New Zealand Journal of Geology and Geophysics, 1994
GEOLOGY:EARTHQUAKES tion due to the tropical climate in Singapore. The granite is usually overlai... more GEOLOGY:EARTHQUAKES tion due to the tropical climate in Singapore. The granite is usually overlain by a thick layer of residual soil. Laboratory tests indicated large reductions in the mechanical properties of the weathered granite. (from Authors)
The study reappraises the deep seismicity of the Tyrrhenian Sea. Careful examination of the quali... more The study reappraises the deep seismicity of the Tyrrhenian Sea. Careful examination of the quality of reported hypocentres shows that the earthquakes define a zone dipping NW, about 200 km along strike, 50 km thick, and reaching a depth of about 500 km. The zone is slightly concave to the NW at a depth of 300 km, but, contrary to many previous reports, is not tightly concave, nor are there significant spatial gaps in the seismicity, which is effectively continuous with depth. Seismicity is, however, concentrated in the depth interval 250–300 km, where the dip of the seismic zone changes from 70° (above 250 km) to a more gentle dip of 45° at greater depths. Seven fault-plane solutions are available for the largest earthquakes in this depth interval, all of them consistent with a P-axis down the dip of the seismic zone, and all of them requiring movement on faults out of the plane of the subducting slab.Two deep earthquakes near Naples lie well outside the main zone of activity; for one of which a fault-plane solution is available that has a P-axis not aligned with the dip of the seismic zone. The tightly concave slab-geometry favoured by other reports is supported mainly by the location of these events near Naples, which we think may represent deformation in a separate, probably shallower dipping, piece of subducted lithosphere.The lack of shallow seismicity, and particularly of thrust faulting earthquakes, at the surface projection of the Benioff zone suggests that active subduction has ceased. Estimates of the convergence rate responsible for subduction in the last 10 Myr far exceed the present convergence rate of Africa and Eurasia, suggesting that the subduction was related instead to the stretching and thinning of the crust in the Tyrrhenian Sea.
We have used body-wave modelling to determine the source parameters of 22 moderate to large earth... more We have used body-wave modelling to determine the source parameters of 22 moderate to large earthquakes that have occurred along the Hikurangi subduction margin and elsewhere in the North Island of New Zealand since 1964. We have also included events from the Harvard CMT catalogue since 1977 as that catalogue contains smaller events than can be modelled with body waves. We have found that shallow earthquakes occurring in the back-arc Taupo Volcanic Zone and its extension to the north show predominantly normal faulting with nodal planes parallel to the regional fabric and T -axes indicating extension in a direction in agreement with geodetic measurements. A normal faulting solution for the 1974 Opunake earthquake is compatible with the mapping of active faults in the offshore Taranaki region, and the initiation depth of 17 km is close to the expected brittle–ductile transition for that area. Earthquakes within the subducted plate at the southwestern end of the margin are dominated by down-dip tension that is rotated towards the deepest part of the subducting plate. Further north, the events in the upper part of the Wadati–Benioff zone (<200 km) show pure down-dip tension, while this association is less obvious for the deeper events, suggesting that the aseismic part of the subducting slab inferred from plate reconstructions has reached the 670 km discontinuity. Normal faulting earthquakes associated with plate bending below the region of interplate contact occur along the length of the margin, while shallow normal faulting events also occur in the trench, indicating that some slab pull propagates to that region. We infer that large interplate earthquakes are not imminent because in other similar areas they are preceded by outer-rise compressional events that have not been recorded in the region in historical times. Slip vectors of interplate thrust earthquakes indicate that the oblique plate motion across the margin is fully partitioned; that is, the plate interface accommodates very little transcurrent motion, which is instead accommodated by faults in the overlying Australian Plate. Extensional geodetic strain has been measured across the margin in the northeast, which has been regarded as unusual in view of the plate convergence. We suggest that the topography at this part of the margin is too steep to be supported by the current horizontal forces due to friction on the plate interface. This may have occurred due to a reduction in friction because of the subduction of seamounts and the subsequent entrapment of sediments, this also being consistent with the idea of underplating at this part of the margin, or due to the subduction of less buoyant crust. Other margins, such as Japan and Mid-America, where there is known to be subduction of sediments, also show extension in the forearc. Compressional strain observed across the margin in the southwest is more usual for a convergent margin, but it does not show the partitioning we see in focal mechanisms. These data might only agree after the long-term compressional strain accumulation has been averaged over a complete seismic cycle. Large historical and palaeo-earthquakes have occurred along the length of the margin associated with deformation in both the axial and the coastal ranges. In addition, along the length of the margin the rate of recent interplate thrust activity is low compared to margins that have a high proportion of aseismic slip. These factors indicate that the entire length of this margin is capable of producing large earthquakes.
