Australia: The Land Where Time Began

A biography of the Australian continent 

Ross Ice Shelf Response to Climate Driven by Tectonic Imprint on Bathymetry of Sea Floor

Over the past 2 decades ice shelves in Antarctica have been thinned by ocean melting at an increasing rate which has led to a loss of grounded ice. It is indicated by geological records that the Ross Ice Shelf can rapidly disintegrate, which would accelerate the loss of grounded ice from catchments and this could produce equivalent to 11.6 m of global sea level rise, though the Ice Shelf is currently close to steady state. For this study data from the ROSETTA-Ice airborne survey and ocean simulations was used in order to identify the principal threats to the stability of the Ross Ice Shelf. Tinto et al. located the tectonic boundary between East and West Antarctica from magnetic anomalies and the use of gravity data to generate a new high resolution map of the bathymetry beneath the ice shelf. Sub-ice shelf circulation is constrained by the tectonic imprint on the bathymetry, which protects the grounding line of the ice shelf from moderate changes in the heat content of the global ocean. In contrast with this, local, seasonal production of upper ocean warm water near the ice front drives rapid melting of the ice shelf east of Ross Island, where more rapid loss of grounded ice would result from thinning of the ice sheets in West as well as East Antarctica. Tinto et al. confirm high melt rates in this region by the use of ROSETTA-Ice radar data. The significance of both the framework and local ocean-atmosphere exchange processes near the ice front in the determination of the future of the Antarctic Ice Sheet is highlighted by these findings.

Though the current contribution to sea level rise by the Antarctic Ice Sheet is small, it is the largest reservoir of potential global sea level rise, and it is the component with the greatest acceleration (Nerem et al., 2018; Shepherd et al., 2018). Rapid thinning that is driven by the ocean of small ice shelves that buttress (Dupont & Alley, 2005) the Amundsen Sea sector of the West Antarctic Ice Sheet (WAIS) (Pritchard et al., 2012), coincides with the largest recent losses of grounded ice (Gardner et al., 2018). The Ross, Filchner-Ronne and Amery ice shelves are the largest ice shelves, all 3 being in approximately steady state (Rignot et al., 2013; Depoorter et al., 2013), though they buttress grounded ice catchments that contain more than half of the entire potential Antarctic contribution to global sea level rise, which highlights a need to understand their stability for future climates that have been predicted.

The geology, glaciology and climatology of the Ross Embayment control the stability and structure of the Ross Ice Shelf, which has an area of ~480,000 km² (Depoorter et al., 2013). Convergent tectonics within ancient Gondwana (500 Ma) and the protracted breakup of this supercontinent (Dalziel & Lawver, 2001; Veevers, 2012) (190-70 Ma) were the origin of the regional geology and physiography, which produced the thinned, subsided lithosphere beneath West Antarctica that at present is adjacent to the thick lithosphere that supports the East Antarctic Ice Sheet (EAIS). Ice from 2 catchments: 1 from the WAIS, that has 2.0 m of potential sea level rise flowing as broad ice streams, and 1 from the EAIS, that has a potential of 9.6 m of sea level rise that flows as narrow glaciers through the Transantarctic Mountains, forms the Ross Ice Shelf that is typically a few hundred m thick. The ice takes something like 1,000 years to flow from the grounding line to the ice front. Currently, the ice shelf is stable (Paolo, Fricker & Padman, 2015) though repeated collapse of the ice shelf (Naish et al., 2009) is documented by geological evidence, as well as large-scale retreat of the grounding line near the margin of the continent at the Last Glacial Maximum (LGM) (Anderson et al., 2014) and more recent substantial changes in the position of the grounding line and the extent of the ice shelf during the Later Holocene (Yokoyama et al., 2016).

Mass is lost equally through basal melting and calving from the East Antarctic side of the Ross Ice Shelf, while on the west Antarctic side mass loss is dominated by calving (Rignot et al., 2013). Melt rates that are derived from satellite are close to zero for much of the shelf, though near the deep grounding lines of large EAIS glaciers as well as along the ice front it can exceed 2 m/yr (Rignot et al., 2013; Paolo, Fricker & Padman, 2015; Horgan et al., 2011; Moholdt, Padman & Fricker, 2014). The largest melt rates that have been observed are about 12 m/yr near the grounding line of Byrd Glacier (Kenneally & Hughes, 2004), and rates have been measured at 8 m/yr, though only in summer, close to Ross Island (Stewart et al., 2019). Delivery of heat to the base of the ice shelf is controlled by ocean circulation that is driven by winds over the ocean to the north of the ice front, exchanges of heat and freshwater at the surface of the ocean and tides (MacAyeal, 1984; Dinniman et al., 2018; Assmann, Hellmer & Beckmann, 2003).

