Large Earthquakes in the Tonga Region Associated With Subduction of the Louisville Ridge

Abstract

Subduction of the Louisville Ridge influences both interplate and intraplate seismicity in the southern Tonga region (22°S–26°S). Five earthquakes with diverse mechanisms and seismic moments greater than 1 × 1027 dyn cm have occurred in this region since 1975. These include two outer rise, one underthrusting, and two intermediate depth events, all of which occurred in close proximity to the subducting ridge. In order to understand better the nature of these events and to determine the role played by the Louisville Ridge in the subduction process, detailed rupture models have been determined for each earthquake by body and surface wave analysis. Three events occurred at shallow depths near the oblique intersection of the Louisville Ridge and the Tonga are. These are the compressional (thrust) outer rise event of October 11, 1975 (Mw = 7.4); the tensional (normal) outer rise event of October 10, 1977 (Mw = 7.4); and the subsequent underthrusting event on December 19, 1982 (Mw = 7.5), the latter being the largest documented interplate event in Tonga. The intermediate depth events of June 22, 1977 (d = 70 km, Mw = 8.1) and April 13, 1980 (d = 160 km, Mw = 7.6) both have normal fault mechanisms; however; the 1977 event has a downdip tension axis, whereas the 1980 event has a downdip compression axis. The 1977 event is unusual due to its large size and the fact that most intermediate depth activity in Tonga involves downdip compression. The body wave radiation for the 1977 event indicates concentrated rupture in the depth range 70–90 km, near the downdip projection of the Louisville Ridge. It appears that the Louisville Ridge perturbs the regional stress regime at both intermediate and shallow depths in the slab. The ridge appears to be a buoyant feature that produces locally strong interplate coupling to the north of its junction with the Tonga trench, which accounts for the 1982 thrust event and the preceding 1975 compressional outer rise event located seaward of the thrust rupture zone. South of the ridge intersection the plate interface appears to be less strongly coupled: tensional stresses due to slab pull are communicated to shallow depths, resulting in the October 10, 1977 (Mw = 7.4), tensional outer rise event. The locally enhanced interplate coupling, along with a downdip stress barrier to the intraplate compressional stresses that arise from deep resistance to the slab subduction, are two effects of the ridge that appear to explain the accumulation of in‐plate tension in the region of the June 22, 1977, event, the April 13, 1980, downdip compressional event reflects a return to in‐plate compression following the great 1977 rupture which relaxed the tensional stress. The unusual stress regime of the subducting ridge and its oblique trajectory down the plate are thus responsible for the diversity of large earthquake occurrence in the southern Tonga arc.

Evaluation of far-field tsunami hazard in French Polynesia based on historical data and numerical simulations

“…The Tonga trench is an exception to this class as the Mw=8.4, 17 November 1865 earthquake is the only known event of this subduction zone to have generated a far-field tsunami observable without instruments . This view of the Tonga subduction is coherent with the analysis of other seismologists (Christiensen and Lay, 1988;Lundgren and Okal, 1988), and of earthquake catalogues which indicate that the vast majority of large earthquakes (Mw≥7) nucleate at depths greater than 30 km. However, the great Indian Ocean tsunami of 26 December 2004 has demonstrated that two centuries of data are not enough to evaluate the maximum expected magnitude on a major thrust fault system.…”

Tsunami risk assessment in the Marquesas Islands (French Polynesia) through numerical modeling of generic far-field events

“…but large thrusting events seem relatively rare (Christensen and Lay, 1988), possibly favouring a model where interplate coupling is low (Lundgren and Okal, 1988). In addition, large tsunamis initiated in this source area have not been reported in the Marquesas archipelago.…”

“…The key parameter controlling the final amplification seems to be the fault strike angle that defines the oceanic domain where the maximum tsunami energy will be trapped and scattered. In order to separate the effect of the strike from the effect of the location, we have finally performed several tests for the southernmost 25 • S location, a place where the Louisville Ridge subducts and where the seismicity and active faulting pattern is rather complicated (Christensen and Lay, 1988). For this reason, the strike angle (180 • ) used in the first modeling may not be the value really involved in future earthquakes; thus, we used Figure 10 displays the amplifications obtained in the four Marquesas bays, compared to amplifications computed for the 16 • S source.…”

The complex 3-D transition from continental crust to backarc magmatism and exhumed mantle in the Central Tyrrhenian basin

“…The association of the seismic gap with the ridge suggests that the stress state of the overriding plate has been altered by the subducting seamounts and their flexural moats (Bonnardot et al 2007). Three large earthquakes (M w = 7.4-7.5) located in close proximity to the subducting Louisville Ridge have been attributed to near-surface locking effects (Christensen & Lay 1988).…”