Tonga Slab Research Articles

Modeling Subduction With Extremely Fast Trench Retreat

AbstractThe Tonga‐Kermadec subduction zone exhibits the fastest observed trench retreat and convergence near its northern end. However, a paradox exists: despite the rapid trench retreat, the Tonga slab maintains a relatively steep dip angle above 400 km depth. The slab turns flat around 400 km, then steepening again until encountering a stagnant segment near 670 km. Despite its significance for understanding slab dynamics, no existing numerical model has successfully demonstrated how such a distinct slab morphology can be generated under the fast convergence. Here we run subduction models that successfully reproduce the slab geometries while incorporating the observed subduction rate. We use a hybrid velocity boundary condition, imposing velocities on the arc and subducting plate while allowing the overriding plate to respond freely. This approach is crucial for achieving a good match between the modeled and observed Tonga slab. The results explain how the detailed slab structure is highly sensitive to physical parameters including the seafloor age and the mantle viscosity. Notably, a nonlinear rheology, where dislocation creep reduces upper mantle viscosity under strong mantle flow, is essential. The weakened upper mantle allows for a faster slab sinking rate, which explains the large dip angle. Our findings highlight the utilizing rheological parameters that lead to extreme viscosity variations within numerical models to achieve an accurate representation of complex subduction systems like the Tonga‐Kermadec zone. Our study opens new avenues for further study of ocean‐ocean subduction systems, advancing our understanding of their role in shaping regional and global tectonics.

Geodynamic Evolution of the Lau Basin

AbstractThe formation of Lau Basin records an extreme event of plate tectonics, with the associated Tonga trench exhibiting the fastest retreat in the world (16 cm/yr). Yet paleogeographic reconstructions suggest that seafloor spreading in the Lau Basin only initiated around 6 Ma. This kinematics is difficult to reconcile with our present understanding of how subduction drives plate motions. Using numerical models, we propose that eastward migration of the Lau Ridge concurrent with trench retreat explains both the narrow width and thickened crust of the Lau Basin. To match the slab geometry and basin width along the Tonga‐Kermadec trench, our models suggest that fast trench retreat rate of 16 cm/yr might start ~15 Ma. Tonga slab rollback induced vigorous mantle flow underneath the South Fiji Basin which is driving the extension and thinning of the basin and contributing to its observed deeper bathymetry compared to neighboring basins.

Stress Variations in Southern Tonga Slab Derived From Deep‐Focus Earthquakes

AbstractTonga is a convergent plate boundary between the Pacific and Australian plates and is the fastest and the most seismically active deep subduction system in the world. We focused on southern Tonga (south of latitude 22°S) and the mantle transition zone (depths of 410–670 km), where seismic activity forms two subparallel bands of events in the east and west. We performed stress analysis by inverting focal mechanisms of earthquakes available in the Global Centroid Moment Tensor catalog and revealed two distinct stress regimes in the slab. While the stress orientation in the eastern slab segment conforms to the down‐dip compressional stress along the entire slab, the stress orientation in the western slab segment is different, having the maximum compression in the vertical direction. This suggests that the western segment can represent a stagnant slab with flattening and bending, as proposed by modeling studies. Its connection with the younger actively subducting slab is supported by the horizontal westward shift at 520 km depth. The stress analysis also indicates substantially different fault orientations in both segments. In the actively dipping slab, the majority of activated faults are predominantly sub‐horizontal. However, they are significantly inclined from vertical in the stagnant slab segment. A higher scatter in fault orientations in the stagnant slab suggests deformation, fragmentation and rheological complexity resulting from bending and flattening.

