Science and Implementation Plan: Chapter 8, Revised

Implementation Plan (1996)

This revision of the WAIS Implementation Plan is based on deliberations of the September 1996 WAIS workshop. Participants assessed the status of the project by reviewing our accumulated knowledge relevant to the WAIS goal and by reviewing last year's Implementation Plan and providing feedback on the list of needed and urgently needed questions to advance toward these goals most rapidly. It is significant that the areas of discussion were not the traditional scientific disciplines as used in the WAIS Science Plan, but the cross-cutting, multidisciplinary objectives of current behavior, physical controls/interactions/stability, and history. The WAIS Working Group finalized the wording of the questions which follow. In deciding the most urgently needed activities, participants rose above narrow disciplinary perspectives and displayed continued integration in this interdisciplinary environment.

Fifteen activities were agreed upon as most urgent. Other than these items, no relative priority is implied by the position of any listed activity. These most urgent activities are described more fully in the following sections. The Support Requirements Table (Section 8.7) is also revised and includes estimates of the annual costs and logistics required to carry out these activities over the next few years. This short-term focus is intended to maximize the plan's utility for both the science community-by describing the most urgently needed research proposals, and for the NSF Office of Polar Programs-by providing both science managers and logistic coordinators the most realistic assessments of near-term funding and logistic requirements. Thus, scientists contemplating involvement in WAIS are encouraged to adhere to the plan as much as possible.

As work proceeds, there will be a continuing need to reassess our state of knowledge as well as the necessary activities and those deemed most urgent. It is expected that the Implementation Plan will be updated annually at future WAIS workshops.

8.1. STATUS REPORT

8.1.1. Current Behavior

It is now clear that the West Antarctic ice sheet drainage into the Ross Ice Shelf is experiencing a complex pattern of thickening and thinning of the ice. To varying degrees of spatial resolution, basic measurements of ice flow, ice thickness and accumulation have been made, and there is change underway nearly everywhere. These imbalances are most apparent on the ice streams where flow rates are highest as the ice has now been measured to slide across a lubricated, and possibly deforming, bed. No single model matches this complex pattern of change.

Sea level is changing but the West Antactic contribution to this change is still undetermined. Changes in the onset area, grounding line and margins all would affect a change in the West Antarctic ice volume, but on different time scales.

The Ross Sea drainage area has been the site of the most comprehensive study of the spatial pattern of accumulation rates. The results clearly demonstrate a spatial complexity that belies the simple parameterizations of precipitation (tied to lapse rate) used in most modern ice-sheet models. Meteorologists have developed the ability to derive moisture fluxes from measured synoptic data, enabling a powerful independent measure of the amount of mass delivered to the ice sheet through the atmosphere.

WAIS has benefited from many extraordinary technological advances, such as GPS, satellite imagery and seismic monitoring. In the cases of borehole geophysics and airborne geophysics, WAIS has driven technological advances of its own. These data sets have been instrumental in assessing the current behavior of the ice sheet. Continued application of "cutting edge" technology will be key to advancement toward WAIS goals.

The following questions point to important activities that need to occur to advance our knowledge of the current behavior of the West Antarctic ice sheet. Starred activities are judged to be most urgent and are discussed in more detail in the next section.

C1* What is the modern spatial distribution of and controls on mass input to the West Antarctic ice sheet?

C2* What is the modern configuration and behavior of the West Antarctic ice sheet and its grounding line; where is it changing and by how much?

C3* How accurately do the current generation of coupled atmosphere-ocean GCMs simulate the climate in the West Antarctic region?

C4* What are the modern sub-ice-shelf budgetsand what factors control their spatial variability?

C5: How should WAIS data sets be incorporated into a user-shared GIS?

C6: How is the iceberg flux changing?

8.1.2. Physical Controls, Interactions and Stability

The ice streams remain the key to West Antarctic behavior. Recent WAIS research has driven a progressive awareness that side shear is significant or dominant in controlling ice streams, and that basal friction ranges from quite small to important with much spatial variability; the older notion that ice-shelf buttressing determined the fate of the ice sheet remains controversial.

