On the role of ice shelves on ice sheet mass balance
The role of ice shelves in buttressing outflow from ice sheets into an ocean has remained a topic of controversy for several decades, mostly due to a lack of observations. With the collapse of Larsen A in 1995 and B in 2002, we observed the glacier response in more details than ever before. The results published by several authors revealed widespread acceleration, of surge type, with no return to the initial conditions. Dryglaski glacier is still accelerating in 2005 after speeding up by 300 percent. Evans/Hektoria/Green glaciers are slowing down but remain 500 percent out of balance. Crane glacier continues to accelerate with an ice front 10km upstream of its 1996 grounding line, but Leppard and Flask glaciers, which are still buttressed by an ice shelf, have remained steady. Ice shelf collapse has therefore a dramatic influence on glacier mass balance. Further south, Fleming Glaciers and others are 80 percent out of balance following the collapse of Wordie Ice Shelf in the 1980s, and maintaining steady speeds. Deep south, Pine Island Glacier accelerated 38 percent from 1974-2005, and its yearly acceleration rate increased fivefold. Thwaites is widening and could double its width if the eastern ice shelf cease to buttress its outflow as suggested by the appearance of new rifts at its grounding line between 1992 and 2005. Smith and Kohler glaciers, buttressed by Crosson/Dotson ice shelves are thinning rapidly and accelerated 10-50 percent in 1996-2005. The retreat/thinning is due to ice shelf thinning caused by a warmer ocean which triggered rapid grounding line retreat, reduction in buttressing force, and ice flow acceleration. A thinning ice shelf from a warmer ocean can therefore have a dramatic influence on ice sheet mass balance well before it breaks up. Add up sufficient atmospheric warming to produce melt water, and the ice shelves will collapse and precipitate the glaciers out of balance. The future of ice shelves will control the mass balance of the Antarctic Ice Sheet.
Ice shelf tides: Modeling, detection and removal
The ocean tide is the dominant source of vertical motion for the floating ice shelves surrounding Antarctica, the most important cause of sub-ice-shelf ocean mixing and basal melt, and a significant contributor to the gravity signal detected by the GRACE satellite mission. For applications such as long-term trend detection from satellite altimetry and gravity, accurate removal of the tide is an essential processing step. However, tide model accuracy for Antarctic regions is lower than for mid-latitude oceans, for two reasons: (1) basic model inputs such as coastline/grounding-line locations, water depth, and sub-ice-shelf water column thickness are poorly known; and (2) there is no high quality data set of sea surface (and ice shelf) elevation comparable to the TOPEX/Poseidon radar altimetry available for the ice-free ocean equatorward of 66 degrees. In this talk we review recent efforts to identify the optimum tide model from presently available models, then discuss methods for further improving the accuracy of Antarctic tide models through assimilation of multiple data types including satellite altimetry from the ERS and ICESat missions and in situ data from ice shelf GPS and ocean moorings. We conclude with a discussion of progress towards prediction of tidal displacements of the ice shelf within the grounding zone (GZ), an effort which depends on improving maps of the grounding line location and GZ width and decreasing tide model grid spacing to order 1 km.
Modelling Ice Shelf Water flow beneath Filchner-Ronne Ice Shelf
A two-dimensional depth-averaged plume model that incorporates Coriolis forces is used to study the flow and thermodynamics of Ice Shelf Water (ISW) under Filchner-Ronne Ice Shelf (FRIS), Antarctica. Buoyant ISW plumes are generated when meltwater is released near the grounding line of large ice shelves. Due to the pressure variation of seawater's freezing temperature, they may become supercooled as they ascend along the underside of an ice shelf, causing frazil ice crystals to form and multiply within the plume. As these frazil crystals are buoyant, they can precipitate onto the ice shelf base to form marine ice (in conjunction with direct freezing). The primary aim of this study is to model the plumes responsible for the thick (>400m) deposits of marine ice observed at the base of FRIS. Results are presented which show both the general effect of Coriolis forces on ISW plumes and the particular importance of rotation on plume flow beneath FRIS.