Included in the ROSETTA-Ice project is a comprehensive airborne survey, with a line spacing to 10-20 Km, of the Ross Ice Shelf that was conducted during 2015-2017. The survey was designed to increase the resolution of seafloor bathymetry for ocean and ice sheet models, as well as to develop new insights the evolution of ice flow and tectonic development in the Ross Embayment. Tinto et al. used the Ice Pod instrument on a New York National Guard LC-130 aircraft in order to acquire gravity, magnetic, ice-penetration radar and laser altimetry data. These measurements, together with new ocean model simulations, revealed the interconnected systems that control the stability of the Ross Ice Shelf on timescales that range from months to millennia.

Geological Structure controls bathymetry beneath ice shelves

An abrupt transition in character across a boundary that is oriented approximately north-south through the centre of the embayment was shown by magnetic anomalies from the ROSETTA-Ice surveys. The West Antarctic side is dominated by high-amplitude anomalies and low amplitude anomalies dominate the East Antarctica side. Immature sedimentary rocks, magmatic ark materials and extended, thinned continental blocks make up the crust of West Antarctica (Siddoway, 2008). It is suggested by Tinto et al. that high amplitude magnetic anomalies could be due to ark magmatism during convergence of Gondwana or the exposure along faults during extension of highly magnetised metamorphic rocks (Luyendyk, Wilson & Siddoway, 2003). The crust of East Antarctica comprises ancient cratonic and orogenic material that has magnetic signatures that are highly variable, which includes a unit of low susceptibility that was identified within the Transantarctic Mountains (Goodge & Finn, 2010) with characteristics that are very similar to the East Antarctic side of the Ross Ice Shelf. Tinto et al. interpret the sharp boundary in magnetic character beneath the middle of the ice shelf, rather than the prominent Antarctic Mountains front (Dalziel & Lawver, 2001, as a marker of the position of the boundary between the crust of East and West Antarctica. There is no obvious boundary in the free air gravity anomaly map, but the difference in character was revealed in the density model that was gravity-derived. Tinto et al. modelled density by inverting the gravity anomaly at sites of known water depth from Ross Ice Shelf Geophysical and Glaciological Survey (RIGGS) measurements in order to show the relative variation in density of a column of rock of constant thickness across the region. The East Antarctic side, that is denser, reflects the thinner crust and a greater contribution from dense mantle material compared to the West Antarctic side. In order to attain the greater seabed depths that were observed on the East Antarctic side, the East Antarctic crust that was initially thick must have undergone a greater amount of extension than the West Antarctic crust. Tinto et al. suggest the different extensional histories of the 2 sides probably correspond to different underlying mantle properties. They interrupt the boundary that was identified in the middle of the Ross Ice Shelf, and extending to the margin of the continent, as the major tectonic boundary between East and West Antarctica.

There is an imprint in the bathymetry beneath the ice shelf of the tectonic boundary, which was revealed in the new bathymetry map that was developed by Tinto et al. through the inversion of the ROSETTA-Ice gravity anomaly field, by use of the density distribution the RIGGS-constrained model that was described above. It has been found that the bathymetry beneath the shelf is typically deeper on the East Antarctic side (670 m mean) and shallower on the West Antarctic side (560 m mean). It is indicated by the fact that large-scale asymmetry in bathymetry coincides with the tectonic boundary, that the asymmetry is a long-term feature that has persisted throughout multiple glacial cycles. The new bathymetry model of Tinto et al. resolves the smaller-scale features that were not present in prior grids, especially close to the grounding line where the new bathymetry is deeper near the Kamb Ice Stream and along the grounding line in the EAIS to the south of Byrd Glacier.

Ocean circulation and basal melting constrained by bathymetry

Tinto et al. ran an ocean circulation model that incorporated the new bathymetry and an updated ice draft. The large-scale patterns of circulation that is modelled, distribution of water mass and melt rate are similar to previous results (Dinniman et al., 2018; Assmann, Hellmer & Beckmann, 2003), though we now know better represents flows into the grounding zones of major outlet glaciers in East Antarctica. Tinto et al. used dyes to track the flow and modification of water masses from deep ocean to the north of the continental margin to the grounding line of the ice shelf. Antarctic Surface Water (AASW), modified Circumpolar Deep Water (mCDW), Ice Shelf Water (ISW) and High Salinity Shelf Water (HSSW) (Orsi & Wiederwohl, 2009) are the principal water masses present along the ice front. The distributions of these water masses vary on a seasonal basis, has been reported previously (Assmann, Hellmer & Beckmann, 2003; Jendersie et al., ?). The simulation identified the relative importance to ice shelf melting of CDW, which is part of the global thermohaline circulation, and HSSW, which is formed locally in polynyas.