Basal lubrication is a difficult subject because it requires careful surface geophysical measurements and the drilling of kilometer-long holes through which direct measurements of the basal till and water are attempted while the ice is moving rapidly. Nevertheless, direct sampling of the subglacial regime has confirmed much of what was inferred from earlier surface seismic work. It is now known that ice at the base of ice stream B is relatively clean, and at least some drag is supported on the sediment. Borehole experiments have revealed complexity in the basal water system but have not detected large, low-pressure water channels. Till sample analysis has shown that this material is extremely weak at the high water pressures measured beneath the ice streams, and preliminary data indicate at least some bed deformation, probably with some localization of displacement near the base of the ice (true sliding or very shallow deformation). The bed of ice stream C has not been sampled directly, but microearthquake observations indicate that high basal shear stress is concentrated over only a small percentage of its bed, probably in response to loss of lubricating water, and that soft sediments still exist beneath much of the ice stream.

The important role of ice-stream margins in resisting ice flow has presented glaciologists with a new challenge because the margins are heavily crevassed, endangering field parties. Misalignments of margins with basal topography pose serious questions as to the controls on marginal positions. Strong evidence for lateral jumps in margin positions indicates poorly understood complexity in their behavior. Strain softening and strain heating of the ice contained within the margin have been confirmed by vertical temperature profiles and force-balance calculations. These processes strongly affect the resistive capacity of the margins.

Satellite imagery has helped increase our knowledge of the West Antarctic ice streams not yet studied by field parties. Margins and onset areas of different ice streams display a variety of appearances, and analysis of velocity fields derived by remote sensing can even locate regions of probable high basal friction. Beyond the limits of the current ice sheet, marine seismic measurements have revealed broad furrows that indicate the course of the ice streams during the Last Glacial Maximum.

To increase our knowledge of the physical controls of the ice sheet and its stability, answers to the following questions need to be pursued.

P1* What determines the location of onset areas and are they fixed?

P2* What is the distribution of till, what is its pore-water chemistry, what are the till's mechanical properties and its response to water pressure?

P3* What is the distribution of basal drag, is it correlated with the subglacial geology or morphology and how is basal drag affected by the subglacial water and till balances?

P4* What determines the positions of ice stream margins and their effect on the discharge of ice through the ice streams?

P5* How can ice-streams be modeled more realistically?

P6* What are the ice-ocean-sediment interactions at the grounding line?

P7: How does velocity vary with depth near the ice bottom?

P8: What is the detailed structure of subglacial flutes?

P9: What is the heat balance in the basal zone underneat an interstream ridge, an ice stream, and an ice-stream margin?

P10: What role does ice-shelf buttressing have on ice-stream behavior and how is this affected by changes in ice rises, ice rumples and confining embayments?

P11: What glaciological variables control the occurrence and disappearance of internal radar-reflecting and seismic-reflecting layers?

P12: How does the ice flow, the warmer water in the deep ocean, and the shelf water interact to determine the sub-ice shelf mass balance and the shape of the sub-shelf cavity?

P13: Are there active subglacial volcanoes?

P14: What would be the East Antarctic response to a collapse, or partial collapse, of the West Antarctic ice sheet?

8.1.3. History

The West Antarctic ice sheet has had a history of repeated advance and retreat from the Ross Sea continental shelf. The most recent advance occurred during the last glacial maximum when the ice sheet was grounded at the continental shelf break in eastern Ross Sea and near Coulman Island in western Ross Sea. Preliminary results indicate a diachronous retreat of the different ice streams Multibeam and seismic data show large-scale geomorphic and subbottom features indicative of a deforming bed with relatively little channelized meltwater at the base of the ice sheet prior to its retreat from the shelf. .Ice stream A was apparently the first to retreat from the shelf. Ice flowed laterally during retreat, shifting to fill voids left by retreating ice streams. Remaining issues concern the timing of ice sheet retreat from the shelf. Ice-sheet and alpine-glacier moraines in the Executive Committee Range place elevation limits on the maximum thickness of the interior ice sheet and are consistent with a once-enlarged ice sheet stretching to the edge of the continental shelf.

