisedas Basin evolution, as for any other basin's history must being driven by horizontal stresses (extensional and explain the subsidenceand sedimentation records con- compressive), thermal processes(including thermal tained within the basin sediments,as well as account for expansionand contraction, and changesin metamorphic the uplift histories of the adjacent exposed cratonic grade of the lower crust), or by passive gravitational blocks for the same time interval as sediments were loading. All mechanisms appear to contribute to a deposited within the basins or immediately prior to this basin's evolution at some time in its history and what deposition. Of particular importance will be the uplift distinguishes one basin from another is the relative histories of the southern part of the Arunta Block and importance of these complementarymechanisms at var- the MusgraveBlock. Also, the Amadeus Basin cannot be ious stagesof the basin's evolution, largely in response treated wholly in isolation from the other central Austra- to the tectonics that shaped the continent as a whole. lian basins, particularly the Ngalia Basin, the northern Furthermore, the responseof the crust to a given force Officer Basin and possibly the Georgina Basin, for it or load may be regionally variable becauseearlier tec- appearsthat all of these basins have had similar subsid- tonic events modifled the crustal properties, or because encehistories in responseto more regional forces (Shaw, of the boundary conditions imposed by the surrounding 1987;Shaw & others,in press;Lindsay & others, 1987). tectonic units. Early crustal and basin models for central Australia The Amadeus Basin is no exception to this and it is were almost entirely based on the large gravity anoma- improbable that a single mechanism can be found that lies observedover the basins and intervening blocks, and will explain all aspectsof the basin (Shaw, 1987, and did not attempt to answer the question as to how these this volume; Shaw & others, in press). Thus if a single structures had evolved to their present geometry. With mechanism model is proposed it can only lead to a peak to trough amplitudes in excessof 180 mgal in some frrst-order model, one which tries to explain the gross locations and with an approximate wavelength of features of the dominant crustal deformation and basin 200 km, these anomalies over the relatively flat terrain sediments.Second-order models must reflect a combina- of the Office4 Amadeus and Ngalia Basins and Musgrave tion of mechanical processeswhose relative importance and Arunta Blocks form the major features in the grav- has not remained unchanged through time in response ity fleld of the Australian continent; clearly the region is to the more regional forces that refined the shaping of out of isostatic equilibrium in any conventional sense, the Australian continent in Late Proterozoic and and the question is where do the subsurface density Palaeozoictimes. By their nature, frrst-order models are anomaliesoccur. Dynamic contributions associatedwith more concernedwith the deformation and tectonic evol- upper mantle small-scaleconvection must be ruled out ution of the crust, and with providing an environment because the last active tectonics occurred more than K. Lambeck t (years)
7 400 -f,.r o \!'ç9 g- ^s9 o L J +, (ü oL o. E |- ^ I t Çt.. 1000 Ë\b"
Depth(km)
Fig. 1. Effective mantle relaxation time as a function of crustal deformation can be tested. In particular, new temperature and stress-difference(for 100 MPa and seismic data, including teleseismictravel-time anomalies 400 MPa) for wet olivine accordingto Chopra & Paterson and deep crustal seismic reflection surveys, permit the (1981), and two geothermal profiles (1,2) according to models of present deep structure to be tested, although Sass & Lachenbruch (1979) spanning a range of values this does not directly test the process by which this that are consistent with surface geology,heat flow and structure was actually reached. The teleseismic data other geophysicalobservations for the central Australian consists of travel-time residuals observed at closely Shield. Qo is the surface heat flow (in heat flow units) spacedsites acrossthe basins and exposed blocks (Fig. and Ao is surface radioactivity (in heat generationunits). 2). Lambeck & Penney(1984) observedsignifrcant vari- The dotted line is the geotherm proposed by Cull & ations in the travel-times of teleseismic-wavesalong the Conþ (1983) for the central Australian region with a part twet' Central Australian line from the southern of the surface heat flow of 75 mWm'2. lThe flow law is Musgrave Block to north of the Ngalia Basin. Major (1981) found that adopted becauseChopra & Paterson azimuth-dependentstation anomalies were observed (0.010/oweight, havea evenvery snall amounts of water, near the Redbank Thrust Zone irt the southern Arunta weakening effect.) At a depth of 50 km the significant Block and near the Woodroffe and Mann Thrusts in the geotherm(1) predicts a temperatureof about 700'C and Musgrave Block. No evidence was found for major a 100 MPa stress-diflerence will relax with a time- crustal offsets or lateral boundaries beneath the constant of about 106 years. At 30 km the corresponding Amadeus Basin sedimentary succession.The result led time-constant exceeds the age of the Earth (from of two closer spaced, parallel, Lambeck, 1986). to the deployment north-south lines of instruments across the northern margin of the Amadeus Basin and across the southern 300 million years ago and, since then, the continent has Arunta Block (the Arunta and Redbank lines; Fig. 2), moved large distances relative to the underlying, sub- and these experiments confirmed the occurrenceof dif- lithospheric mantle. The source of the anomalous struc- travel-times of up to 1.5 seconds across the ture must therefore be in the lithosphere, and because ferential Zone (Lambeck & others, 1988). A the stress relaxation time constant also decreaseswith Redbank Thrust southern Amadeus