Seismicity and fault-plane solutions show that the active deformation in the Adriatic region is v... more Seismicity and fault-plane solutions show that the active deformation in the Adriatic region is very varied. West of Messina, N-S shortening occurs with slip vectors representative of the overall Africa-Eurasia motion. Along the length of peninsular Italy, NE-SW extension on normal faults is the dominant style of deformation, but changes t o N-S shortening in N. Italy. Inland central and northern Yugoslavia is deforming on strike-slip and thrust faults, and an intense belt of NE-SW shortening continues south along the coast from central Yugoslavia into Albania. South of Albania the shortening in coastal regions is in a more easterly direction. The most remarkable feature of the region is the low level of seismicity in the Adriatic Sea itself, compared with the intense activity in the hgh topographic belts that border it on the SW, NW and NE. The relatively rigid behaviour of the Adriatic allows its motion relative t o Eurasia to be described by rotation about a pole in N. Italy. Anticlockwise rotation about this pole accounts, in a general way, for the change in style and orientation of the deformation in the circum-Adriatic belts. Historical and recent seismicity account for approximately equal rates of extension in central Italy and shortening in southern Yugoslavia of about 2 mm yr-' ; however, these are uncertain by at least a factor of two, and are anyway likely t o be underestimates of the true motion, because of the unknown contribution of aseisniic creep.
New Zealand Journal of Geology and Geophysics, 1989
Long-period body-wave modelling of the main shock of the 1987 March 2 Edgecumbe earthquake sequen... more Long-period body-wave modelling of the main shock of the 1987 March 2 Edgecumbe earthquake sequence reveals a multiple rupture event with a moment of 4. 3 × 10 N m. The first rupture originated at a depth of about 8 km and propagated to the surface and to the southwest. A second subevent occurred about 3 s later with an epicentre about 9 km to the southwest of the main-shock epicentre. The focal mechanism of the first subevent shows dominantly normal faulting. Comparison with the surface fault break indicates that the northeast-striking, northwest-dipping nodal plane was the fault plane and hence the horizontal slip vector trend was approximately 342°. The second subevent may correspond to the faulting observed on the Te Teko trace.
bThe plate motion model NUVEL-1 predicts oblique convergence between the Pacific and Australian p... more bThe plate motion model NUVEL-1 predicts oblique convergence between the Pacific and Australian plates in the South Island of New Zealand. We used P and SH body waveform analysis to constrain the focal mechanisms of the 15 largest earthquakes (Ms > 5.8) that have occurred in this region since 1964, in order to see how the plate motion is accommodated. At the southern end of the Alpine Fault, convergence is achieved by oblique slip movement along a concentrated zone of deformation. In the southern offshore region one event may be related to thrusting of the Australian plate beneath the Pacific plate, and another strike-slip event probably demonstrates movement on an active strike-slip fault system parallel to, but offset from, the southern limit of the Alpine Fault. This geometry provides a possible mechanism for the rapid uplift of the Fiordland region. Deformation in the northern South Island is more distributed. In the south-west Marlborough region partitioning occurs between strike-slip faulting in the SE and reverse faulting farther NW in the Buller region. We suggest that the partitioning developed as a consequence of an increasing component of shortening that was accommodated by slip on reactivated pre-existing normal faults in the Buller region. Shortening in the Buller region may have deflected the NE end of the Alpine Fault towards the NW, forming the prominent bend. The Marlborough Fault System, with its youngest and most active faults to the SE, probably developed in an attempt to maintain a through-going strike-slip structure as each of the strike-slip faults was transported towards the north-west. Partitioning of the opposite polarity (with reverse faulting SE of the strike-slip faulting) occurs in north-east Marlborough. The boundary between the two different styles of partitioning in NE and SW Marlborough appears to coincide with a change in the nature of the downgoing slab and a change in strike of faults of the Marlborough Fault System. A normal faulting earthquake on the northern edge of the Chatham rise probably results from a complex interaction of the buoyant continental crust in that region with the subduction zone and the overlying Marlborough Fault System.