Where the ice sheet presently exists, internal layers reveal historical behavior. Thinning at Byrd Station over the past 14,000 years has been relatively minor, and deposition on Siple Dome has been continuous for at least the past 10,000 years. Yet these internal layers revealed in radar soundings show contortions and relict disturbances that point to past ice streams flowing in places and along directions that differ from the modern positions and directions. These internal layers are not always seen, however. Satellite images contain an extensive array of flow-related features that underscores the temporal variability of flow patterns in West Antarctica.

The following questions define the research activities that are needed to learn essential facts about the history of the West Antarctic ice sheet.

H1* What is the climate history of West Antarctica and was there an ice sheet present there during past interglacials, particularly stage 11?

H2* What is the deglaciation history in the eastern Ross, the Bellingshausen and Amundsen Seas?

H3* What is the ice-surface elevation history revealed in the mountains within West Antarctica?

H4* What is the contemporary rate of relative sea level rise and does it help constrain the deglaciation history of West Antarctica?

H5* What do internal layers detected by radar reveal about the ice-sheet history?

H6: What are the leads and lags of West Antarctic ice sheet advance and retreat with global sea level and with glacial fluctuations in the Northern Hemisphere and in other parts of the Southern Hemisphere?

H7: Do additional subglacial sediment samples share similar mechanical and compositional properties with the samples underneath Upsteam B?

H8: What is the history and character of interaction between the East and West Antarctic ice sheets?

H9: Can subglacial material be dated more precisely witn diatoms than has been accomplished to date?

H10: What is the cause and significance of the V-shaped feature on Roosevelt Island?

H11: How do diamictons in the Ross Sea compare with those beneath the ice sheet?

8.2. MOST URGENT RESEARCH

While all of the questions listed above are important and must be answered to fully attain the WAIS goals, those starred items comprise a more urgently needed subset. Each question is described below, with a fuller explanation of the particular science to be accomplished and a recommendation of how it could be accomplished. The Support Requirements Table summarizes the funding and logistic requirements.

C1* What is the modern spatial distribution of mass input to the West Antarctic ice sheet and what controls this pattern?

The West Antarctic ice sheet results from the accumulation of snow. Variations in net accumulation (the difference between precipitation and sublimation) contribute to the resulting pattern of flow. Meteorological techniques have been developed that calculate moisture flux from standard products available from the international weather centers. Thus, the annual and seasonal accumulation rates for sectors of West Antarctica can be calculated. In addition, the planned ice cores will yield records of snow accumulation, regional climate, globally mixed gases, and solar interactions. These data, when combined with the observations described below, will provide the basis for describing and understanding the spatial and temporal variations of precipitation, sublimation and temperature over the West Antarctic ice sheet. The El Ni–o-Southern Oscillation (ENSO) phenomenon is strongly and variably manifested over West Antarctica, and understanding how its variations arise on the synoptic and hemispheric scales is essential for interpreting the climate records from the planned deep ice cores (see H1), as well as for understanding how the ice-sheet mass input could change under altered boundary conditions.

To support the ice-core interpretation and the regional meteorology data base, the University of Wisconsin Antarctic Meteorological Research Center (AMRC) will continue to archive the Antarctic composite IR images, the National Meteorological Center Global Gridded Analysis, Antarctic station observations including radiosonde soundings, Advanced Very High Resolution Radiometer (AVHRR) data, and automatic weather station (AWS) data. The European Centre for Medium-range Weather Forecasts (ECMWF) daily output will need to be collected. The field operations will be the deployment of one or more AWS units each year for the next three years or more over West Antarctica. The first AWS unit will be installed at Siple Dome during the 1995-1996 field season. The Siple Dome AWS unit will measure the snow temperature profile to a depth of 10 meters and the snow accumulation at the surface as a function of time. In addition, the usual AWS measurements will be made of air temperature, relative humidity, wind speed, wind direction, vertical temperature difference in the air, and air pressure. The Byrd AWS unit will be continued and the eight-AWS-unit array on the slope between Byrd Station and the Siple Coast will be continued. Once a second site for ice-core drilling has been selected, AWS units will be installed at the site and at four locations 100 kilometers from the drilling site. These data collection activities are already funded and logistically supported by NSF-OPP. However, the required analysis efforts are new and will need funding, but do not involve field work.