New Insight into Ice Shelf Rift Propagation from Geodetic and Seismic Monitoring
Rifts in Antarctic ice shelves are large through-cutting fractures that penetrate the entire ice thickness. These rifts can grow to be hundreds of kilometers long, eventually forming the boundary from which large tabular bergs detach. Despite the important role that iceberg calving plays in the mass balance of the Antarctic ice sheet (icebergs account for up to two thirds of the mass loss), very little is known about the forces involved in their initiation and subsequent propagation. During the 2002-2003 and 2004-2005 seasons we jointly deployed arrays of GPS and seismometers around the tip of an actively propagating rift on the Amery Ice Shelf, East Antarctica. Our observations show strong clustering of seismicity along the rift axis, extending far ahead of where the rift tip is visible on the surface. We also find episodic swarms of seismicity accompanied by rapid rift widening, which we interpret as bursts of rift propagation. The locations of events during the seismic swarm show that during each burst, the rift propagated approximately 100-200 meters. Previous studies have shown no direct triggering of bursts of propagation by tides or winds. Serendipitously, during the 2004-5 our instruments were deployed one week before the magnitude 9.3 Sumatra earthquake. Not only is the earthquake clearly visible in our seismic records, but we also see the arrival of T-waves (acoustic waves which propagate through the ocean) as well as the tsunami triggered by the earthquake. This presents us with a novel opportunity to study the influence of both the earthquake and the tsunami on rift propagation. We present preliminary results showing that neither the earthquake nor the T-waves had any effect on propagation. However, one of the bursts occurs several hours after the tsunami arrives at the ice shelf, suggesting a possible connection and raising questions about the potential influence of large storms and swell on propagation.
A Blowing Snow Model for Ice Shelf Rifts
Ice mélange (a mixture of snow, marine ice, and ice talus) may play various roles in the rates of propagation of iceberg-calving rifts through Antarctic ice shelves. This modeling study examines the role of windblown snow in the formation and maintenance of ice mélange in the "nascent rift" in the Ross Ice Shelf (78 08'S, 178 29'W). The rift axis is perpendicular to the regional wind direction, allowing us to employ a two-dimensional blowing snow model. The Piektuk-Tuvaq blowing snow model (Dery and Tremblay, 2004) adapted the Piektuk blowing snow model for use in sea ice environments by including parameterization for open-water leads within the sea ice. This version of the model was used to study the initial conditions of a freshly-opened rift, as the input of blowing snow into the seawater within the rift promotes marine ice formation by cooling and freshening the surface water. We adapted the Piektuk-Tuvaq model both for the local climatic conditions and to incorporate the geometry of the rift, which is 30m deep and 100m wide (far deeper than a lead). We present the evolution of the topography within the rift for two cases. The first is an ice mélange composed exclusively of snow and marine ice, the second uses an initial topography including large chunks of ice talus.
Monitoring Iceberg B15a With Airborne Radar Sounding
Understanding the migration, ablation and disintegration of large tabular icebergs requires knowledge of their thickness and basal character over time. In 2000, the ~5000 square km iceburg B15a calved off the western edge of the Ross Ice Shelf. The University of Texas, Institute for Geophysics (UTIG) first acquired phase-coherent radar imaging of iceberg B15a along with laser altimetry for accurate surface elevations during the early austral summer of 2001/02. At that time, the 20-month old iceberg was in one piece and was lying nearly perpendicular to the front of the Ross Ice Shelf north of Ross Island. A 140 km profile was flown parallel to and about 10 km from the then-recent ice shelf break, while an adjacent 160 km profile bisected the iceberg. These two profiles were connected with two shorter profiles (approximately 35 and 45 km) running perpendicular from the shelf break to the former shelf edge. Our objective of simultaneously imaging high-resolution surface and bottom scattering as well as the extremely subtle englacial layering was achieved. Ice thickness and internal layering implied B15a was subject to heterogeneous basal melting prior to calving from the Ross Ice Shelf. Also, major vertical cracking extending from the surface to the base was identified at only a single site (where the iceberg subsequently split) whereas basal cracking was pervasive over approximately two thirds of the iceberg base. On the two shorter profiles, maximum ice thicknesses ranged from 285 meters to 279 meters. Three years later, when a second survey opportunity arose, the iceberg had fragmented into three major pieces, B15a, B15j and B15k. During this field season, UTIG acquired new radar sounding data over the three-iceberg suite, with the goal of repeating the previous profiles as closely as possible; the objective was to detect any changes in ice thickness due to melting as well as to monitor the state of the stress-induced basal fractures. At the time of the most recent data acquisition, the largest portion (B15a) had drifted northward to between Franklin Island and Beaufort Island, large chunk B15j was rotated in place where the Ross Ice Shelf meets Ross Island, and sliver B15k had drifted westward from B15a. Due to relatively rapid iceberg rotation, precisely repeating the lines was impossible; cross-overs, minimum and maximum ice thicknesses were obtained, as well as the freeboard from laser altimetry. The repeated shorter lines had maximum thickness of 266 meters and 274 meters, respectively, compatible with several meters of ablation.
