Abstract Rayleigh‐Taylor instabilities within the upper mantle are investigated as a mechanism for the uplift of rift‐margin mountain ranges within the continents. Our focus is on the Transantarctic Mountains (TAM). These mountains form at the margin between East and West Antarctica and have been free from the effects of subduction for about 500 million years. Lying behind and subparallel to the TAM is the Wilkes Subglacial Basin. The TAM–Wilkes Subglacial Basin pairing, with a characteristic wavelength of ∼500 km over a strike length of about 3000 km, is identified as one of the key physiographic signatures predicted by a migrating Rayleigh‐Taylor instability in the mantle. Such a persistent wavelength ( λ ) can be attributed to a maximum, and therefore dominant, growth rate for a Rayleigh‐Taylor instability given by λ ≈ πh, where h is the thickness of the layer that participates in the instability. 2D finite‐element analysis shows such an instability is plausible as it evolves from a step‐like displacement, initiated by extension or strike‐slip motion in the mantle lithosphere, and migrating laterally with time. Data from both Antarctica and New Zealand are synthesized to propose a rift and transform (sinistral strike‐slip) plate boundary within the Ross Embayment, which would have initiated the instability. A key finding of this study is that a 50 my gap between two dominant exhumation episodes in the TAM is common to other passive margins and is a natural outcome of the viscous, non‐linear, growth rate for a Rayleigh Taylor instability.

Continued ground uplift long after the drying out of the Aral Sea demonstrates that human activity can provoke a response deep inside our planet, in this case by causing rock in Earth’s mantle to flow.

Densely populated urban coastal strips are most at risk from the effects of relative sea-level rise. At the same time, anthropogenic activities associated with urbanisation, such as groundwater withdrawal, land reclamation and cut and fill operations can lead to local land subsidence (LLS), further exacerbating the risk to urban infrastructure. We generate the first high-resolution urban maps (10 m) of LLS using Sentinel-1 InSAR data between 2018 and 2021. This analysis reveals $77\percnt$77% of urban coastlines are subsiding at rates of ${-}0.5$-0.5 mm/yr or greater, and $10\percnt$10% are subsiding faster than $-3.0$-3.0 mm/yr. This analysis documents highly-localised hotspots of LLS, with subsidence rates exceeding $-10.0$-10.0 mm/yr in some cases. Highest subsidence rates occur on land that was reclaimed during the twentieth century, areas occupying Holocene sediment, or locations that were recently cut and filled. Time-series analysis of LLS at sites of reclaimed land shows both linear and non-linear rates of deformation over time periods of up to 6-8 years. This study reveals the spatial variability of exposure to sea-level rise hazard within New Zealand cities, and demonstrates that in many cases current rates of VLM should be expected to continue for the next few decades.

New mapping of the Barberton Greenstone Belt in South Africa shows that the central part is a pseudo-stratigraphy made of shallow-water and deep-water siliciclastic and volcanic slide blocks, with individual blocks ranging in size from tens of meters to >10 km in length. The outcrop pattern and scale are remarkably similar to those of large-scale Miocene to recent submarine landslides in New Zealand along the active Hikurangi subduction zone that are periodically triggered by earthquakes on the subduction megathrust, providing evidence for megathrust earthquakes in the Paleoarchean.

Along the Hikurangi Margin in New Zealand, the Pacific plate is obliquely (~45°) subducting beneath the Australian plate at ~40 mm/yr. The overlying deforming wedge has long term strain partitioning, with some trench-parallel plate motion focused on major dextral strike-slip faults in the back part of the deforming wedge. These faults pose a major seismic hazard, capable of rupturing as M6 to M8+ earthquakes. Here, I analyse GNSS measurements of the decadal interseismic strain rates in part of eastern North Island, where the northern Wellington Fault is the only major dextral strike-slip fault. The fault plane is subvertical, with a slip rate over the Holocene of 5 – 10 mm/yr. The interseismic deformation is well modelled by a simple 2-D elastic dislocation model in which the underlying subduction megathrust is fully locked at depths between 30 and 8 km. The creeping (or with repeated slow slip events) deeper and shallower parts of the megathrust slip at the full relative plate velocity. Importantly, the model contains no information about the slip rate and geometry of the major faults in the overlying deforming wedge, and it cannot predict these slip rates; the pattern of interseismic strain rates appears to be dictated solely by the locking on the megathrust. However, the strike-slip fault is precisely located where the interseismic shear strain rate is at its maximum, directly above the deeper locking line, and so locking on the megathrust appears to determine the pattern of long-term strain partitioning too. During a large magnitude earthquake on the northern Wellington Fault, the entire thickness of the wedge must rupture. The existence of several inactive parallel faults adjacent to the active trace suggests that the position of the megathrust locking line has moved over time. The paleoseismic record indicates that earthquakes on the Wellington Fault are independent of megathrust slip events. However, only about 30% of the component of trench-parallel relative plate motion is taken up on the northern Wellington Fault, so there must also be some oblique slip on the megathrust itself. The degree of partitioning is a consequence of the relative frequency of large magnitude earthquakes on the megathrust compared to those in the overlying deforming wedge, assuming constant stress drops for individual earthquakes. The interseismic megathrust locking must be a key factor here, because this controls the stress field at the point where a large magnitude earthquake is most likely to nucleate, and thus which faults will rupture.