Beneath the ice shelf the dominant inflow by volumes is HSSW that flows beneath the ice front near Ross Island, then moving to the south along the base of the Transantarctic Mountains. This water is responsible for high rates of melting at deep grounding lines of major glaciers in EAIS including the Byrd Glacier, even though HSSW is at the surface freezing temperature  (about -1.9^oC), due to the suppression by pressure of the freezing temperature. Tinto et al. found that the mixture of HSSW and ISW does not cross the tectonic boundary because of dynamic constraints that are imposed by the thinner water column on the West Antarctic side. Instead, it continues flowing to the north and exits the ice shelf cavity in the vicinity of Glomar Challenger Trough.

Hayes Bank steers a subsurface layer of mCDW southwards across the continental shelf to the ice font. It is shown by the simulation of Tinto et al. that some mCDW circulation and basal melting beneath the ice shelf to the west of Roosevelt Island. The penetration of the mCDW is limited, however, to a region within about 100 km of the ice shelf front. Further to the south, the WAIS side of the ice shelf is isolated from the source of oceanic heat and is dominated by a sluggish pool of very cold ISW, and this leads to negligible rates of melting at that location.

Modelled melt rates that are relatively high along the ice front are consistent with estimates based on satellite data (Horgan et al., 2011; Moholdt, Padman & Fricker, 2014). Rapid melting in summer due to warmer inflows of mCDW along Hayes Bank, as well as the presence of AASW that has been seasonally warmed along the ice front, dominate annual-average rates. The highest seasonal melt rates are located on the EAIS side close to Ross Island where flows of AASW under the ice shelf are permitted by thinner ice at the front (Stewart et al., 2019; Assmann, Hellmer & Beckmann, 2003; Stern et al., 2013).

Radar observation of basal melt near the ice shelf front

Tinto et al. used cross-sections of ice shelf vertical structure from the ROSETTA-Ice Shallow Ice Radar in order to identify thinning along the flowlines of East Antarctica, which provided a direct measurement of changing ice thickness, which was interpreted as basal melt, averaged over timescales of decades to centuries. The internal boundary between the lower layer of ice that formed on the continent, and younger ice that formed from snowfall onto the ice shelf was identified by the radar. Along the Mulock Glacier flowline the continental ice layer thins by more than 75 m over a distance of 40 km over a period of ~82 years, to reach 0 thickness about 50 km south of the ice front. Based on these observations, the steady thinning rate for this layer over the last ~80 years is 0.9 m/year. Change in thickness is a combination of basal mass balance and ice divergence. It is suggested by strain rates calculated from a satellite-derived ice velocity field (Moholdt, Padman & Fricker, 2014) that in this region compressive flow causes thickening of this layer of 0.33 m/yr. A basal melt rate of 1.23 m/yr is found by applying this strain correction to the observed thinning rate, which matches the 1.2 ± 0.2 m/yr basal melt rate that was derived from satellite altimetry (Moholdt, Padman & Fricker, 2014). It is suggested by the close match between the method of Tinto et al. and the satellite altimetry result that the processes that are currently melting the EAIS ice near Ross Island have persisted throughout the last century.

Future vulnerability and past ice sheet processes

It is indicated by the results of the study by Tinto et al. that the asymmetry in the bathymetry that is controlled by tectonics will prevent melt rates at the grounding line from changing substantially for future moderate change in climate, which agrees with Dinniman et al. (Dinniman et al., 2018). In this case melt rates will remain high, though stable, at the deep grounding lines of EAIS glaciers, as they are controlled by HSSW for which temperature remains constant (about -1.9^oC) and whose circulation is controlled strongly by bathymetry. Near grounding lines of the WAIS melt rates will remain low because the large-scale circulation accumulates very cold meltwater in this region, and the thinness of the cavity on the West Antarctic side of the tectonic boundary provides a strong dynamic barrier to incursions of the global ocean heat from the mCDW inflows.

The primary sensitivity of the mass balance of the Ross Ice Shelf will be, in the near term, to variations in local climate that change melt rates near the ice front (Stewart et al., 2019; Assmann, Hellmer & Beckmann, 2003; Stern et al., 2013). Tinto et al. suggest that changes in frontal melt may be driven by changes in the amount of mCDW flowing south across the continental shelf along Hayes Bank, and by variable production of AASW in summer. The mCDW heat flux depends on large-scale climate processes that determine the rate at which CDW (with temperature greater than 0^OC) is forced onto the continental shelf and subsequent loss of heat from the mCDW by mixing and convection of the upper ocean in winter. The production rate and properties of AASW are influenced strongly by conditions of the local sea ice, influx of freshwater from the Amundsen Sea (Jacobs & Giulivi, 2010) which influences heat content and stability of the upper ocean, and the net atmospheric heat flux (Schneider & Reusch, 2016).