Toroidal flow around the Tonga slab moved the Samoan plume during the Pliocene

Abstract Age-progressive seamount tracks generated by lithospheric motion over a stationary mantle plume have long been used to reconstruct absolute plate motion (APM) models. However, the basis of these models requires the plumes to move significantly slower than the overriding lithosphere. When a plume interacts with a convergent or divergent plate boundary, it is often deflected within the strong local mantle flow fields associated with such regimes. Here, we examined the age progression and geometry of the Samoa hotspot track, focusing on lava flow samples dredged from the deep flanks of seamounts in order to best reconstruct when a given seamount was overlying the mantle plume (i.e., during the shield-building stage). The Samoan seamounts display an apparent local plate velocity of 7.8 cm/yr from 0 to 9 Ma, 11.1 cm/yr from 9 to 14 Ma, and 5.6 cm/yr from 14 to 24 Ma. Current fixed and mobile hotspot Pacific APM models cannot reproduce the geometry of the Samoa seamount track if a long-term fixed hotspot location, currently beneath the active Vailulu’u Seamount, is assumed. Rather, reconstruction of the eruptive locations of the Samoan seamounts using APM models indicates that the surface expression of the plume migrated ~2° northward in the Pliocene. Large-scale mantle flow beneath the Pacific Ocean Basin cannot explain this plume migration. Instead, the best explanation is that toroidal flow fields—generated by westward migration of the Tonga Trench and associated slab rollback—have deflected the conduit northward over the past 2–3 m.y. These observations provide novel constraints on the ways in which plume-trench interactions can alter hotspot track geometries.

2D stress rotation in the Tonga subduction region

Despite their enigmatic origins, deep earthquakes provide unique information about the stress distribution in slabs. Down-dip compression is expected at depth, due to resistance from endothermic phase transition near 660 km and increasing viscosity. Focal mechanisms for ordinary deep (620-680 km) earthquakes in the Tonga subduction zone exhibit down-dip compressional stresses in agreement with this expectation, but unusually deep (≥ 680 km) earthquakes exhibit vertical tension and horizontal compression. Here we present a numerical modelling study of the Tonga slab focused on stress patterns in the transition zone. The subducting Pacific plate is one of the oldest and thus coldest plates in the world, so our parametric study tests the effects of rheology and phase transitions on stress evolution for an exceptionally cold subducting plate. Our results suggest that direct buoyancy effects from the endothermic phase transition at 660 km depth are overprinted by bending-related forces. A stress pattern best fitting seismogenic stresses is found for the coldest tested subducting plate (150 Myr old) and a viscosity interface decoupled from the 660-km phase transition and shifted to 1000-km depth. An abrupt change of stress orientations is observed when the slab, temporarily deflected by the phase transition, penetrates to the shallow lower mantle and the fold of the flat-lying part is tightening.

Structure and dynamics of the Tonga subduction zone: New insight from P-wave anisotropic tomography

The Tonga-Lau-Fiji region is important to study plate-plume and subduction-ridge interactions, but its deep mantle structure is still not very clear. Here we present high-resolution tomography of 3-D P-wave azimuthal anisotropy down to 400 km depth of the Tonga subduction zone derived from arrival-time data of local earthquakes recorded at seafloor and land seismometers. The subducting Tong slab is imaged as high-velocity anomalies at depths of 100-400 km, whereas large-scale low-velocity anomalies down to 400 km depth are revealed in the mantle wedge beneath the backarc basin and volcanic arc. Trench-parallel anisotropy beneath the Lau Basin extends southwards to ∼140 km depth at ∼20.5°S, representing the extent of both southward flow of the Samoan plume and toroidal flow by the slab rollback. At depths of 140-400 km, the Lau Basin and Fiji Plateau mainly exhibit plate-parallel fast-velocity directions (FVDs) north of ∼20.5°S, indicating strong corner flow in the mantle wedge driven by the slab subduction and dehydration. The Tonga slab exhibits trench-parallel FVDs at depths of <200 km, reflecting fossil fabric formed during the plate spreading stage, whereas, at greater depths, the slab mainly exhibits trench-normal FVDs, which may reflect complicated deformations within the slab. These results suggest that the Samoan plume has a significant impact on the Tonga-Lau-Fiji region, leading to variations in the scale and depth extent of mantle flows.