depth (due to increasing temperature with depth), many fourth line of instruments acrossthe (the of these anomalies are likely to originate from within Basin and the Musgrave Block Musgraveline; Fig. the upper lithosphere (Fig. 1). Sucþ density anomalies 2) produced comparable large-amplitude travel-time must either be kept in place by the finite strength of this anomalies over relatively short distances. The second layer or be balanced by horizontal stresses.In either important new geophysical data set is the deep crustal case,maximum stress-differenceswithin this layer must seismic reflection profile recorded across the Arunta be of the order of 100MPa. Block and Amadeus Basin by the Bureau of Mineral Significant new geophysical data have become avail- Resources(line 5, Fig. 2). Only the deep structure for able in the past few years against which the models of the northern part ofthis line, acrossthe southern Arunta Tëleseismictravel-time anomalies and deepstructure of the northern and southern margins of the Basin
NORTHERN TERRITORY Mt. Allan
Napperby
ARUNTABLOCK
AMADEUSBASIN o Palm valley
Tempe Downs
( ...Central Australianline f I OFFICERBASIN SOUTHAUSTRALIA Fig. 2. Location map of the teleseismic travel-time sur- is recorded onto analoguetape together with time infor- veys: (1) central Australian line (Lambeck & Penneyo mation. All station positions have been surveyed with 1984); (2,3) Arunta and Redbank lines (Lambeck & either the Doppler or Global Positioning System (GPS) others, f988); (a) Musgrave Iine. Averagerecorder spac- satellite navigation methods. Unfrltered records of selec- ing is about 30 km along line I and about 7 km along the ted earthquake signals with high quality impulsive other three lines. Line 5 is the BMR seismic reflection onsets were digitised to produce the characteristic line across the Arunta Block. records (Fig. 3) for earthquakes from the Marianas, Japan,Tonga and Macquarie Ridge regions and recorded Block, has been examined in detail-so far (Goleby & along the Arunta line illustrated in Figure 2. The others,1989). observedquantity is the'station anomaly',the departure of the travel-time of the seismic wave to a station from Teleseismicobservat¡ons the nominal value predicted for a radially symmetric Earth model. These station anomalies are determined to Each earthquake recording site consistsof a single verti- within a constant value and as such they actually indi- cal component, short period, seismometerwhose output cate differential values. For closely spaced stations and 412 K. Lambeck 13 14 15 ttl lla¡lanas 721 1 2 3 a 5 7 Ee to fl 13 2 I 633 | 16t3r2tOEct.32t Fig. 3. Seismic records for four events recorded on the strongly azimuth dependent and, to a lesser degree, Arunta line. Station 1 correspondsto the southernmost dependent on epicentral distance, and mean station site in the Amadeus Basin (near Gosses Bluff) and anomalies are computed for earthquakesfrom the same station 15 lies at the no¡thern margin of the Ngalia Basin general source region. Figure 4 illustrates the station (near Mt Altan). The first three records are from three anomalies recorded along the north-south central Aus- deepolarge magnitude earthquakesarriving from north- tralian line for which the averagestation separation was erly (events 721 and 687) and easterly (event 633) azi- about 30 km. While this experiment indicated that large muths, respectively,while the fourth record is from a amplitude anomalies occurred, the spacing was inad- small magnitude, shallow crustal earthquake on the equate for detailed modelling and in subsequentexperi- Macquarie Ridge to the south-southeastof the line. ments this spacingwas reduced to 10 km and less. Without a better distribution of seismic recordersover for earthquakesfrom within a nanow distance and azi- the region and without better azimuthal coverageofthe muth range, they are indicative of lateral structure earthquakes source regions, it is not possible to carry beneath the stations. Observationsof such anomalies as out comprehensive3-D modelling for crustal and upper a function of azimuth and distance permit the P-wave mantle structure. Consequently only 2-D models are relative velocity structure to be determined beneath the consideredin which it is assumedthat there is structural stations with a resolution approximately equal to the symmetry along the tectonic strike. Surface geology and station separation. The methods of data reduction and gravity data indicate that the east-west structure across analysis for the station anomalies are discussed in the northern margin of the basin is relatively uniform, Lambeck & Penney (1984) and Lambeck & others and the similarity of the travel-time residuals recorded (1988). Tables l-4 summarise the data for the along the two parallel north-south lines confirms that, to teleseismic experiments. The station anomalies are a first ordel the deep crustal structure here can be Tþleseismictravel-time anomalies and deepslruclure of the northern and southern margins of the Basin I ro r9 20 2l 22 ,:[ Fig. 4. Observed seismic fravel-time anomalies across central Australia (solid circles with error bars) for different azimuths.Station 1 lies near EverardPark in the south and station 25 lies near Anningie. The Fiji-Tonga eventsarrive from the east, the Japan and Mariana events from the north, and the Macquarie Ridge event from the south-southeast.The Macquarie Ridge result is based on a single earthquake record of relatively poor quality. The vertical scale (in seconds)establishes the amount by which arrivals are relatively early or late. 414 K. Lambeck Thble 1. Central Australia line: data for teleseismic Täble 2. Arunta line: data for teleseismicexperiments. experiments Station coordinates Station coordinates Station Station Number Name Latitude Longitude Number Name Latitude Longitude I Undandita -23.8550 t32.16s0 -23.6607 I Eve¡ard Park -26.91 t32.'10 2 TÌrrkey Bore t32.2665 -23.5907 r32.1580 J Kenmore Park -26.70 132.68 3 Erajaku Well -25.98 + Stokes Well -23.4883 t32.1283 Sentinel Hill L5¿.5 I -23.3807 5 Mulga Park -25.96 t3t.74 5 Glen Helen 132.1092 -25.72 6 Derwent -23.29s0 t32.tr67 6 No. 3 Bore 13r.73 -23.1683 7 Curtin Springs -25.34 131.86 7 Derwent t32.t561 8 Angus Downs -25.t7 132.1 8 Station 9 Liddle Hills -24.89 r32.28 8 Mikes Bo¡e -23.0742 t32.r640 l0 Liddle Hills -24.72 132.32 9 Ace Bore -23.0202 132.1653 North 10 Ace Bore North -22.88s8 132.1447 -24.60 11 Mount Wedge -22.7428 132.1605 ll Wallara Ranch 132.39 -22.6317 t2 Ayers Lookout -24.s0 t32.61 t2 CassidyBore 132.2308 IJ Tèmpe Downs -24.40 132.45 South 14 Illara Creek -24.29 r32.35 13 Cassidy Bore -22.5458 132.2r50 Missionary -23.88 132.39 I4 No. I Bore -22.4702 r32.1 980 15 -22.3885 Plain 15 Mt Allan t32.t832 Bluff -23.77 16 Gosses r32.36 region, azimuth and distancerange MacDonnell -23.61 Source t7 132.36 Source Distance Azimuth Number of Range regíon range () range (') events 18 Stokes Well -23.49 r32.13 t9 Papunya -23.27 r32.03 Fiji-Tonga 43-52 87-1 05 15 20 Ace Bore -23.02 132.16 Kermadec 45-4) 108-1l6 9 2l Mt Wedge -22.75 r32.t7 Japan Täble 3. Redbank line: data for teleseismicexperiments mately with the Redbank Thrust and the granulite ter- rain immediately to the north of it, and relatively low Station coordinates Station velocity material to the south of this zone. The north- Number Name Latitude Longitude ward dipping surface separating the high velocity zone must I Palm Valley -24.01 132.62 from the underlying lower velocity zone dip 2 8-Mile Yard -23.89 132.67 steeply,at about 45', down to depthsof 40-50 km and Bore possibly down to 60-80 km. There is no evidencethat -23.7 J Rodina Bore 5 132.67 this surface shallows with depth. The resol- -23.66 the dip of 4 Ormiston 132.74 ution of the geometry of the downthrust wedge of lower Gorge poor. 6 Redbank Hut -23.51 132.8r crust, howeve4is Beneath the Amadeus Basin and 7 No. l0 Bore -23.38 r32.74 the Arunta Block south of the Redbank Thrust, the 8 No. 9 Bore -23.24 t32.72 Moho appears to be shallower than it is beneath the o -23.17 No. 35 Bore 132.70 northern part of the line and to dip gently northward, 10 No. l7 Bore -23.07 r32.66 but farther north, starting south of the Ngalia Basin, this lt QueensBore -23.02 t32.71 t2 Herbert Well -22.94 \ 5¿-O I dip is greater.Resolution north of the Ngalia Basin is \ñ'est poor. Figure 8 illustrates the results of ray tracing -22.77 l3 CabbageTiee r32.68 through such a model for seismic wavesemanating from Bore the Fiji-1bnga, Japaî and South Sandwich regions. The t4 8-Mile Soak -22.63 132.7| South predicted travel-time anomalies for this and other azi- l6 GeorgesYard -22.45 132.72 muths are in satisfactory agreementwith the observed l7 Mica Dam -22.34 132.68 values. Variations of this model have been discussedby Lambeck & others (1988). The depth and gradientsof Source region,- azimuth and distance Source Distance Azimuth Number of the various surfacesare functions ofthe adopted veloc- region range () range f) Events ities for the crustal layers and upper mantle, but the Japan 51-65 0-8 9 principal characteristics illustrated appear to be essen- Fiji-Tonga 42-50 91-100 8 tial. The model is consistentwith the residuals observed Macquarie 33-41 t66 J along the Redbank line (Fig. 8) although small differ- Ridge encesdo occur. In particular, the wedge of downthrust Station anomalies lower crust is less extensivebelow this easternline, con- Statìon sistent with the east-west variation in the amplitude of Number Japan Fiji-Tonga Macquarie the gravity maximum observed to the north of the I 0.48 0.24 0.18 RedbankZone (Lambeck & others, 1988). 2 0.55 0.2r 0.19 The model is consistent with the gravity observations J 0.43 0.19 0.09 À 0.36 0.04 0.00 (Fig. 9) although short wavelengthsmall amplitude vari- o 0.40 0.r4 0.05 ations in the observed gravity are not represented by 7 -0.43 0.06 - 0.04 such a model. These variations are readily attributed to -0.67 -0.04 - 8 0.15 the observed changes in the densities of the surface 9 -0.80 -0.08 - 0.02 10 -0.76 -0.60 - 0.02 rocks acrossthe structure (Lambeck & others, 1988). ll -0.62 -0.69 0.00 The residual gravity also exhibits a systematic gradient t2 -0.53 -0.70 - 0.21 13 -0.33 -0.46 - 0.28 14 0.42 0.17 0.16 r6 0.02 0.59 0.l4 t1 0.03 0.53 station anomalies for events arriving from the north (Japan and Marianas); arrivals to the north of the RedbankZone are early by as much as 1.5 s compared with arrivals to the south of this zone. The station anomalies from easterly azimuths (Fiji, Tonga and Kermadec) are smaller in azimuth but still vary signif,- cantly along the line. An important observation is that the peak early arrivals do not occur at the same sites for the two azimuths. Earthquakes from southern azimuths (Macquarie Ridge and South Sandwich Islands) are few and the shallow Macquarie eventsproduce poor quality records.Nevertheless, the resulting station anomalies are indicative of a very different variation along the line. Figure 7 illustrates the features of the model that are required by the Arunta line observations. Nominal crustal and upper mantle velocities have been adopted (6.3 km s-1 for upper crust, 7.0 km s-1 for the lower crust, 8.3 km s-1 for the upper mantle). The division of the crust into two layers is arbitrary and the deformation of the interface between these two zones is assumedto follow that of the Moho. The boundaries are assumedto be sharp discontinuities rather than gradational. The significant features of this model are a zone of relatively high-velocity material dipping steeply northward such Fig. 5. Detailed location map of the Arunta and Redbank that its continuation to the surface coincides approxi- lines. K. Lambeck Thble 4. Musgrave line: data for teleseismicexperiments Station coordinates Station Number Name I Mingeemealinna -27.00t0 r31.0940 2 Pantanya -26.9548 13t.0924 3 Weinna Hill -26.9030 13r.0949 4 Weinna Hill -26.8t82 t3r.0973 North 5 Carbeena -26.7507 6 Witjunnga -26.677| 7 Mt Harriet -26.6t61 South 8 Mt Harriet -26.s3t8 l 31.0603 9 Wanngatinya -¿o.