The effect on the stability of ice sheets of changing ice shelf melt rates depends on the local contribution of the ice to net buttressing of the grounded ice flow. ‘Passive shelf ice’ comprises most of the ice front. Any loss of ice from this region will have only a small effect on the acceleration of grounded ice flow. Contrasting with this, ice shelf thinning or retreat near Ross Island will reduce the buttressing of nearby glaciers in EAIS as well as the ice streams of WAIS that are more distant (Reese et al., 2018). Tinto et al. proposed that the grounded ice catchments around Ross Embayment are most vulnerable to ice shelf loss near the ice shelf front around Ross Island and Minna Bluff, due specifically to increased duration and intensity of the production in summer of warm AASW and its subsequent flow beneath the ice shelf (Stewart et al., 2019; Dinniman et al., 2018; Assmann, Hellmer & Beckmann, 2003; Stern et al., 2013).

The role of climate variations on longer timescales in the destabilisation of the Ross Ice Shelf of will depended on the position of the ice front. The grounding line of the ice sheet during the Last Glacial Maximum (LGM) was near the edge of the continental shelf (Anderson et al., 2014). Water masses that were formed locally in this configuration are likely to have played a lesser role as the globally controlled, CDW that was relatively warm, could flow into the cavity beneath the ice, thereby generating high melt rates at the grounding line similar to those that are observed at present in the Amundsen Sea (Dutrieux et al., 2014). The ability of the wind-forced ice-front polynyas, during the subsequent retreat of the ice sheet, to produce colder HSSW would have been established. The Ross Ice Shelf system would have shifted to a locally controlled, cold, sub-ice cavity as the HSSW filled the ice cavity (MacAyeal, 1984). This switch from global to local controls should have been preserved in the geological record of former ice shelf extent, including existing sediment cores (Naish et al., 2009).

According to Tinto et al. following the Last Glacial Maximum, as the grounded ice retreated, the ice sheets in East and West Antarctica would have responded differently to the bathymetry on either side of the tectonic boundary. To the north of the modern ice shelf, in the Ross Sea, extensive sediments from the Cainozoic buried the bathymetric expression of the boundary and so would not have had direct influence on the retreat of the grounding line across this region. Instead, the bathymetry of this region has been sculpted by glacial deposition and erosion (Anderson et al., 2014). The bathymetry beneath the Ross Ice Shelf of the present reflects clearly the tectonic boundary. The rapid retreat of the grounding line, which is inferred on the Eastern Side (Spector et al., 2017), would have been aided by the deep bathymetric troughs connected to the Nimrod and Byrd glaciers. Slower ice sheet retreat will have been experienced by the shallow West Antarctic side, as it has pinning points such as Roosevelt Island, Steershead Ice Rise and Crary Ice Rise.

Different boundary conditions are also introduced by the contrasting properties of the crust across the tectonic boundary, modulating the flow of grounded ice from East and West Antarctica during prior glacial epochs. The production and localisation of geothermal heat flux are controlled by crustal properties as well as influencing the isostatic response of the lithosphere to ice loading and unloading. A key role is played by the response to changing ice load in the WAIS grounding line history around Crary Ice Rise (Kingslake et al., 2018), which lies on the boundary between East and West Antarctic crust. As the ice sheets retreated across different material on either side of the tectonic boundary they will have produced different isostatic responses.

It has been shown by the results of the survey by Tinto et al. that the bathymetry and basal boundary conditions beneath the Ross Ice Shelf have a tectonic origin which indicates that the contrasting conditions under the WAIS and EAIS sectors have endured through the entire glacial history of the Ross embayment. Beneath the Ross Ice Shelf of the present, the shape of the sub-ice-shelf cavity is controlled by the newly identified tectonic boundary, which enables circulation that insulates the groundling line from the influence of global ocean heat. Tinto et al. have identified that for the East and West ice sheets in the Ross Sea sector that the greatest vulnerability is to local, seasonal, upper ocean warming and deepening of the surface layer at a key region of the ice front, near Ross Island. The need to incorporate the ice shelf response to local climate processes in large-scale predictions of ice sheet behaviour in the broader tectonic framework.

Sources & Further reading

Tinto, K. J., et al. (2019). "Ross Ice Shelf response to climate driven by the tectonic imprint on seafloor bathymetry." Nature Geoscience 12(6): 441-449.