... Aveleen Donohue, Mr. Ernesto Guereque, Miss Agnes Halasz, Mr. Klaus Kurschner, Miss Gertrude ... more ... Aveleen Donohue, Mr. Ernesto Guereque, Miss Agnes Halasz, Mr. Klaus Kurschner, Miss Gertrude Neumann, Miss Mary Lu Wilson and Mrs. Julia Wittenberg for technical assistance, to Dr. Virginia Mitty, of New York City, who referred the propositus to Dr. Helen Anderson of St. ...
], the details of which are described in . During this study, total chemistry and mass spectromet... more ], the details of which are described in . During this study, total chemistry and mass spectrometry blanks were 6 pg for Re and 2 pg for Os, with a 187 Os/ 188 Os blank ratio of 0.175. Blank corrections were insignificant for all samples analyzed in this study. Over a 3-year period, repeated analyses of an Os isotopic mass spectrometry standard provided by the Carnegie Institution of Washington, Department of Terrestrial Magnetism (DTM), made with Johnson-Matthey ammonium hexachlorosmate batch 5.56870-A, yielded mean 187 Os/ 188 Os ϭ 0.17367 Ϯ 0.00058 [external reproducibility at the 2 level (n ϭ 24) using a peak-jumping routine and a secondary electron multiplier], within error of the DTM value of 0.17429 Ϯ 0.00055 [S. B. Shirey, Can. J. Earth Sci. 34, 489 (1997)] (5). There is no measurable bias introduced into the data from our mass spectrometer (Finnigan MAT 262 N-TIMS) relative to results produced at DTM.
The Alpine fault is the Pacific-Australian plate boundary in the South Island of New Zealand. Thi... more The Alpine fault is the Pacific-Australian plate boundary in the South Island of New Zealand. This study analyzes 195 earthquakes recorded during the 6 month duration of the Southern Alps Passive Seismic Experiment (SAPSE) in 1995/1996 and two Mr. 5.0 earthquakes and aftershocks in 1997, which occurred close to the central part of the Alpine fault. Precise earthquake locations are derived by simultaneous inversion for hypocenter parameters, a one-dimensional velocity model, and station corrections. Together with focal mechanisms calculated using a first motion and amplitude ratio method, these results provide a picture of the seismotectonics in the central South Island over a 6 month period. Moment tensor inversions of three earthquakes provide an independent means of comparison to the focal mechanisms derived using the amplitude/first motion method. To validate our observations over time, we compare the SAPSE seismicity with the seismicity recorded by the New Zealand National Seismic Network (NZNSN) and a local network at Lake Pukaki east of the Southern Alps (6 months versus 8 years). Our study indicates that the Alpine fault releases elastic strain seismically from the surface down to 10-12 km depth between Milford Sound in the south and the Hope fault in the north. The seismicity rate of the Alpine fault is low but comparable to locked sections of the San Andreas fault, with large earthquakes expected. Seismicity decreases north of Bruce Bay at the Alpine fault and within a triangular region along the Alpine fault located between the Hope and Porters Pass fault zones. We interpret this as the result of deformation distributed on the Alpine fault and the Hope and Porters Pass fault zones. The base of the seismogenic zone is fairly uniform at 12 km _+ 2km over large parts of the South Island. The high Alps region has a shallower base of the seismogenic zone, indicating localized elevated temperatures east of the Alpine fault. Most of the study region deforms under a uniform stress field with a maximum principal horizontal shortening direction of 110ø-120 ø, similar to geodetic observations and plate motions. This confirms that the region is not undergoing strain partitioning. The earthquake data show that the deformation away from the Alpine fault is distributed on mainly NNE trending thrust faults and strike-slip transfer faults with a maximum seismogenic depth of 12km. 1995]. The thrust component in this region is accommodated by offshore structures and by a wide zone of crustal deformation extending east into central Otago [Norris et al., 1990]. Farther north it is a moderately eastward dipping oblique thrust fault [Berr)wzan et al., 1992; Norris and Cooper, 1995] with highest relative uplift rates at Paringa River [Simp, on etaL, 1993] and between Fox Glacier and Hare Mare River [Cooper and Norris, 1994; Wellman, 