+ I JJ 131.0655 10 -26.4193 131.0792 1t Pampili -26.3306 13t.0745 t2 No. 25 Bore -26.2706 131.0832 13 Amata -26.t960 r31 .l 156 t4 Mt Woodward -26.r 008 131.1661 l5 Border -26.0216 t3r.2ts6 l6 Opparinna -25.9314 r31.2t79 Creek t7 Britten Jones -2s.8362 13t.23t4 C¡eek Benda Hill -25.75t2 13t.2209 Benda Hill -25.6594 13r.2153 North 20 -25.5702 131.t952 2l Uluru South -25.4799 13 1.1 987 22 Uluru -25.392r 131.1931 ZJ Petermann -25.3t78 13 1.1 849 Road aÁ Lassiter -25.19r1 131.1849 Highway 25 Lake Amadeus -24.9455 131.1758 Sourceregion, azimuth and distance range Source region Distance Azimuth Number of range (") range f) events Fiji-Tonga 44-51 90-97 IJ Japan 54-61 359-1l l0 Mindanao 3¿-JO 349-352 4 South 93-96 r67-168 ¿+ Sandwich Islands Kermadec 44-45 106-108 J New Zealand 4r-43 116-118 2 SE Indian 29-32 2r2-22r 2 Ocean Station anomalies South Station Sandwich Kermadec New SE Indian number Islønds Islands Zealand Ocean I -0.34 -0.22 0.01 - 0.05 - 0.20 - 0.06 - 0.11 2 -0.25 -0.09 0.05 J -0.15 -0.r2 0.06 - 0.18 0.00 0.04 - 0.40 ^ -0.17 -0.22 0.03 - 0.09 0.00 0.13 - 0.46 5 -0.09 -0.24 0.1I 0.13 6 0.13 -0.40 0.20 - 0.06 - 0.38 1 0.21 -0.50 0.21 - 0.21 - 0.49 8 0.19 -0.32 0.13 - 0.27 0.43 -n.qs -n.n 9 0.l0 0.20 0.09 - 0.15 0.40 - 0.33 - 0.29 t0 0.l0 0.04 0.40 0.38 - 0.46 11 0.08 -0.04 0.0s 0.42 - 0.08 0.r2 0.63 l2 0.07 0.00 0.06 0.26 - 0.08 0.13 0.56 l3 0.03 0.01 0.05 0.17 - 0.05 - 0.02 l4 0.09 0.12 0.02 0.17 0.00 0.r0 0.19 l5 0.10 0.18 0.05 0.08 0.04 0.33 r6 0.13 0.27 0.05 0.12 0.10 0.09 0.r2 t7 0.36 r8 0.09 0.24 0.0s 0.06 0.1I 0.18 0.13 19 0.16 0.26 0.04 0.08 0.10 20 -0.03 0.15 0.20 0.00 - 0.0r 0.02 n.n 2I 0.06 0.24 0.06 0.02 0.09 22 0.03 0.20 0.1I - 0.01 0.1I 0.07 0.l6 z) -0.12 0.03 0.28 - 0.24 0.04 - 0.08 - 0.24 aÁ -0.22 -0.03 - 0.37 - 0.06 25 -0.18 -0.13 0.41 - 0.30 - 0.07 - 0.34 - 0.14 Tþleseismictravel-time anomalies and deepstructure of the northern and southern margins of the Basin A ARUNTALINE B REDBANK LINE Fig. 6. Station anomalies recorded along A) the Arunta of the basin extends southward from Lake Amadeus. line and B) the Redbank line. (Station locations are indi- across the Woodroffe, Davenport and Mann Thrusts cated in Fig. 5.) Negative values correspond to early near Amata to about latitude 27'S (Fig. 11). Figure arrivals. t 2 illustrates the station anomalies for eventsfrom both the north and south of the line as well as from the east. acrossthe structure and an iteration between the gravity Large amplitude anomalies are observed that are and teleseismic modelling is warranted. The model is strongly dependenton the azimuth of the incoming seis- also consistent with the deep crustal seismic reflection mic waves.As for the Redbank Thrust, the occurrence sectionsrecorded acrossthe structure (Goleby & others, of the rapid changesin arrival times are offset from the 1989; Goleby, 1989). In particular, this latter data set surface expressionsof the thrust and faults, indicating supports some of the principal characteristics of the that these structures dip steeply and extend into the model: the Redbank Thrust dipping at about 40' down lower crust and possibly involve the mantle. Figure to at least 30-35 km depth and possibly as deep as l3illustrates the model inferred from these results, 50km; a signifrcant variation (about 10-15km) in using the same crustal and mantle velocities as before. Moho depth acrossthe structure; and the existenceof a The model bears many similarities to the Arunta model, seismicallytransparent zone immediately to the north of particularly for that segment of the Redbank Thrust the thrust that is suggestiveof lower crust and mantle crossedby the Redbank line. The features of the model material to the north of the thrust overriding lower and that are required by the Musgrave data set include: (i) a upper crustal material to the south (Fig. l0A,B). gently sloping Moho from below Lake Amadeus to south of the Woodroffe Thrust; (ii) a wedge of relatively low Southern marg¡n of the Amadeus Bas¡n and the velocity material dipping down to a depth of about Musgrave Block 50-60 km immediately to the north of high velocity The teleseismic experiment acrossthe southern margin material, with the boundary between the two bodies 4r8 K. Lambeck Japan Tonga South Sandwlchlslands *xxxxxxx¡xaxx. Fig. 7. Crust-mantle models based on the Arunta line maximum lies well to the south of the Woodroffe-Mann travel-time anomalies and results of ray tracing (x) for deformed zone and resemblesa short-wavelengthanom- Japan, Fiji-Tonga and South Sandwich Island earth- aly superimposed on a broader feature, similar to that quakes.Observed values are indicated bV (Â). seen elsewherein the Musgrave Block over Giles Com- plex type bodies (e.9. Mt Caroline and Mt Davies). dipping southward at about 45" and corresponding Central Amadeus Basin approximately at depth with the position of the Mann The only travel-time information available for the cen- Fault; (iii) considerable variation in crustal thickness tral part of the Amadeus Basin, between about Lake north of the Woodrofe Thrust, with the possibility that Amadeus and Hermannsburg, is from the Central Aus- the crust beneath the basin is thinner than average;(iv) tralian line where the averagestation spacing was about lower or intermediate crust near the surfaceimmediately 30 km. Here the station anomalies show little variation, to the south of the Woodroffe-Mann thrust zone and with arrivals near the centre of the basin, between also at the southernmost part of the line. Agreement of Liddle Hills (station 9) and Wallara Ranch (station l0), this model with the gravity data (Fig. 13B) is satisfac- being relatively early and exhibiting no strong azimuthal tory although iterative modelling of both the gravity and variation (Fig. a). This indicates that no major crustal teleseismic data is warranted here as well. The gravity velocity anomalies nor major offsets in Moho depth Tþleseismictravel-time anomalies and deepstructure of the northern and southern margins of the Basin 4r9 Tonga uppef cfust lower crust mantlo Tonga Fig. 8. Crust-mantle model based on the Redbank line travel-time anomalies and the predicted (x) and observed (.) travel time anomaliesfor Fiji-Tonga, Japan and South Sandwich Islands. = 1 6 I o E E .9 o ì *;oo @ ztr E J -50 6 o o T .9 -5 .2 o o dt -100 Fig. 9. Gravity observationsalong the Arunta line and the = o difference with the conrputed anomalies for the o crust-nrantle model shown in Figure 7. o E K. Lambeck A s 15000 16000 17000 .,iii,* tr, o goB r 3 o--,-,-ls km MIGRATED TRAVERSE14. ENERGY ARUNTA BLOCK STACK SOUTHERN ARUNTA CENTRALARU NTA PROVINCE NORTHERNARUNTA PROVINCE PROVINCE NGALIABASIN 0 ì:-N'." -ì\\ rì ->{ã '--f: *-' - ^'1, _ ts( ì'-Ì-:ìJ> ' ).