Stress Drops of Intermediate‐Depth and Deep Earthquakes in the Tonga Slab

AbstractMultiple physical mechanisms have been proposed to explain the cause of intermediate‐depth and deep earthquakes, but they are still under debate. Source parameters such as stress drop, have the potential to provide insight into their physical mechanisms. We develop a modified spectral decomposition method to analyze 1‐year seismic data from temporary land‐based and ocean bottom seismographs in a complex subduction zone. By applying this method to investigate 1,083 intermediate‐depth and deep earthquakes in the Tonga slab, we successfully resolve the source spectra and stress drops of 743 MW 2.6–6.0 earthquakes. Although the absolute stress drops are subject to the choices of source model parameters, the relative stress drops are more reliably resolved. The median stress drop of Tonga earthquakes does not change with respect to magnitude but decreases with depth by 2–3 times in two separate depth ranges of 70–250 and 400–600 km, corresponding to intermediate‐depth and deep earthquakes, respectively. The median stress drops show spatial variations, with two high‐stress‐drop (five times higher than the ambient value) regions, coinciding with strong local deformation where the Tonga slab bends or tears. In the Tonga double seismic zone at 120–300 km depths, the median stress drop appears smaller in the lower plane than in the upper plane, suggesting a slower rupture velocity or a higher fluid content in the lower‐plane region. Our results suggest that intermediate‐depth and deep earthquakes in the Tonga slab generally follow the earthquake self‐similar model and favor the fluid‐related embrittlement hypothesis for both groups of earthquakes.

Tonga Slab Morphology and Stress Variations Controlled by a Relic Slab: Implications for Deep Earthquakes in the Tonga‐Fiji Region

AbstractDeep focus earthquakes within the Tonga‐Fiji subduction system account for about two‐thirds of the global total and provide significant constraints on slab deformation. The factors controlling the intense deformation remain unclear. Here, we use two‐dimensional, time‐dependent geodynamic models to study the morphology, stress state, and thermal structure of the Tonga‐Fiji subduction zone. The results, consistent with tomographic images and focal mechanisms, demonstrate that collision between a relic slab from the Vanuatu Trench and the Tonga slab may control the steeper dip of the Tonga slab and earthquakes in the mantle transition zone. We suggest that the magnitude 8.2 and 7.9 earthquakes in 2018 mostly ruptured within the warm rim of the Tonga slab and occurred beneath the folding relic slab with high temperatures of at least ∼900°C and ∼1100°C, respectively. The findings support the hypothesis that local slab temperature likely controls rupture of deep earthquakes.

The 2018 Fiji Mw 8.2 and 7.9 deep earthquakes: One doublet in two slabs

The cold Fiji-Tonga subduction zone accounts for >75% of cataloged deep earthquakes but none of the largest ten in the last century. On 19 August 2018 and 06 September 2018, a deep earthquake doublet with moment magnitude (Mw) 8.2 and 7.9 struck the Fiji area, providing a rare opportunity to interrogate the behaviors of great deep earthquakes in cold slabs. While the aftershock productivity of the 2018 Mw 8.2 event is similar to the 2013 Okhotsk Mw 8.3 event in a cold slab, the inferred compact rupture dimensions of both the Mw 8.2 and 7.9 events appear to be similar to the 1994 Bolivia earthquake in a warm slab. This seems to contradict the traditional view that slab temperature controls deep earthquakes. However, we find that neither event was confined within the cold Tonga slab core: the Mw 8.2 ruptured mostly in the warmer rim of the Tonga slab and the Mw 7.9 occurred in a warm relic slab leaning on top of the Tonga slab. The Fiji doublet demonstrates local slab temperature as the critical factor for deep earthquakes, and reveals complex interaction of subducted slabs in Tonga.

Seismic Evidence for Plume‐Slab Interaction by High‐Resolution Imaging of the 410‐km Discontinuity Under Tonga

AbstractThe Tonga‐Samoa system provides a unique tectonic context to study how a cold subducting slab interacts with a hot rising mantle plume. Here we present a 3‐D high‐resolution image of the 410‐km mantle discontinuity (the 410) using seismic signals excited by deep‐focus earthquakes. The 410 is found to be ~30 km shallower inside the Tonga slab relative to the ambient mantle and ~20 km deeper further to the northwest under Fiji Islands. The downward deflection of the 410 under Fiji supports the hypothesis of a plume migration around the northern edge of the Tonga slab from Samoan hot spot to under Fiji due to fast trench rollback. The 50‐km topography difference in the 410 between the plume and the slab corresponds to a temperature difference of ~500 ± 100 K. The Samoan plume is inferred to be 200 ± 50 K hotter than the ambient mantle and supports a thermal origin for the plume.