1979]. The region with highest uplift rates is connected •,•.ith a high thermal gradient [Allis and Shi, 1995; Koons, 1987a; Shi et al., 1996]. Thermal modeling [Allis' and Shi, 1995; Bart and Braun, 1999; Shi et al., 1996], data from fluid inclusion studies [Craw, 1988; Craw et al., 1994; Holm et al., 1989; JenAins et al., 1994], zircon reset ages [Tippett and Kamp, 1993] and heat flow measurements (R.H. Funnell and R.G. Allis, Thermal regime of the southeast South Island, New Zealand, submitted to New Zealand Journal of Geology and Geophysics, 2000) (hereinafter referred to as Funnell and Allis, submitted manuscript, 2000) all indicate •levated temperatures and predict a thermally weakened crust close to the Alpine fault. Allis and Shi [ 1995] point out that thermal models predict elevated temperatures near the surface, but temperatures below 20 km should be depressed as a result of the building of a crustal root beneath the Southern Alps. The estimated brittle-ductile transition /one range• from 4 to 12 km based on different models. This study provides important nee, constraints for the brittle-ductile transition zone from precise determination of earthquake depths. ß ". BROADBAND STATIONS ' ø., ß ' . [] SHORT PERIOD STATIONS .,,"' _•e ß . ß ,• 23 SHOTS ß ß ß 5461 EARTHQUAKES TRIGGlaRED' O 1997 MT COOK EQ + AFTE!•HO•K'6 "/d "!
Long-period seismic source parameters of the May 23, 1989, Macquarie Ridge earthquake are determi... more Long-period seismic source parameters of the May 23, 1989, Macquarie Ridge earthquake are determined using a surface wave inversion procedure that incorporates detailed source-time functions obtained from shorter-period body waves. The seismic source model obtained using this method is consistent with observations of both body waves and long-period Rayleigh waves from the earthquake. The Macquarie Ridge earthquake rupture has a centroid time (28 s) and a rightdateral strike-slip fault mechanism with a rake of 175 ø on a vertical hult plane striking N38øE. This mechanism is consistent with P wave first motions of the event.
New Zealand Journal of Geology and Geophysics, 1994
GEOLOGY:EARTHQUAKES tion due to the tropical climate in Singapore. The granite is usually overlai... more GEOLOGY:EARTHQUAKES tion due to the tropical climate in Singapore. The granite is usually overlain by a thick layer of residual soil. Laboratory tests indicated large reductions in the mechanical properties of the weathered granite. (from Authors)
The study reappraises the deep seismicity of the Tyrrhenian Sea. Careful examination of the quali... more The study reappraises the deep seismicity of the Tyrrhenian Sea. Careful examination of the quality of reported hypocentres shows that the earthquakes define a zone dipping NW, about 200 km along strike, 50 km thick, and reaching a depth of about 500 km. The zone is slightly concave to the NW at a depth of 300 km, but, contrary to many previous reports, is not tightly concave, nor are there significant spatial gaps in the seismicity, which is effectively continuous with depth. Seismicity is, however, concentrated in the depth interval 250–300 km, where the dip of the seismic zone changes from 70° (above 250 km) to a more gentle dip of 45° at greater depths. Seven fault-plane solutions are available for the largest earthquakes in this depth interval, all of them consistent with a P-axis down the dip of the seismic zone, and all of them requiring movement on faults out of the plane of the subducting slab.Two deep earthquakes near Naples lie well outside the main zone of activity; for one of which a fault-plane solution is available that has a P-axis not aligned with the dip of the seismic zone. The tightly concave slab-geometry favoured by other reports is supported mainly by the location of these events near Naples, which we think may represent deformation in a separate, probably shallower dipping, piece of subducted lithosphere.The lack of shallow seismicity, and particularly of thrust faulting earthquakes, at the surface projection of the Benioff zone suggests that active subduction has ceased. Estimates of the convergence rate responsible for subduction in the last 10 Myr far exceed the present convergence rate of Africa and Eurasia, suggesting that the subduction was related instead to the stretching and thinning of the crust in the Tyrrhenian Sea.