- 6sJ DEEPSE|SMtC REFLECTTON MODEL F'ig. 10. ,4,)Migrated line drawing of an energy stack of Discussion the seismic section across and north of the Redbank Thrust. (From Goleby & others, 1989.) B) Interpreted Crustal model: Relat¡onw¡th grav¡ty,topography and seismic section of the migrated line drawing of a coher- surface geology ency stack along the same section. ONTZ, Ormiston Nappe Thrust Zone; RDZ, Redbank Deformed Zone. The significantly early station anomalies correspond roughly with surface outcrops of basic granulites of the beneath the basin can be identifred by this method, and Arunta and Musgrave Blocks (Figs 6, l2), but in both that the observations are consistent with a model in casesthe earliest arrivals are offset from the occurrence which the Moho is relatively shallow near the centre of of the highest metamorphic gradesand they occur over a the basin and deepens toward both the northern and wider zone than the surface outcrops. The relatively late southern margins. A closer spacing of receivers across arrivals occur not over the basins but over the granitic this part of the basin is neverthelessdesirable because gneiss outcrops between the surface expressionsof the resolution of the model here is poor and minor Moho deformed zones and the basin margins; that is, to the ofsets or lateral crustal structure could go undetected. south of the Redbank Thrust Zone and to the north of No constraints are provided on the absolutethickness of the Woodroffe-Mann deformed zone. No major anoma- the crust in this region. lies occur at the basin margins as deflned by the Tëleseismictravel-Íime anomalies and deepstructure of the northern and southern margins of the Basin 421 highs would be expectedto correspondto major topo- AMADEUSBASIN graphic depressionsand vice versa, and the region \, .2s__@__ between the Woodroffe-Mann and Redbank Thrust zones would stand as much as 3 km above the adjacent blocks. The gravity maximum in the southern Arunta .24 does correspond to the small topographic depressionof Mt Olga Lake Lewis and related features (Fig. 9) and there is . .23 of the mid-basin gravity high with the Ayers Roðk correspondence ¡22 Lake Amadeus depression. The regional topographic .21 high in the northern part of the Amadeus Basin coin- cides roughly with the gravity minimum. Similar rela- tions are not evident in the Musgrave Block and southern basin margin. Stress implications A measure of the maximum stress-differenceor of the provided anom- WoodroffeThrust non-isostatic stressin the crust is by the alous mass in adjacent vertical columns of the lithosphere, which is estimated to reach 150 MPa \ \ (Lambeck & others, 1988).This anomalousstress ñeld ____-. .10 .9 is primarily north-south whereas the east-west com- \a .8 lf/ ponent is much smaller and is essentially in isostatic / equilibrium (Stephenson& Lambeck, 1985). There is '6 wintiqinna L' / some evidence from the previously mentioned relation -___:5-_------/ between gravity and topography that the crust is trying .4/ to reach an equilibrium state, but the rates of adjust- ^.s- g-i1 - ment are orders of magnitude less than would normally be expected. The rebound process is retarded either becausethe crust or lithosphere has reached a state of mechanical equilibrium between the vertical buoyancy O'+ 50 100km and surface loads and horizontal compressiveforces, or the structure has been frozen into a very thick and cold lithosphere such that the stress-differencesare supported regionally within this layer. That the central Australian region is now in an overall state of horizontal compres- OFFICERBASIN sion is left in little doubt with the occurrence of the recent north-south thrusting earthquakes in the the Musgrave Fig. 11. Location map of stations along Musgrave Block and farther to the north near Tènnant gravity lineaments tine. The major positive and negative Creek (Bowman, 1988). But can such forceshave per- are shown by the dashed lines. sisted since the last major orogenies,the Alice Springs Orogeny of Late Devonian-Carboniferous age and the Amadeus (or MacDonnell) Homocline in the north and Petermann RangesOrogeny of Early Cambrian or older the outcrop of the Dean Quartzite in the south (between age? stations 18and lgon the Musgraveline, Fig' 11). The cooling lithosphere model requires that after the Neither do anomalies occur where the major thrusts occurrenceof the orogenythe crustal structure was held reach the surface. In particular, there is no seismic in place for a sufficiently long interval for the lithosphere signal acrossthe shallow-dipping Vy'oodroffeThrust nor to cool to the point where its finite strength was across the Petermann Ranges Nappe recognised by adequate to support the anomalous stress-state.This Forman (1966). Crustal thickness beneath the basin could be achieved if the thrusting events produced sig- appears to be somewhat less than beneath the adjacent nifrcant topography and a syn-orogenic sediment load exposedbasement. acrossthe deformed zone, so that af the completion of A general agreement exists between the travel-time the tectonism the crust was essentially in a state of anomalies and the gravity anomalies,with early arrivals isostatic equilibrium. Then, assuming that no subse- corresponding to gravity highs and vice