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 to 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 high topographic belts that border it on the SW, NW and NE. The relatively rigid behaviour of the Adriatic allows its motion relative to 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−1; however, these are uncertain by at least a factor of two, and are anyway likely to be underestimates of the true motion, because of the unknown contribution of aseismic creep. The Adriatic region resembles, in some ways, other relatively stable continental blocks, such as Central Iran and the Tarim Basin, that are caught up within the distributed deformation of the Alpine–Himalayan Belt. The Adriatic, however, is bounded on three sides by the relatively stable Eurasia plate. Its boundary with the African plate is short and ill‐defined by seismicity, but is likely to be located in the Southern Adriatic, near the Strait of Otranto. The present day seismicity shows that the Adriatic, although once perhaps ‘a promontory of Africa', is no longer behaving in this way, and the motions on its boundaries do not directly reflect the Africa–Eurasia convergence.

No abstract

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 st...

The upgrade of the New Zealand National Seismograph Network in the late 1980s has enabled more accurate earthquake locations to be determined. The catalogue data for events occurring from January 1990 until the end of February 1993 show some new patterns that have not been identified in previous observation periods, and also confirm the persistence of some phenomena observed previously, such as the aseismic corridor through the Nelson region. The deep seismicity data show spatial patterns remarkably similar to those for higher magnitude events recognised by Reyners in 1989. The Hikurangi Benioff zone is marked by intense seismic activity at depths between 150 and 200 km beneath the Central Volcanic Region; it has a sharp discontinuity beneath northwest Nelson and it extends as far southwest as Westport. The Fiordland Benioff zone is distinctly more seismically active in its northern block, and activity is noticeably concentrated in a zone to the west of Lake Te Anau.Shallow earthquakes (depth <15 km) for the period 1990 to February 1993 outline the active eastern boundary of the Central Volcanic Region and an east-west band running from Mt Ruapehu to Mt Taranaki. Earthquakes in this latter group also extend to deep crustal levels and they, and some recently recognised faults in this area, are probably related to a crustal discontinuity in this region. The Cape Egmont Fault Zone has been particularly active during this observation period. In the South Island, the 1990 Tennyson earthquake appears to have triggered activity along the Awatere Fault. The Alpine Fault is seismically quiet in the section from Harihari to Jackson Bay. A band of earthquakes lying to the east of the fault north of Harihari may represent activity associated with the Alpine Fault at depth although this cannot be confirmed by existing data. At the southern end of the Alpine Fault, two subparallel lineaments appear to form the boundaries of the western end of the Otago Range and Basin Province, but they are not associated with any known tectonic feature. In the Wellington region, the earthquakes shallower than 15 km are concentrated in the area between the Wellington and Wairarapa Faults. A group of earthquakes near Carterton possibly represents the interaction between two fault systems of different trends in

b The 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.

Seismic anisotropy and P-wave delays in New Zealand imply widespread deformation in the underlying mantle, not slip on a narrow fault zone, which is characteristic of plate boundaries in oceanic regions. Large magnitudes of shear-wave splitting and orientations of fast polarization parallel to the Alpine fault show that pervasive simple shear of the mantle lithosphere has accommodated the cumulative strike-slip plate motion. Variations in P-wave residuals across the Southern Alps rule out underthrusting of one slab of mantle lithosphere beneath another but permit continuous deformation of lithosphere shortened by about 100 kilometers since 6 to 7 million years ago.

No abstract

l'he 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 aP-axis not aligned with the dip of the seismic zone. The tightly concave slabgeometry 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.