Table of published paleomagnetic data used in this study.

Satellite-based measuring systems are making it possible to monitor deformation of the Earth's surface at a high spatial resolution over periods of several decades and a significant fraction of the seismic cycle. It is widely assumed that this short-term deformation directly reflects the long-term pattern of crustal deformation, although modified in detail by local elastic effects related to locking on individual faults. This way, short-term deformation is often jointly inverted with long-term estimates of fault slip rates, or even stress, over periods of 10 s to 100 s kyrs. Here, I examine the relation between these two timescales of deformation for subduction, continental shortening and rifting tectonic settings, with examples from the active New Zealand and Central Andean plate boundary zone. I show that the relation is inherently non-unique, and simple models of locking on a deep-seated megathrust or decollement, or mantle flow, provide excellent fits to the short-term observations without requiring any information about the geometry and rate of surface faulting. The short-term deformation, in these settings at least, cannot be used to determine the behaviour of individual faults, but instead places constraints on the forces that drive deformation. Thus, there is a fundamental difference between the stress loading and stress relief parts of the earthquake cycle, with failure determined by dynamical rather than kinematic constraints; the same stress loading can give rise to widely different modes of long-term deformation, depending on the strength and rheology of the deforming zone, and the role of gravitational stresses. The process of slip on networks of active faults may have an intermediate timescale of kyrs to 10 s kyrs, where individual faults fail piecemeal without any characteristic behaviour. Physics-based dynamical models of short-term deformation may be the best way to make full use of the increasing quality of this type of data in the future. This article is part of a discussion meeting issue 'Understanding earthquakes using the geological record'.

New passive- and active-source seismic experiments reveal unusually high mantle P-wave speeds that extend beneath the remnants of the world's largest known large igneous province, making up the 120-million-year-old Ontong-Java-Manihiki-Hikurangi Plateau. Sub-Moho Pn phases of ~8.8 ± 0.2 km/s are resolved with negligible azimuthal seismic anisotropy, but with strong radial anisotropy (~10%), characteristic of aggregates of olivine with an AG crystallographic fabric. These seismic results are the first in situ evidence for this fabric in the upper mantle. We show that its presence can be explained by isotropic horizontal dilation and vertical flattening due to late-stage gravitational collapse and spreading in the top 10 to 20 km of a depleted, mushroom-shaped, superplume head on a horizontal length scale of 1000 km or more. This way, it provides a seismic tool to track plumes long after the thermal effects have ceased.

The observed variations in the thickness of the conductive lithosphere, derived from surface wave studies, have a first‐order control on the elevation of the continents, in addition to variations in the thickness of the crust—this defines whole lithosphere isostasy (WLI). Negative buoyancy of the mantle lithosphere counters the positive buoyancy of the crust, and together, their respective thicknesses and density contrasts determine elevation of the continents both in their interiors and at their edges. The average density contrasts for lithospheric mantle with crust and with asthenosphere are typically 300 to 550 and 20 to 40 kg m−3, respectively, with a ratio 10 to 16, suggesting moderate average depletion of lithospheric mantle. We show that a crustal model for Antarctica, assuming WLI and using these density contrasts, provides a close fit to estimates of crustal thickness from surface wave tomography and gravity observations. We use a global model of WLI as a framework to assess factors controlling topography, showing that plausible regional variations in crustal and mantle densities, together with uncertainties in the crustal and conductive lithospheric thicknesses, are sufficient to account for global elevations without invoking dynamic topography greater than a few hundred meters. Estimates of elastic thickness Te in the continents are typically 25–50% of the thickness of the conductive lithosphere, indicating that the mantle part supports some of the elastic strength of the lithosphere.