versa. Signifi- quent rebound occurred in response to erosion, the cant offsets do, howeve4 occur across both deformed minimum topographic load would equal the anomalous zones.Using nominal densitiesof 2.7, 2.81 and 3.3 g stress-differenceinherent in the present model, or about cm-3 for the upper crust, lower crust and mantle respec- 3-4 km. As this eroded away with a regional erosional tively, and 2.6 and 2.5 g cm-3 for the Proterozoic and time constant of the order of 108 years (Stephenson& Palaeozoic sediments respectively,the model-predicted Lambeck, 1985;Lambeck & Stephenson,1986), roughly gravity anomalies are in general agreement with the equal to or greater than the thermal conduction time observations,and the model explains the essentialchar- constantofthe lithosphere,the build up ofnon-isostatic acteristics of the gravity field over the region. stress differences would be supported by the cooling The variation in topography over the region as a mechanical lithosphere. It is important to note that the whole is generally mild. Averageelevations, for 10' lati- magnitude of the maximum stress-differencesdo not tude by 10'longitude blocks, range flom about 600 m in change significantly throughout this processbut that the the north to about 400 m to the south of the Musgrave depth distribution of these stressesthroughout the Block, although individual mountains rise up to 1000 m lithospheredoes. above the surrounding plains of alluvium, sand dunes and clay pans. There is not the clear relation between Timing of orogenic events topography and gravity that would be expected if the It is reasonableto assumethat the present crustal struc- region was in isostatic equilibrium. Typically,the gravity ture acrossthe two deformed zonesformed largely at the À N) N.) Travel-timeresiduals (sec) Travel-timeresiduals (sec) Travel-timeres¡duals (sec) I I I Late Early I Late Earlv I Late Early Ol o (¡ ; o (rl o N À Ø=4 E ô.'.v .d Fä- !?lP i:t= o ?8 å Ê'oq õ' U) itã Þ o IhE á c FØ O Eoã U'¿ ÈEÞ !'ß F õúD aØ 3q¡ @ 5ã Q-5 vta C- Ée 9) rÞ= € ox o õ' 9) ú=r>rÞ = f E= 3i'+ -w tÞ x a7 Þrì fr Ëã Fç ¡^Z (\ .;'(Þ ñ,ñ Tþleseismictravel-time anomalies and deepstructure of the northern and southern margins of the Basin 423 time of the last major movements on the thrusts. A trarily permitted to occur on these zones of weakness. knowledgeof the timing of these movementsis therefore Movement was assumed to occur at the time of, or of importance in understanding the responseof the con- preceding, the Petermann RangesOrogeny at the south- tinental lithosphere to tectonic forces; the further back ern margin, and again during or before the Alice Springs in time these events occurred, the greater becomesthe Orogeny at the northern margin of the Amadeus Basin. question of how these structures have been supported Other limitations of the model were that the lithosphere through time (Fig. l). In the caseof the Redbank Thrust was assumedto be rheologically homogeneous,that the the timing of the Alice Springs Orogeny appears to be isotherms remained horizontal at all stagesof the defor- well constrained as being of Late Devonian-Early mation, and that the rheology was lineaq corresponding Carboniferous age (Shaw, 1987). These movements to a viscoelastic (Maxwell) layer. The principal conse- occurred over a 30-50 km wide zone but particularþ on quence of these assumptionsis that the estimates of the the Ormiston and Redbank Thrusts (Shaw, 1987), and required stressare excessive.The features of the defor- produced the characteristic up-turned northern margin mation are, howeve4 preserved.A further limitation of of the Amadeus Basin. A minimum age for the current the model was that the in-plane compressiveforce had crustal structure is therefore about 300 Ma and the non- remained constant throughout the basin-forming inter- hydrostatic stressesmust have persisted since this time. val. However,as tectonic processesat the plate bound- The timing of the Petermann RangesOrogeny, when the aries change, the regional stressfield must also change last movements occurred on the Woodroffe-Mann and the horizontal stress state is consequently time deformed zone, is less well constrained. Following the dependent;the resulting subsidencecurves for the basin work of Forman (1966, 1972) this orogeny is commonly may be (and usually are) quite complex. In particular, believed to be of Early Cambrian age and to have periods of reduced stress following periods of intensive involved the folding and metamorphism to kyanite compression, will result in subsidencecurYes that are grade of the Dean Quartzite and overlying Pinyinna similar to those predicted from thermal models. Beds along the northern margin of the Musgrave Block. Despitethese limitations certainaspects of this model Recent work by Maboko (1988) suggeststhat the last remain important. First, the role of erosion and sedi- major movement may have occurred somewhat later. mentation in enhancing crustal deformation is very sig- There are, however, a sufficient number of curious nificant and this will also be true in any alternate model. aspects about the timing of this event to warrant its Second, the notion that the lithosphere acts as a stress re-examination. Whatever the age, the implication is guide to forces originating at plate boundaries, and that large stress differences have persisted for at least therefore leads to similar subsidencecycles in adjacent 500-600 Ma. basins, will also be applicable to other models (see also Lambeck & others, 1987). Third, the combination of Models of formation flexural deformation and movement on faults also seems which attempted to link the Late A flrst-order model to be a requirement of successfulmodels. One of the tectonic evolution of the central Proterozoic-Palaeozoic more controversial points of the original model was the (the Amadeus, Officer and Ngalia Australian basins requirement for significant ductile deformation on time and of the intervening cratonic crust (the South- Basins) scales of 108 