We have used body-wave modelling to determine the source parameters of 22 moderate to large earth... more We have used body-wave modelling to determine the source parameters of 22 moderate to large earthquakes that have occurred along the Hikurangi subduction margin and elsewhere in the North Island of New Zealand since 1964. We have also included events from the Harvard CMT catalogue since 1977 as that catalogue contains smaller events than can be modelled with body waves. We have found that shallow earthquakes occurring in the back-arc Taupo Volcanic Zone and its extension to the north show predominantly normal faulting with nodal planes parallel to the regional fabric and T -axes indicating extension in a direction in agreement with geodetic measurements. A normal faulting solution for the 1974 Opunake earthquake is compatible with the mapping of active faults in the offshore Taranaki region, and the initiation depth of 17 km is close to the expected brittle–ductile transition for that area. Earthquakes within the subducted plate at the southwestern end of the margin are dominated by down-dip tension that is rotated towards the deepest part of the subducting plate. Further north, the events in the upper part of the Wadati–Benioff zone (<200 km) show pure down-dip tension, while this association is less obvious for the deeper events, suggesting that the aseismic part of the subducting slab inferred from plate reconstructions has reached the 670 km discontinuity. Normal faulting earthquakes associated with plate bending below the region of interplate contact occur along the length of the margin, while shallow normal faulting events also occur in the trench, indicating that some slab pull propagates to that region. We infer that large interplate earthquakes are not imminent because in other similar areas they are preceded by outer-rise compressional events that have not been recorded in the region in historical times. Slip vectors of interplate thrust earthquakes indicate that the oblique plate motion across the margin is fully partitioned; that is, the plate interface accommodates very little transcurrent motion, which is instead accommodated by faults in the overlying Australian Plate. Extensional geodetic strain has been measured across the margin in the northeast, which has been regarded as unusual in view of the plate convergence. We suggest that the topography at this part of the margin is too steep to be supported by the current horizontal forces due to friction on the plate interface. This may have occurred due to a reduction in friction because of the subduction of seamounts and the subsequent entrapment of sediments, this also being consistent with the idea of underplating at this part of the margin, or due to the subduction of less buoyant crust. Other margins, such as Japan and Mid-America, where there is known to be subduction of sediments, also show extension in the forearc. Compressional strain observed across the margin in the southwest is more usual for a convergent margin, but it does not show the partitioning we see in focal mechanisms. These data might only agree after the long-term compressional strain accumulation has been averaged over a complete seismic cycle. Large historical and palaeo-earthquakes have occurred along the length of the margin associated with deformation in both the axial and the coastal ranges. In addition, along the length of the margin the rate of recent interplate thrust activity is low compared to margins that have a high proportion of aseismic slip. These factors indicate that the entire length of this margin is capable of producing large earthquakes.
Seismicity and fault-plane solutions show that the active deformation in the Adriatic region is v... more Seismicity and fault-plane solutions show that the active deformation in the Adriatic region is very varied. West of Messina, N-S shortening occurs with slip vectors representative of the overall Africa-Eurasia motion. Along the length of peninsular Italy, NE-SW extension on normal faults is the dominant style of deformation, but changes t o N-S shortening in N. Italy. Inland central and northern Yugoslavia is deforming on strike-slip and thrust faults, and an intense belt of NE-SW shortening continues south along the coast from central Yugoslavia into Albania. South of Albania the shortening in coastal regions is in a more easterly direction. The most remarkable feature of the region is the low level of seismicity in the Adriatic Sea itself, compared with the intense activity in the hgh topographic belts that border it on the SW, NW and NE. The relatively rigid behaviour of the Adriatic allows its motion relative t o Eurasia to be described by rotation about a pole in N. Italy. Anticlockwise rotation about this pole accounts, in a general way, for the change in style and orientation of the deformation in the circum-Adriatic belts. Historical and recent seismicity account for approximately equal rates of extension in central Italy and shortening in southern Yugoslavia of about 2 mm yr-' ; however, these are uncertain by at least a factor of two, and are anyway likely t o be underestimates of the true motion, because of the unknown contribution of aseisniic creep.