Model outputs from Lamb, S., Moore, J., Perez-Gussinye, M., Stern, T. (2020). Global whole lithosphere isostasy: implications for surface elevations, structure, strength and densities of the continental lithosphere, Geochem, Geophys, Geosyst., doi :10.1029/2020GC009150 Data sets supplied here are the outcome of modelling described in the text. Files are given in either ASCII or GMT grd format. Data Set S1 (ds01.grd). Gridded crustal model of Antarctica based on whole lithosphere isostasy described in this study, and used to construct Figure 7c. Data columns are: x distance, y distance, crustal thickness. In GMT grd format with bounds in km -R-3000/3000/-3000/3000 -I5. Uses same projection as Bedmap 2 - see Fretwell et al. (2013) for details of projection. Suggested colour palette in GMT: seis -T0/60/2.5 -I Data Set S2 (ds02.xyz). Average elevation and crustal thickness of continental interiors calculated in this study, used to plot Figure 3c and described in text, using a standard lithospheric thickness of 100 km. Data columns are: Name, area, average elevation (m), average reduced elevation for 100 km standard lithosphere (m), average lithospheric thickness (km), average crustal thickness (km), 1 sigma uncertainty in elevation (m) or reduced elevation (m), 1 sigma uncertainty in lithospheric thickness (km), 1 sigma uncertainty in crustal thickness (km). ASCII file. Data Set S3 (ds03.grd). Gridded elevation anomalies (observed elevation – elevation calculated from whole lithosphere isostasy), as described in text and used to construct Figure 8. Data columns are: Longitude, Latitude, elevation anomaly (m). In GMT grd format with bounds in degrees -R-179/179/-60/80 -I1. Suggested colour palette in GMT: seis -T-2000/2000/200 -Z -I -M -D --COLOR_NAN=white Data Set S4 (ds04.grd). Gridded global crustal density perturbation model calculated to give zero elevation anomaly, based on elevation anomalies in Data Set 3, used to plot Fig. 9a. Data columns are: Longitude, Latitude, crustal density perturbation in kgm-3. In GMT grd format with bounds in degrees -R-179/179/-60/80 -I1. Suggested colour palette in GMT: seis -T-200/200/10 -Z -I -M -D --COLOR_NAN=white Data Set S5 (ds05.grd). Gridded global conductive lithosphere mantle density perturbation model calculated to give zero elevation anomaly, based on elevation anomalies in Data Set 3, used to plot Fig. 9b. Data columns are: Longitude, Latitude, mantle density perturbation in kgm-3. In GMT grd format with bounds in degrees -R-179/179/-60/80 -I1. Suggested clour palette in GMT: seis -T-50/50/5 -Z -I -M -D --COLOR_NAN=white Data Set S6 (ds06.grd). Gridded global crustal thickness perturbation model calculated to give zero elevation anomaly, based on elevation anomalies in Data Set 3, used to plot Fig. 9c. Data columns are: Longitude, Latitude, crustal thickness in km. In GMT grd format with bounds in degrees -R-179/179/-60/80 -I1. Suggested colour palette in GMT: seis -T-10/10/1 -Z -I -M -D --COLOR_NAN=white Data Set S7 (ds07.grd). Gridded global conductive lithosphere thickness perturbation model calculated to give zero elevation anomaly, based on elevation anomalies in Data Set 3, used to plot Fig. 9d. Data columns are: Longitude, Latitude, thickness in km. In GMT grd format with bounds in degrees -R-179/179/-60/80 -I1. Suggested colour palette in GMT: seis -T-100/100/5 -Z -I -M -D --COLOR_NAN=white Data Set S8 (ds08.grd). Compilation of gridded ratios of elastic thickness to conductive lithospheric thickness in the continents used to construct Figure 10c and d. Data columns are: Longitude, Latitude, ratio of elastic thickness to conductive lithospheric thickness from sources cited below. In GMT grd format with bounds in degrees -R-179/179/-60/80 -I1. Suggested colour palette in GMT: rainbow -T0/1/0.05 -Z -I -D --COLOR_NAN=white Data references: Lowry, A.R. and Pérez-Gussinyé, M., 2011. The role of crustal quartz in controlling Cordilleran deformation. Nature, 471(7338), 353-357. Pérez‐Gussinyé, M., Lowry, A.R., Watts, A.B. and Velicogna, I., (2004). On the recovery of effective elastic thickness using spectral methods: examples from synthetic data and from the Fennoscandian Shield. Journal of Geophysical Research: Solid Earth, 109(B10). Pérez-Gussinyé, M. and Watts, A.B., (2005). The long-term strength of Europe and its implications for plate-forming processes. Nature, 436(7049), 381. Pérez‐Gussinyé, M., Lowry, A.R. and Watts, A.B., (2007). Effective elastic thickness of South America and its implications for intracontinental deformation. Geochemistry, Geophysics, Geosystems, 8(5). Pérez‐Gussinyé, M., Lowry, A.R., Phipps Morgan, J. and Tassara, A., (2008). Effective elastic thickness variations along the Andean margin and their relationship to subduction geometry. Geochemistry, Geophysics, Geosystems, 9(2). Pérez-Gussinyé, M., Metois, M., Fernández, M., Vergés, J., Fullea, J. and Lowry, A.R., (2009). Effective elastic thickness of Africa and its relationship to other proxies for lithospheric structure and surface tectonics. Earth and Planetary Science Letters, 287(1-2), 152-167. Swain, C.J. and Kirby, J.F., 2006. An effective elastic thickness map of Australia from wavelet transforms of gravity and topography using Forsyth's method. Geophysical Research Letters, 33(2).