years that permitted the initially small Arunta and Musgrave Blocks) is that of Lambeck ern elastic deformations to grow with time to significant (1983a, 1984). In this model the displacementof the amounts. In the original model a viscoelastic rheology crust, predominantly vertical, is driven primarily by the was adopted but the principal resulf, that the layer lithosphere acting as a stress guide to horizontal com- became effectively weaker with time under a constant pressionalforces whose origin lies outside the region. In stressñeld, is equally valid for other ductile models (see, the presence of crustal heterogeneities,such forces, for example,McAdoo & Sandwell,1985). aided by the surface transport of material eroding from In the original model the basin evolution was consid- uplifting basement and deposited into subsiding basins, have been driven primarily by more-orJesscon- can lead to signiflcant crustal warping or flexure if the ered to in-plane compression.The careful work by Shaw crust respondsto these loads and forces in a non-elastic tinuous (1987; & others, in press; Shaw, this manner.A primary motivation for this model was that it see also Shaw had a considerablymore coupledthe Late Proterozoic-Phanerozoic tectonic evol- volume) revealedthat the basin at least nine distinct episodes ution of the various basins within the central Australian complex history involving be correlated region to provide a unified model for the region as a of subsidence.Many ofthese episodescan the plate mar- whole. This model implied that basins and eroded with major compressivetectonic eventsat phases with cratons are out of isostatic equilibrium and predicted a gins but some appear to be associated predominantly and the crustal structure broadly consistent with the gravity periods of extensionaltectonics, observations(Stephenson & Lambeck, 1985).Finally, it subsequent basin evolution has been episodic rather phases seemedto satisfy the gross subsidencecharacteristics of than continuous. Not all of basin formation the basins; namely, an initially slow evolution culminat- appear to be fault controlled. ing in rapid subsidence and thrusting near the newly A model which contrasts with the ductile deformation formed basin margins where the final formation process model discussedabove is the tilt-block model in which is comparable to foreland basin models. As such, the moderately dipping thrust faults penetratedeep into the model provides a mechanism for producing or reactivat- crust, and possibly through it, and break the crust into a ing major thrusts and for producing orogens in the number of blocks that respond to an applied inplane interior of cratonic blocks, away from plate boundaries stressby tilting relative to each other (Fig. 14; McQueen (Lambeck,1983b). & Beaumont, 1989). The tilt-block model shareswith Even within this restricted sense,'the model had a the previous model the important aspect of responding number of major limitations. One is that the adopted to distantly generated stressfields so that coupled evol- analytical description could not handle discontinuities ution of adjacent basins and blocks is a natural conse- in the crust; movements on existing faults were not per- quence.In addition, movement is signif,cantly enhanced mitted until flexural stressesreached failure limits of the by erosion and sedimentation, and is further modified lithosphere as a whole, and movement was then arbi- by the flexural deformation of the tilted blocks and by 424 K. Lambeck .F- j to c J (Ú o, c ,oo (U (Ú E '.3È at, o, õ 'e c c c Eg3 jJ- = (ÚC" > s3 15 ll 0 E Jzo -c 8¿o o = (ú c') E -50 (Ú CI o J o, :t o -100 dl Fig. 13. ( A) Crust-mantle model basedon the Musgrave nately, the crustal structure predicted by both models is line travel-time anomalies. (8, on facing page) Gravity very similar and consistent with the seismic models. observationsand predictions along the Musgrave line. Much tends to be made of distinctions between so- (C,D,E) Results of ray tracing through the model illus- called 'thick-skinned' and 'thin-skinned' models of trated in (A) for Japan, Fiji-Tonga and South Sandwich crustal deformation. Insofar as the seismic evidence Islands earthquakes, respectively.The observedarrivals (both travel-time anomalies and deep reflection data) are boundedby the 11 standarddeviation from the mean. points to an involvement of the crust as a whole in the deformation, and to the major faults being planar down 'thick-skinned' any eventual stress relaxation. The primary conceptual to at least the Moho, models must be distinction between this model and the crustal warping favoured, and the 'thin-skinned' model proposed by model is the role played by the shear zones. In the TÞyssier(1985) must be ruled out for this central part of 'thin-skinned' crustal warping model, flexural deformation occurs until the basin (Goleby, 1989). However, stressdiferences locally exceedthe strength of the crust models are likely to have validity toward the eastern and the subsequentmovement occurs on new or reacti- (e.g. the Arltunga area) and possibly western limits of vated thrusts. Deformation and basin formation is poss- the basin-block interface becauseof the complex stress ible without faulting occurring but, unless stress frelds that may develop at the limits of the thrust in relaxation is important or unlesslong time intervals are responseto changing boundary conditions. involved, these basins are not very deep. The spacing of An impofant question concerningthe basin evolution the reactivated faults will be strongly influenced by the is the tectonic condition that led to deposition of the location of previously establishedzones of relative weak- earliest sediments - the Heavitree Quartzite in the ness, and possibly by the flexural wavelength of the Amadeus Basin and the Vaughan Springs Quartzite in crust. In the tilt-block model the primary movement the Ngalia Basin. Widespread mafic dyke swarms pre- occurs on the fault or over a broad shear zone. Unfortu- ceded these deposits, but the sediments lack the rapid (\¡\ì ñ G Travel-timeres¡duals (sec) Travel-time residuals(sec) g (sec) I Travel-time residuals I c: I o e¿o o or or d Late Early i¡ Late o Early i¡ i¡ Late o Early ¿n ïcò ñ È !