C2* What is the modern configuration and behavior of the West Antarctic ice sheet and its grounding line; where is it changing and by how much?

Migration of ice-stream margins, shutdown of ice streams, changes in grounding-line position, and a discordance of flow bands and current ice-flow directions are all indications of the changes that are occurring in West Antarctica. Calculations of the rate of ice thickness change are usually non-zero. Added to this, numerous disintegrating ice shelves in the nearby Antarctic Peninsula emphasize the need for the establishment and maintenance of an effective, comprehensive and coordinated monitoring effort. To the greatest extent possible, WAIS should exploit current and future satellite sensors which routinely collect data over the Antarctic. Field and airborne measurements of ice elevation and ice thickness are needed, both as basic data for mass-balance calculations and for verification of change.

Satellites, in particular the Landsat series, have been used very effectively to document coastal conditions in the past. A continuation of this work is important, on a routine basis, as much of what is recognized as change depends entirely on the existence of an historical record. With a new generation of Landsat and SPOT spacecraft, and EOS imagers to be launched in 1998/99, monitoring activities that utilize imagery at the visible wavelengths should flourish. An entirely new aspect of this sort of activity was introduced with the advent of regular coverage of coastal Antarctica with satellite-based imaging radars. The European satellites ERS-1 and ERS-2, the Japanese satellite JERS-1, and the Canadian satellite Radarsat, when used with the new receiving station in McMurdo, provide the first chance to have scheduled high-resolution imagery that is unaffected by clouds. These imagers will greatly simplify the work required for routine coastal monitoring. The radar altimeters on ERS-1, ERS-2, and the past missions of Seasat and Geosat, should allow a reasonable monitoring of changing conditions on the large ice shelves, where low surface slopes simplify the interpretation of these data. Finally, the EOS laser altimeter will begin an era of substantially increased precision in measuring elevation changes across the entire Antarctic ice sheet. Many of the requisite remote-sensing activities are currently underway, funded by either the NSF, USGS or NASA. However, the repeat surveys are generally not supported and require specific funding for image purchase and analysis.

Radar soundings of ice thickness are logistically difficult to obtain, but are essential in mass balance assessment. The areas most in need of coverage are ice stream E, Pine Island and Thwaites Glaciers, and a repeat set of measurements across the grounding lines of ice streams A-E and the ice plain of B. The Support Office for Aerogeophysical Research (SOAR) instrument suite is well suited to this data collection task by providing both precise surface elevations and ice thicknesses from a Twin Otter positioned accurately by GPS.

C3* How accurately do the current generation of coupled atmosphere-ocean GCMs simulate the climate in the West Antarctic region?

Atmospheric dynamics control snow accumulation rates over ice sheets. Simple thermal parameterizations cannot adequately predict the changes in accumulation rates that would result from shifts in the storm tracks in a greenhouse-gas-enhanced atmosphere. It is necessary to couple an ice-sheet model with an atmosphere General Circulation Model (GCM) that adequately includes sea-ice in order to be able to predict changes in the Antarctic net accumulation. Predictive GCMs and ice-sheet models hold the promise of understanding the nature of the glacial-interglacial cycle, but such large-scale models need to be augmented by mesoscale models (see P5) and tested by collected data (see C1) to be most useful. Ice-sheet models forced by the GCM-generated net accumulation should provide a new ice topography as an updated boundary condition for a GCM, thus forming a feedback in a coupled ice-atmosphere system (see P5).

Currently, two modeling strategies exist for simulating atmospheric processes. Regional (or mesoscale) models are used for relatively short time integrations (a few days to approximately 1 week) to understand fine-scale details of atmospheric processes, such as evolution of cyclones and related synoptic processes. GCMs are used for longer term integrations of about a month or more with applications up to geologic time scales. Both modeling strategies should be employed in WAIS. Regional modeling efforts should be focused on the synoptic-scale environment, assessing the patterns of moisture transport on a case-study basis using data from available observation networks. These case studies would be used to refine the parameterization schemes in GCM simulations.

The primary requirement is for the modeling activities but on-ground sampling of aerosol tracers is vital to understand the roigins of various air masses sampled in ITASAE ice cores. Modeling work will depend on refined GCMs with coupling of the atmosphere and ocean and is in need of additional modelers.