Analyses of anomalous amplitudes of antipodal PKIIKP waves

Approaching the distance of 180°, seismic focusing greatly amplifies the normally weak PKIIKP phase (underside reflection from the inner core boundary). Anomalously strong amplitudes of the PKIIKP phase reported previously at near antipodal distances (at seismic station TAM in North Africa) have been interpreted to infer anomalous structure(s) of the inner core boundary (including a sharp drop of compressional wave speed in the bottommost outer core or a near-zero shear wave speed in the topmost inner core). However, our observations of 12 earthquakes located antipodal to TAM (including the previously cited four events) suggest, for several reasons, that the anomalous PKIIKP energy might be a seismic phase misidentification. The anomalous phase appeared at distances less than 179.6° but not at larger distances (~179.8°). The phase appears consistently from antipode to distances less than 160° and has horizontal slowness similar to the PKIKP phase (going straight through the inner core). Its travel times vary greatly and show a systematic difference between two groups of events at different distances. A simple point scatter provides a good match to the travel times and the systematic variation of the anomalous phase at most stations, suggesting that it could originate from scattering off strong heterogeneities in the mantle wedge above the subducting Tonga slab. The phase misidentification suggests that the previously proposed inner core boundary structure(s) based on the anomalous phase need to be re-evaluated.

Gutenberg–Richter law for deep earthquakes revisited: A dual-mechanism hypothesis

Deep earthquake b values appear to vary with slab thermal state and earthquake magnitude. The physical reason for the variations and the relation with deep rupture mechanisms are still unclear. Here I confirm the spatial variations of b value and the dependence on slab temperature using about 40 yr of data from the Global Centroid Moment Tensor catalog. A new bimodal pattern is observed for the 500–700 km depth range: b is close to 1 in the cold Tonga slab, while in warmer slabs (e.g., South America, Japan–Kuril, Izu–Bonin–Mariana), b is close to 0.5 for intermediate magnitudes (Mw5.3–6.5) and increases to ∼1 for large magnitudes (Mw>6.5). To explain these observations, I propose a dual-mechanism hypothesis in which deep earthquakes nucleate only within the metastable olivine wedge (MOW), but can rupture outside MOW by a different mechanism. The fractal dimension of earthquake size distribution changes from 2 to 1 as the thermally controlled MOW thickness decreases, and back to 2 as the mechanism outside MOW dominates.

Upper mantle structure of the Tonga‐Lau‐Fiji region from Rayleigh wave tomography

AbstractWe investigate the upper mantle seismic structure in the Tonga‐Lau‐Fiji region by jointly fitting the phase velocities of Rayleigh waves from ambient‐noise and two‐plane‐wave tomography. The results suggest a wide low‐velocity zone beneath the Lau Basin, with a minimum SV‐velocity of about 3.7 ± 0.1 km/s, indicating upwelling hot asthenosphere with extensive partial melting. The variations of velocity anomalies along the Central and Eastern Lau Spreading Centers suggest varying mantle porosity filled with melt. In the north where the spreading centers are distant from the Tonga slab, the inferred melting commences at about 70 km depth, and forms an inclined zone in the mantle, dipping to the west away from the arc. This pattern suggests a passive decompression melting process supplied by the Australian plate mantle from the west. In the south, as the supply from the Australian mantle is impeded by the Lau Ridge lithosphere, flux melting controlled by water from the nearby slab dominates in the back‐arc. This source change results in the rapid transition in geochemistry and axial morphology along the spreading centers. The remnant Lau Ridge and the Fiji Plateau are characterized by a 60–80 km thick lithosphere underlain by a low‐velocity asthenosphere. Our results suggest the removal of the lithosphere of the northeastern Fiji Plateau‐Lau Ridge beneath the active Taveuni Volcano. Azimuthal anisotropy shows that the mantle flow direction rotates from trench‐perpendicular beneath Fiji to spreading‐perpendicular beneath the Lau Basin, which provides evidence for the southward flow of the mantle wedge and the Samoan plume.