New Zealand Journal of Geology and Geophysics, 1989
Long-period body-wave modelling of the main shock of the 1987 March 2 Edgecumbe earthquake sequen... more Long-period body-wave modelling of the main shock of the 1987 March 2 Edgecumbe earthquake sequence reveals a multiple rupture event with a moment of 4. 3 × 10 N m. The first rupture originated at a depth of about 8 km and propagated to the surface and to the southwest. A second subevent occurred about 3 s later with an epicentre about 9 km to the southwest of the main-shock epicentre. The focal mechanism of the first subevent shows dominantly normal faulting. Comparison with the surface fault break indicates that the northeast-striking, northwest-dipping nodal plane was the fault plane and hence the horizontal slip vector trend was approximately 342°. The second subevent may correspond to the faulting observed on the Te Teko trace.
bThe plate motion model NUVEL-1 predicts oblique convergence between the Pacific and Australian p... more bThe plate motion model NUVEL-1 predicts oblique convergence between the Pacific and Australian plates in the South Island of New Zealand. We used P and SH body waveform analysis to constrain the focal mechanisms of the 15 largest earthquakes (Ms > 5.8) that have occurred in this region since 1964, in order to see how the plate motion is accommodated. At the southern end of the Alpine Fault, convergence is achieved by oblique slip movement along a concentrated zone of deformation. In the southern offshore region one event may be related to thrusting of the Australian plate beneath the Pacific plate, and another strike-slip event probably demonstrates movement on an active strike-slip fault system parallel to, but offset from, the southern limit of the Alpine Fault. This geometry provides a possible mechanism for the rapid uplift of the Fiordland region. Deformation in the northern South Island is more distributed. In the south-west Marlborough region partitioning occurs between strike-slip faulting in the SE and reverse faulting farther NW in the Buller region. We suggest that the partitioning developed as a consequence of an increasing component of shortening that was accommodated by slip on reactivated pre-existing normal faults in the Buller region. Shortening in the Buller region may have deflected the NE end of the Alpine Fault towards the NW, forming the prominent bend. The Marlborough Fault System, with its youngest and most active faults to the SE, probably developed in an attempt to maintain a through-going strike-slip structure as each of the strike-slip faults was transported towards the north-west. Partitioning of the opposite polarity (with reverse faulting SE of the strike-slip faulting) occurs in north-east Marlborough. The boundary between the two different styles of partitioning in NE and SW Marlborough appears to coincide with a change in the nature of the downgoing slab and a change in strike of faults of the Marlborough Fault System. A normal faulting earthquake on the northern edge of the Chatham rise probably results from a complex interaction of the buoyant continental crust in that region with the subduction zone and the overlying Marlborough Fault System.
... Aveleen Donohue, Mr. Ernesto Guereque, Miss Agnes Halasz, Mr. Klaus Kurschner, Miss Gertrude ... more ... Aveleen Donohue, Mr. Ernesto Guereque, Miss Agnes Halasz, Mr. Klaus Kurschner, Miss Gertrude Neumann, Miss Mary Lu Wilson and Mrs. Julia Wittenberg for technical assistance, to Dr. Virginia Mitty, of New York City, who referred the propositus to Dr. Helen Anderson of St. ...
], the details of which are described in . During this study, total chemistry and mass spectromet... more ], the details of which are described in . During this study, total chemistry and mass spectrometry blanks were 6 pg for Re and 2 pg for Os, with a 187 Os/ 188 Os blank ratio of 0.175. Blank corrections were insignificant for all samples analyzed in this study. Over a 3-year period, repeated analyses of an Os isotopic mass spectrometry standard provided by the Carnegie Institution of Washington, Department of Terrestrial Magnetism (DTM), made with Johnson-Matthey ammonium hexachlorosmate batch 5.56870-A, yielded mean 187 Os/ 188 Os ϭ 0.17367 Ϯ 0.00058 [external reproducibility at the 2 level (n ϭ 24) using a peak-jumping routine and a secondary electron multiplier], within error of the DTM value of 0.17429 Ϯ 0.00055 [S. B. Shirey, Can. J. Earth Sci. 34, 489 (1997)] (5). There is no measurable bias introduced into the data from our mass spectrometer (Finnigan MAT 262 N-TIMS) relative to results produced at DTM.