C4* What are the modern sub-ice-shelf budgets and what factors control their variability?

Ice shelves are in direct contact with both the atmosphere and the ocean, so they are potentially the most vulnerable parts of WAIS to environmental change. While the surface regime may be observed and sampled with relative ease, direct studies of the basal regime present a considerable challenge. Nevertheless, recent work has highlighted several important facts: (i) basal melting of ice shelves accounts for 25-35% of current accumulation on the Antarctic Ice Sheet; (ii) current mean melt rates for individual ice shelves vary over orders of magnitude (from ~0.1 to ~10 m/yr); (iii) an ice shelf may experience high melting and high freezing rates (both ~1 m/yr) at different locations on its base. The spatial variability occurs because ocean waters that interact with the ice shelves have temperatures that range more than 3 degrees above the surface freezing point. Regions of melting and freezing on the same ice shelf are controlled mainly by a process known as "ice pumping", which is driven by the pressure dependence of the freezing point of seawater. This induces melting of thick ice near the grounding zone and freezing beneath thinner parts of an ice shelf, nearer the ice front. The rates appear to be strongly affected by the pattern and strength of the ocean circulation in the sub-ice shelf cavity. The circulation is itself strongly influenced by the cavity shape, which in turn is determined by both the dynamics and the basal mass balance of the ice shelf (see P12).

Studies required to address the above question are wide-ranging and multi-disciplinary. There is a pressing need to expand the current database of observed melting and freezing rates. While radar sounding surveys give a broad picture of regions of an ice shelf that are influenced by melting or freezing, the measurement of rates is much harder. The most widely used techniques rely on the assumption that the ice shelf is in steady state, which is clearly inappropriate for large areas of the Ross Ice Shelf, for example. Direct observations require access holes to be made through the ice cover, so obtaining a good spatial coverage would be time-consuming and logistically challenging. Oceanographic observations beneath and in front of ice shelves are a high priority if measured melt rates are to be understood in terms of processes operating in the water column. Widespread observations within the cavity may prove impossible without the development of suitable autonomous submersible vehicles (ee P6), but much can be learned from oceanographic sections along the ice fronts, which enable estimates to be made of net melting in the cavity. Coincidental geochemical measurements can help to determine the residence time of water in the sub-ice cavities and the origin of meltwater outflows.

Time-series observations are also important, as many measurements of melt rates yield only spatial and temporal averages, yet seasonal and interannual variability of continental shelf water properties is known to be high. Finally, the development of suitable oceanographic models is a necessary step in relating the glaciological and oceanographic measurements. These models must cope with the unique sub-ice-shelf environment, where the thick ice cover induces a significant surface slope and isolates the ocean from atmospheric forcing.

P1* What determines the location of onset areas and are they fixed?

The acceleration of slowly moving inland ice marks the initiation of an ice stream. Learning the reason for this phenomenon is crucial for understanding why ice streams occur and is centrally relevant in understanding why they persist or, in the case of ice stream C, stop. The hypothesis of geologic controls, such as a fault zone or enhanced geothermal heat flow, suggests a stagnant onset zone where either a shift in sediment properties or an increase in subglacial water determines the onset location. Competing hypotheses appeal to the extreme thinning that accompanies the accelerating flow or the transport of deforming subglacial till, both of which require some migration of the onset area.

The answer to this question requires a concerted field study of the onset process. An initial surface survey, guided by imagery, is already funded to identify the spatial pattern of acceleration. This will be completed in 1997. Subsequent field seasons will require more detailed surface deformation work, radar and seismic investigations in those regions where the coarse survey identifies onset begins. Finally, the direct sampling of the bed in areas both upstream and downstream of onset will determine the change in basal conditions that accompanies onset. These studies should include any physical characterization of the onset process that enables numerical simulation for numerical models.

P2* What is the distribution of till, what is its pore-water chemistry, what are the till's mechanical properties and its response to water pressure?