Dynamic triggering of deep earthquakes within a fossil slab

AbstractThe 9 November 2009 Mw 7.3 Fiji deep earthquake is the largest event in a region west of the Tonga slab defined by scattered seismicity and velocity anomalies. The main shock rupture was compact, but the aftershocks were distributed along a linear feature at distances of up to 126 km. The aftershocks and some background seismicity define a sharp northern boundary to the zone of outboard earthquakes, extending westward toward the Vitiaz deep earthquake cluster. The northern earthquake lineament is geometrically similar to tectonic reconstructions of the relict Vitiaz subduction zone at 8–10 Ma, suggesting the earthquakes are occurring in the final portion of the slab subducted at the now inactive Vitiaz trench. A Coulomb stress change calculation suggests many of the aftershocks were dynamically triggered. We propose that fossil slabs contain material that is too warm for earthquake nucleation but may be near the critical stress susceptible to dynamic triggering.

Upper- and mid-mantle interaction between the Samoan plume and the Tonga–Kermadec slabs

Mantle plumes are thought to play a key role in transferring heat from the core–mantle boundary to the lithosphere, where it can significantly influence plate tectonics. On impinging on the lithosphere at spreading ridges or in intra-plate settings, mantle plumes may generate hotspots, large igneous provinces and hence considerable dynamic topography. However, the active role of mantle plumes on subducting slabs remains poorly understood. Here we show that the stagnation at 660 km and fastest trench retreat of the Tonga slab in Southwestern Pacific are consistent with an interaction with the Samoan plume and the Hikurangi plateau. Our findings are based on comparisons between 3D anisotropic tomography images and 3D petrological-thermo-mechanical models, which self-consistently explain several unique features of the Fiji–Tonga region. We identify four possible slip systems of bridgmanite in the lower mantle that reconcile the observed seismic anisotropy beneath the Tonga slab (VSH>VSV) with thermo-mechanical calculations.

Dominant role of temperature on deep earthquake mechanics for the Tonga slab near the bottom of the upper mantle

The effects of temperature on the mechanics of deep earthquakes are investigated based on detailed seismic images of a horizontal portion of the Tonga slab (19°S to 22°S). The hypocenter distribution of deep earthquakes and tomographic models have shown that the Tonga slab in this region stagnates laterally around the upper- and lower-mantle boundary. We analyze data from seismic networks in the United States and Japan from the deepest earthquakes that occurred in the stagnating part of the slab (focal depths 658 to 678 km). We observe a clear arrival approximately 8 s after the direct P wave and find that it is an S-to-P converted wave emitted upward from the deepest earthquake foci and reflected downward by a horizontal interface located nearly 30 km above the foci. The S wave speed drops upward across the interface by a few percent, so the interface is best interpreted as the former oceanic Moho in the stagnating Tonga slab. A numerical model of the thermal structure of the Tonga slab is constructed based upon the slab geometry, with the dynamics of the slab motion being incorporated. We find that if the deepest earthquakes occur in the coldest part of the slab, the vertical distance between the foci and the Moho should be approximately 30 km, which agrees well with the observations. The temperature around the foci is determined to be 750°C. We suggest that deep earthquakes are stringently restricted below this temperature, near the bottom of the upper mantle.

Concave slab out board of the Tonga subduction zone caused by opposite toroidal flows under the North Fiji Basin

Abstract An alternative scenario is proposed for the origin of a concave NE-facing slab under the North Fiji Basin between the Tonga subduction slab and Vanuatu Arc. During rollback of the Australian Plate, Vanuatu Arc rotated clockwise, whereas Fiji Platform rotated counterclockwise from 12/10 Ma until 1.5 Ma ago. Thereafter, only Vanuatu Arc rotated until the present. During the period of opposite rotations, toroidal flows entered the mantle around the northwest slab edge of Vanuatu Arc and from the northeast slab edge of Fiji Platform. The latter lies close to the northern end of the Tonga slab where arc-parallel flows are shown by volcanic geochemistry and mantle anisotropy. Opposite toroidal flows with upwelling and downwelling components generate the concave form of the combined Vanuatu/Fiji Platform slab, match its extent to mapped deep seismicity, explain its position overlying the Tonga slab, provide a mechanism for high heat flow in the North Fiji Basin as well as enriched MORB and OIB basalts in the northern NFB, and obviate slab collisions invoked to produce slab curvature.