The Alpine fault is the Pacific-Australian plate boundary in the South Island of New Zealand. Thi... more The Alpine fault is the Pacific-Australian plate boundary in the South Island of New Zealand. This study analyzes 195 earthquakes recorded during the 6 month duration of the Southern Alps Passive Seismic Experiment (SAPSE) in 1995/1996 and two Mr. 5.0 earthquakes and aftershocks in 1997, which occurred close to the central part of the Alpine fault. Precise earthquake locations are derived by simultaneous inversion for hypocenter parameters, a one-dimensional velocity model, and station corrections. Together with focal mechanisms calculated using a first motion and amplitude ratio method, these results provide a picture of the seismotectonics in the central South Island over a 6 month period. Moment tensor inversions of three earthquakes provide an independent means of comparison to the focal mechanisms derived using the amplitude/first motion method. To validate our observations over time, we compare the SAPSE seismicity with the seismicity recorded by the New Zealand National Seismic Network (NZNSN) and a local network at Lake Pukaki east of the Southern Alps (6 months versus 8 years). Our study indicates that the Alpine fault releases elastic strain seismically from the surface down to 10-12 km depth between Milford Sound in the south and the Hope fault in the north. The seismicity rate of the Alpine fault is low but comparable to locked sections of the San Andreas fault, with large earthquakes expected. Seismicity decreases north of Bruce Bay at the Alpine fault and within a triangular region along the Alpine fault located between the Hope and Porters Pass fault zones. We interpret this as the result of deformation distributed on the Alpine fault and the Hope and Porters Pass fault zones. The base of the seismogenic zone is fairly uniform at 12 km _+ 2km over large parts of the South Island. The high Alps region has a shallower base of the seismogenic zone, indicating localized elevated temperatures east of the Alpine fault. Most of the study region deforms under a uniform stress field with a maximum principal horizontal shortening direction of 110ø-120 ø, similar to geodetic observations and plate motions. This confirms that the region is not undergoing strain partitioning. The earthquake data show that the deformation away from the Alpine fault is distributed on mainly NNE trending thrust faults and strike-slip transfer faults with a maximum seismogenic depth of 12km. 1995]. The thrust component in this region is accommodated by offshore structures and by a wide zone of crustal deformation extending east into central Otago [Norris et al., 1990]. Farther north it is a moderately eastward dipping oblique thrust fault [Berr)wzan et al., 1992; Norris and Cooper, 1995] with highest relative uplift rates at Paringa River [Simp, on etaL, 1993] and between Fox Glacier and Hare Mare River [Cooper and Norris, 1994; Wellman, 1979]. The region with highest uplift rates is connected •,•.ith a high thermal gradient [Allis and Shi, 1995; Koons, 1987a; Shi et al., 1996]. Thermal modeling [Allis' and Shi, 1995; Bart and Braun, 1999; Shi et al., 1996], data from fluid inclusion studies [Craw, 1988; Craw et al., 1994; Holm et al., 1989; JenAins et al., 1994], zircon reset ages [Tippett and Kamp, 1993] and heat flow measurements (R.H. Funnell and R.G. Allis, Thermal regime of the southeast South Island, New Zealand, submitted to New Zealand Journal of Geology and Geophysics, 2000) (hereinafter referred to as Funnell and Allis, submitted manuscript, 2000) all indicate •levated temperatures and predict a thermally weakened crust close to the Alpine fault. Allis and Shi [ 1995] point out that thermal models predict elevated temperatures near the surface, but temperatures below 20 km should be depressed as a result of the building of a crustal root beneath the Southern Alps. The estimated brittle-ductile transition /one range• from 4 to 12 km based on different models. This study provides important nee, constraints for the brittle-ductile transition zone from precise determination of earthquake depths. ß ". BROADBAND STATIONS ' ø., ß ' . [] SHORT PERIOD STATIONS .,,"' _•e ß . ß ,• 23 SHOTS ß ß ß 5461 EARTHQUAKES TRIGGlaRED' O 1997 MT COOK EQ + AFTE!•HO•K'6 "/d "!
Long-period seismic source parameters of the May 23, 1989, Macquarie Ridge earthquake are determi... more Long-period seismic source parameters of the May 23, 1989, Macquarie Ridge earthquake are determined using a surface wave inversion procedure that incorporates detailed source-time functions obtained from shorter-period body waves. The seismic source model obtained using this method is consistent with observations of both body waves and long-period Rayleigh waves from the earthquake. The Macquarie Ridge earthquake rupture has a centroid time (28 s) and a rightdateral strike-slip fault mechanism with a rake of 175 ø on a vertical hult plane striking N38øE. This mechanism is consistent with P wave first motions of the event.
Uploads
Papers by Helen Anderson