Basal lubrication is the primary causative mechanism of ice-stream motion. To model ice-stream behavior and predict its time development in response to internal and external variables, the basal lubrication mechanism, which is composed of additive contributions from basal sliding and sub-sole till deformation, must be understood and quantitatively formulated. The chief ingredients of this formulation are the distribution of till over the bed, the source and transport controls on this distribution, the rheological properties of the till, and the basal-sliding law. The till thickness distribution can be measured in detailed seismic reflection profiles, the interpretation of which can be spot-checked in boreholes to the bottom. Via such boreholes, rheological properties of the till can be measured both in situ and in recovered core samples, as has been done on ice stream B at UpB. By basal measurements, sliding and till deformation can be measured and their contributions to basal lubrication, distinguished. Further, it is necessary to measure the basal shear stress, which enters into the sliding and rheological laws; this can best be done by laboratory creep tests on basal ice obtained by deep-core sampling in conjunction with basal shear-strain-rate measurements.

A crucial additional measurement is the basal water pressure and the closely related pore water pressure of the till and the ice overburden pressure, which have a strong effect on the strength (hence rheology) of the till, as shown by laboratory experiments on the till. Because of this effect, water pressure must be included in modeling ice-stream motion, which therefore requires a model of the basal water-conduit system in which the pressure is generated. To obtain the observational basis for constructing such a model, basal measurements that can constrain the system geometry are particularly important. Measurement of the basal melting rate, which provides the source term for water flow in the basal conduit system and the generation of basal water pressure, is also needed. In any model, the melting rate will be closely related to the vertical profile of temperature in the ice, which can be measured in boreholes. Sedimentological and paleontological studies of the till shed light on the bedrock sources of the till and the transport of the till in conjunction with ice-stream motion. Interpretive work should include formulation of basal-sliding and till-deformation laws, and relations governing basal water pressure and till generation and transport under the ice streams, suitable for use in ice-stream/ice-sheet models.

At a minimum, the above observations need to be made on an ice stream at two or three sites that are physically quite distinct, in order that the control-response relations among the several physical variables can be reasonably well revealed. Of greatest urgency for the immediate future are a second site on ice stream B, probably about 150 km upstream from UpB (already funded), and two sites on a second ice stream, either D or E, to test the generality of the conclusions from ice stream B and to extend the range of the physical conditions encountered. High-resolution seismic profiling to define the basal till distribution should be done in the region around each borehole site.

P3* What is the distribution of basal drag, is it correlated with the subglacial geology or morphology and how is basal drag affected by the subglacial water and till balances?

To understand the interplay between interior ice that is "stuck" to the bed and the ice streams with their dramatic basal slippage, we must understand the distribution and causes of basal drag. Important lubrication is provided by the till that underlies at least portions of ice streams B and C, and probably other ice streams. These tills are thought to be derived from subglacial sedimentary units combining with subglacial meltwater. If both factors contribute to the mechanical behavior of the till, then basal drag should be partly controlled by the distribution of sedimentary basins and geothermal flux. In the limit, it is possible that subglacial geology dictates areas of low basal drag and therefore the location of the ice-stream onsets (see P1) and margins (see P4).

Such simple interpretations do not account for the subglacial water distribution driven by the coupled effects of surface and bed morphology or simple bed roughness. These contributors to basal drag almost certainly play a dominant role beneath the interior ice as well as for "sticky" spots on the ice streams. Ice stream C has shut down and remains nearly stagnant, even though most of the bed remains very well lubricated. Diversion to ice stream B of basal water that once lubricated the "sticky" spots of C is supported by several data sets, providing a possible explanation for some of the nonsteady behavior of both. To illuminate the differences (and similarities) between the evolution of interior ice and ice streams, we must correlate basal drag measurements gleaned from ice-sheet morphology and velocity observations with geophysical indications of subglacial geology and the flow pattern of subglacial water.