Upper bounds of seismic anisotropy in the Tonga slab near deep earthquake foci and in the lower mantle

Upper bounds of seismic anisotropy in the Tonga slab near deep earthquake foci and in the lower mantle

Shear wave anisotropy in textured phase D and constraints on deep water recycling in subduction zones

Regions of low seismic velocity and high shear anisotropies in cold subducted slabs have often been related to anisotropic fabrics in hydrous phases mainly induced by slab deformation. The interpretation of these seismic anomalies in terms of hydration thus relies on a better knowledge of the elasticity and plastic deformation mechanisms of candidate hydrous phases. Here we investigate the development of lattice preferred orientations (LPO) in phase D [MgSi2H2O6, 10–18 wt% H2O], the ultimate water carrier in hydrous subducted peridotite. The samples were deformed non-hydrostatically up to 48 GPa in a diamond anvil cell and the texture and strength were obtained from analysis of the X-ray diffraction patterns collected in radial diffraction geometry. We find that at low strains the layered structure of phase D displays strong 0001 texture, where the stacking fault axis (c-axis) preferentially align parallel to the compression axis. A subsidiary 101¯0 texture develops at higher strains. Plasticity simulations in polycrystalline aggregates using a viscoplastic self-consistent model suggest that these LPO patterns are consistent with shape preferred orientation mechanism during the first compaction steps and, with dominant easy glide on basal planes and harder first order pyramidal slip, respectively, upon further compression. We find that phase D displays the lowest strength and the highest anisotropy among phases in hydrous peridotite in the uppermost lower mantle and might thus control the shear wave anisotropy generated in subducted slabs below the transition zone. We further evaluate the effect of textured phase D on the seismic velocity structure and shear wave anisotropy of deformed hydrous peridotite and compare the results to seismic observations in Tonga subduction. We show that 16 vol% of phase D in hydrous subducted peridotite is required to explain the negative velocity anomalies of 3%, the extent of shear wave splitting (0.9±0.3%) and the shear wave ray polarization geometry (VSH>VSV) observed in a detached fragment of the Tonga slab laying sub-horizontally below the transition zone. Seismic observations may thus place constraints on the degree of hydration of the Tonga slab beyond the transition zone, leading to the requirement that a minimum of 1.2 wt% H2O is retained in the slab by hydrous peridotite below 670 km depth.

Lower mantle seismic scatterers below the subducting Tonga slab: Evidence for slab entrainment of transition zone materials

We show evidence that materials with significantly different elastic properties are juxtaposed in the lower mantle immediately below the subducting Tonga slab (depths ⩽1000km), like the anomalies preferentially located beneath the lower mantle slabs at other Pacific subduction zones (Kaneshima, 2009). Array analyses of wave form data of short period seismic networks at western United States and Japan for deep earthquakes at the Tonga slab reveal S-to-P scatterers with a size less than the wavelengths (∼10km). The scatterers are located mostly outside of the slab by several tens of kilometers. Assuming a locally planar interface for the geometry of the scatterers, the amplitudes and polarities of the S-to-P waves are modeled to constrain the properties of the scatterers. We find that the scatterers are steeply dipping, the Vs increases oceanward across the interface, and the Vs contrasts are at least comparable to that associated with the post-spinel transformation (⩾6%). It is unclear at this stage what these subslab scatterers represent, so we discuss about three mechanisms which seem plausible from mantle dynamics viewpoints: (1) they may represent basaltic rocks which were emplaced by partial melting immediately beneath the former oceanic lithosphere–asthenosphere boundary (LAB) before the slab started subducting; (2) alternatively the elastic anomalies of the scatterers may be caused by localized presence of dehydrated water; or (3) the scatterers may correspond to a sharp boundary between fine-grained isotropic rocks in the immediate vicinity of the slab and coarse grained anisotropic rocks more distant from the slab. The presence of pronounced and localized elastic anomalies preferentially beneath the slabs in the shallow lower mantle, whatever its mechanism is, implies that a geophysically observable amount of transition zone material is entrained by the subducting slabs into the lower mantle.