Fast-Ice Formation in Presence of Tabular Icebergs in McMurdo Sound, Antarctica
Icebergs calved from the Ross Ice Shelf during last five years (B15, C16 and C19) and currently adrift in the Ross Sea have had an impact on the formation and break-Up of fast ice (sea ice that is immobile and attached to the shore) in the vicinity of McMurdo Sound. Between October 2004 and May 2005, a great wall of icebergs consisting of B15A, the Dragalski Ice Tongue, B15K and C16 fortuitously developed along the Victoria Land Coast and effectively isolated McMurdo Sound, from the effects of the open Ross Sea. During this time period, 40% to 90% of the seaward boundary of this area was blocked by these icebergs. The iceberg barrier occurred during the warmest summer months (December - March) when sea ice is normally cleared from the area; thus, as the 2005 austral winter progressed, land-fast multiyear sea ice has remained in the region, despite a subsequent break-down of the iceberg barrier (i.e., B15A drifting north). This barrier reduced the effect on the sea ice of southerly winds by presenting a mechanical buttress to nortward flow, and also blocked currents from the Ross Sea which normally contribute to the break up of fast ice in the austral summer. In addition, lateral melting of the icebergs resulted in the freshening of the ocean surface layer, enhancing stratification and sea-ice formation. This study shall Geographic Information System analysis of satellite imagery to quantify the length of the iceberg barrier, and to demonstrate the resultant impact on the development and total extent of fast ice. A simple numerical model of one-dimensional water-column/sea-ice interaction will be used to quantify the effects of fresh-water flux from iceberg melting and associated effects on sea-ice formation.
Chemotrophic Ecosystem Beneath the Larsen Ice Shelf, Antarctica
The first living chemotrophic ecosystem in the Southern Ocean was discovered in a region of the seafloor previously occupied by the Larsen-B Ice Shelf. A towed video survey documents an ecosystem characterized by a bottom-draping white mat that appears similar to mats of Begiattoa, hydrogen sulfide oxidizing bacteria, and bivalves, 20-30 cm large, similar to vesicomyid clams commonly found at cold seeps. The carbon source is unknown; three potential sources are hypothesized. First, thermogenically-produced methane may occur as the marine shales of this region are similar to hydrocarbon-bearing rocks to the north in Patagonia. The site occurs in an 850 m deep glacially eroded trough located along the contact between Mesozoic-Tertiary crystalline basement and Cretaceous-Tertiary marine rocks; decreased overburden could have induced upward fluid flow. Also possible is the dissociation of methane hydrates, a process that might have occurred as a result of warming oceanic bottom waters. This possibility will be discussed in light of the distribution of early diagenetic ikaite in the region. Third, the possibility of a biogenic methane source will be discussed. A microstratigraphic model for the features observed at the vent sites will be presented; the system is comprised of mud mounds with central vents and surrounding mud flow channels. A series of still image mosaics record the dynamic behavior of the system, which appears to demonstrate episodic venting. These images show the spatial relationship between more and less active sites, as reflected in the superposition of several episodes of mud flow activity and the formation of mud channels. In addition, detailed microscale features of the bathymetry of the site will be presented, placing the community within the context of glacial geomorphologic features. The Larsen-B Ice Shelf persisted through the entire Holocene, limiting carbon influx from a photosynthetic source. Tidal modeling of both pre and post breakup scenarios will be used to document oceanic circulation of the region, critical to an understanding of the role of advective processes. However, one consequence of recent ice shelf collapse is the increased downward flux of phytoplankton debris, as documented by the pockets of algal fluff observed at the sea floor and diatom counts that show a several order of magnitude increase in diatom concentration in the uppermost few cm of the sediment column. The consequences of this new source of carbon on the existing chemosynthetic community are yet to be realized, though already signs of benthic colonization are observed. Coupled to burial by dropstones, silt and clay released from glacial ice during the March 2002 ice shelf collapse, the future of this newly discovered ecosystem is uncertain. Finally, the broader implications of this discovery will be discussed, particularly with regard to the potential existence of similar ecosystems in other sub-ice settings.