Significant efforts to correlate basal drag with geology and water flow are already underway. A program of aerogeophysics, begun in 1994, includes radar and laser investigations of surface and bed morphology and roughness, as well as gravity investigations of subglacial sedimentary units and magnetics studies of subglacial volcanism and the thermal structure of the crust. It is focused on the region extending from the ice divide above the catchment for ice stream D to ice stream D's grounding zone, including the interstream ridges between ice streams C and E. These data will be augmented by recently completed satellite measurements of D's velocity, work approved for the onset area of D (see P1), and new work recommended at the margins (see P4) and on the ice stream itself (see P2) to give an excellent picture of the basal drag distribution. Other existing grants are aimed at using these data to calculate the likely flow path of subglacial water under both C and D--currently data collection and reduction for the 5-km-grid surface elevations and ice thicknesses desired for this work is complete only for the upper portion of C. These data, collected most efficiently by SOAR, should be collected over the remainder of C and also E. High-resolution comparison of these data with velocity fields derived primarily from satellite observations will provide important insights to basal resistance and ice flow. Photoclinometric techniques on satellite imagery can be used to interpolate surface elevations between flight lines for higher-resolution studies.

Basal seismicity patterns of ice streams B and C are providing evidence on the degree of basal lubrication, the stress concentrations, and the flow properties of sediments between "sticky spots". When combined with high-frequency, high-resolution seismic surveys validated by point borehole measurements, these greatly improve our understanding of the bed properties that control the basal drag.

P4* What determines the positions of ice stream margins and theireffect on the discharge of ice through the ice streams?

The discharge of ice through the ice streams depends on the width of the flow and the flow speeds. Although Siple Coast ice streams flow along lows in the basal terrain, the locations of the margins and the basal topography revealed by echo sounding show no clear relation. Furthermore, the existence of relict margins and discovery of the gradual outward motion of part of the margin of ice stream B shows that the margins can shift, both abruptly and gradually. The velocity pattern in ice streams based on satellite and on-the ground measurements now indicate that force transmitted from the margins to the interior of the ice streams is a major and sometimes dominant component of the control on the speed. A very strong concentration of the marginal shearing requires that the ice in the margins is softened, probably by a combination of elevated temperature and anisotropic ice texture induced by the high rates of shearing. The softening mitigates the retarding effects of the margins on the ice stream discharge. All of this information points to the importance of the margins in the ice stream system. However, understanding is still inadequate to identify the mechanism by which ice stream margins are fixed or allowed to migrate.

There is still no answer to the key issue of what are the essential differences at the bed under inter-ice-stream ridges compared to the ice streams. Quantitative analysis of the marginal drag is hindered by incomplete knowledge of the rheological and thermodynamic processes acting in the margin.

Three more margins should be studied. Each site entails a small field camp on the out-stream (safe) side of the margin. Work will entail a survey of the surface elevation and deformation fields, the collection of cores and measurement of the temperature profiles within the core holes. Science field teams will include glaciologists, geophysicists, seismologists, and drillers.

P5* How can the West Antarctic ice sheet and, in particular, the ice streams be modeled more realistically?

The collection of WAIS data must be aligned with parallel modeling projects that adequately incorporate the physics governing the system being studied. The validity of approximations and assumptions made which simplify model equations must be tested. There need to be models of the ice sheet itself, models of the surrounding oceans (see P12), including sea ice, and models of the atmosphere (see C1).

The most important aspect of the WAIS ice-sheet model is the basal boundary condition (see P2), yet it is the most difficult (and expensive) to observe. Thus, there needs to be close coordination between modelers and field scientists to optimize the return on focused studies of subglacial hydrology and sediment dynamics. These models should also strive to incorporate isostatic interaction of the ice with the lithosphere and interaction with a variable sea level.

The European Ice Sheet Modeling Initiative (EISMINT) has addressed each of the fundamental boundary conditions affecting the ice sheet as a whole, but has not concentrated on the peculiar complexities of marine ice sheets. Nevertheless, coordination of effort between the European modeling community and the U.S. modeling community could be extremely synergistic, including possible exchanges of computer-codes, modelers and workshops. Workshops on model intercomparisons and WAIS-specific model developments should be organized to accelerate model development and disseminate information and techniques to the various modeling communities.

Specific field work is not necessary, but models of the type necessary for WAIS require considerable computing resources. No single department, and few individual institutions, can support the level of hardware required to perform these modeling experiments. Collectively, computing power at the 1-YMP-equivalent processor full time is required. The processor should be piggybacked at a site where all the "extras'' (software licenses, maintenance, gurus, network connections) are already available. NASA/Goddard or NCAR are logical choices. Internet access for modelers at remote sites is, in general, sufficient for such modeling exercises.