\ .ß* 4 ñ oj s" N ò \ s. cò\ s, cò ua (\ I'i q 426 K. Lambeck AO \ \\\\\r\ ì \ \ \\ \ t \ t \ tr \ \ I \ a I \ì \\\\ \\rìr\\\ \\\\\.\\ \ \\\\r\\\ \\\ \\\\\\ \ rl\\rlì \ -/ \ \- \-.2 \ / / \ \/ rr . ir'r\' \\\rrrr\ \\\\\\r \ \ \ \ \ \ \ \\ I ìì.,'ì.,' Fig. 14. Hypothetical models of tilted block style of Goleby, 8., 1989 - The crustal structure of the Arunta Block, deformation of a plate characterisedby varying strength central Australia: Results from deep seismic reflection pro- with depth (Flom McQueen & Beaumont, 1989.) The frling. Ph.D Thesis, Australian National University, Can- density and velocity contrasts across the Moho produce berra (unpublished). Goleby, 8., Shaw, R.D., Wright, C., Kennett, B.L.N., & the gravity and travel-time anomalieswhere this surface 'thick- K., 1989 - Geophysical evidence for is offset. Lambeck, skinned' deformation in central Australia. Nature, 337, 32s-330. facies and thickness variations that are characteristic of Lambeck, K., 1983a - Structure and evolution of the extensional mechanisms, and upper crustal extension intracratonic basins of central Australia. Geophysical seems to have been minimal (Shaw, 1987; Shaw & Journalof the RoyalAstronomical Society,74,843-886. others, in press). The tentative interpretation of the Lambeck, K., 1983b - The role of compressiveforces in teleseismic travel-times, that the crustal thickness is intracratonic basin formation and mid-plate orogenies. somewhat less under the basin than under the adjacent GeophysicalResearch Letters, 10, 845-848. basement,does support some crustal extension at some Lambeck, K., 1984 - Structure and evolution of the Amadeus, very early stageofbasin evolution, and this point bears Officer and Ngalia Basins of central Australia. Australían closer examination. Journalof Earth Sciences,3l, 25-48. Lambeck, K., 1986 - Crustal structure and evolution of the central Australian basins. 1n Dawson, J.G., Carswell, D.4., References Hall, J., & Wedepohl, K.H. (Editors), The nature of the lower continental crust. Geological Society of London, Beaumont, C., Quinlan, G., & Hamilton, J., 1987 - The and its relationship to the evolution of SpecialPublication 24, 133-145. Alleganian orogeny - the eastern interior, North America. In Beaumont C., & Lambeck,K., Burgess,G., & Shaw,R.D., 1988 Tèleseismic Tankard, A.J. (Editors), Sedimentary basins and travel-time anomalies and deep crustal structure in central basin-forming mechanisms.Canadian Society of Petroleum Australia. GeophysicalJournal, 94, 105-124. - Geologists,Memoir 12, 425-445. Lambeck, K., Cloetingh, S., & McQueen, H.WS., 1987 Bowman, J. R., 1988 - Constraints on locations of large Intraplate stressesand apparent changesin sea level: The intraplate earthquakesin the Northern TÞrritory Australia, basins of northwestern Europe. Canadian Society of Pet- from observations at the Warramunga Seismic Array. roleum Geologists,Memoir, 12, 259-268. - GeophysicalResearch Letters, 15, 1475-1478. Lambeck, K., & Penney,C., 1984 Tèleseismictravel-time Chopra, P.N., & Paterson,M.S., 1981 - The experimental anomalies and crustal structure in central Australia. Phys- deformation of dunite. Tectonophysics,78, 453-47 3. ics of the Earth and Planetary Interiors,34, 46-56. - Cull, J., & Conley, D., 1983 - Geothermal gradients and heat Lambeck,K., & Stephenson,R., 1986 The post-Palaeozoic flow in Australian sedimentary basins. BMR Journal of uplift history of south-easternAustralia. Australian Journal Australian Geology& Geophysics,8,.329-337. of Earth Sciences,33, 253-270. Forman, D.J., 1966 - Regional geology of the south-west Lindsay, J.F., Korsch, R.J., & Wilford, J., 1987 - Timing of margin, Amadeus Basin, central Australia. Bureau of Min- the breakup of a P¡ote¡ozoic supercontinent: Evidence eral Resources,Australia, Bulletin, 81, 54 pp. from Australian intracratonic basins. Geology, 15, Forman, D.J., 1972 - Petermann Ranges,Northern Tèrritory t06t-1064. - L:250000 Geological Series. Bureau of Mineral Maboko, M., 1988 - Metamorphic and geochronologicalevol- Resources,Australia, Explanatory Notes, SheetSG52-7. ution in the Musgrave Ranges, central Australia. Ph.D Tëleseismictravel-time anomalies and deepstructure of the northern and southern margins of the Basin 427 Thesis, Australian National University, Canberra (unpub- Shaq R., this volume - The tectonic development of the lished). Amadeus Basin. .Iz Korsch, R.J., & Kennard, J.M. (Edi- McAdoo, D.C., &. Sandwell,D.T., 1985 - Folding of oceanic tors), Geological and geophysicalstudies in the Amadeus lithosphere. Journal of Geophysical Research, 90, Basin, central Australia. Bureau of Mineral Resources,Aus- 8563-8569. tralia, Bulletin,236. McQueen, H.W.S., & Beaumont, C., 1989 - Mechanical - models of tilted block basins. 1z Price, R.A. (Editor), Shaq R.D., Etheridge, M.E., & Lambeck, K., in press Origin and evolution of sedimentary basins and their Development of the Late P¡oterozoic to Mid-Palaeozoic energy and mineral resources. American Geophysical intracratonic Amadeus Basin in central Australia: A key to Union, GeophysicalMonograph 48, 65-7 l. understandingtectonic forces in plate interiors. Tëctonics. - Sass,J.H., & Lachenbruch, 4.H., 1979 Thermal regime of Stephenson,R.S., & Lambeck, K., 1985- Isostaticresponse the Australian continental crnst. In McElhinny, M.W. of the lithosphere with in-plane stress:Application to cen- (Editor), The Earth: Its origin, structure and evolution. tral Australia. Journal of Geophysical Research, 90, AcademicPr¿ss, 30 I -35 1. 8581-8588. Shaw, R., 1987 - Basement uplift and basin subsidence in central Australia. Ph.D Thesis, Australian National Uni- Teyssier,C., 1985 - A crustal thrust systemin an intracratonic versity, Canberra (unpublished). environment. Journal of Structural Geology,7, 689-700.