P6* What are the ice-ocean-sediment interactions at the grounding line?

Several surprising observations are focusing interest on the grounding line or grounding zone, where the Siple Coast ice begins to float as the Ross Ice Shelf. The availability of high-resolution sonar images and cores of the deglacial sedimentary sequence of the Ross Sea is revealing channel-form features and possibly washed sediments that are indicative of large water flows. If this water was subglacial, then the existence of these features requires concentrated flows of large meltwater volumes and places important constraints on ice-flow models. If tidal processes were involved at the grounding line, then confident interpretation of the sedimentary record in terms of ice-stream processes requires removal of the grounding-line overprint.

The detection of tidal influences on ice motion far inland, past the distance at which the direct tidal raising and lowering of the ice shelf would be dissipated, provides a sensor for basal processes. The generation and propagation of tidal signals are not well-understood, however, and grounding-line studies may illuminate these processes.

Sedimentation leading to grounding and enhanced drag on the ice has been suggested as a mechanism that enhances the stability of the ice, providing a negative feedback on grounding-line retreat, possibly even in the face of locally or globally rising sea level. Whether such stabilization is active depends on the strength, supply, and water-lubrication of the deposits. The supply rate also places major constraints on basal models of ice streams and deforming beds; given the technical difficulties of studies of the ice-stream bed from the ice surface, grounding-line sediment fluxes may be required to solve some extant problems.

In addition, it is likely that grounding-line studies will improve the understanding of ice-shelf processes, freshwater and nutrient supply to the Ross Sea, and other oceanographic or biologic as well as geologic processes. For example, models and limited observations indicate that sub-ice-shelf circulation is important for melting near the grounding line (see P12). The efficiency and location of this process may control the site and rate of sedimentation. This grounding-line melting is coupled in models to net freezing farther from the grounding line, which can produce thin or thick marine ice that would trap any debris not already melted out. Understanding such processes will aid in interpreting present and past Antarctic behavior, and also the sediment transport in the enigmatic Heinrich events of the North Atlantic. Salt rejection during such freeze-on may contribute to water densification, with implications for formation of the globally important Antarctic Bottom Water. Such sub-ice-shelf models have not been tested extensively; grounding-line experiments could provide important tests.

Borehole access to the Ross Sea near the grounding line is likely the best path to obtaining such observations. Concentrated geophysical and perhaps remote-sensing work would lead up to, supplement, and perhaps continue past such borehole work.

H1* What is the climate history of West Antarctica and was there an ice sheet present there during past interglacials, particularly stage 11?

The WAISCORES element of WAIS seeks to recover and analyze two cores through the West Antarctic ice sheet, and to conduce related studies (ice-flow, geophysics, air-snow transfer, etc.) needed to interpret the cores. The main impetus is to obtain the history of climate change from this part of the world. This will improve our knowledge of the regional or global footprint of the rapid climate changes observed in Greenland ice cores, and of their timing and magnitude, information which is critical in predicting global climate change. The high-resolution nature of the available West Antarctic records will produce results not possible in East Antarctica. The dynamic nature of the West Antarctic ice sheet requires that we understand ice flow to interpret the cores; the climate-change records of the ice cores will provide a history of forcing and flow response needed to understand the ice sheet. Joint interpretation of the WAISCORES data with other WAIS results thus will improve understanding of the ice sheet and of climate change.

The WAISCORES coring is planned to begin at Siple Dome in the 1996-97 season with deep drilling in 1997-98. Concurrently, site-selection activities will be completed at an inland site somewhere near the Ross/Pine Island Bay drainage divide. An aerogeophysical survey of the inland site has been completed, but a higher resolution surface-based geophysics program to measure detailed surface strain rates, accumulation and bed topography is also needed. After completion of coring activities at Siple Dome, coring is planned to begin at the inland site. The full program of the WAISCORES drilling effort is described in the WAISCORES plan prepared by the U.S. Ice Core Working Group.

H2* What is the deglaciation history in the eastern Ross, the Bellingshausen and Amundsen Seas?

While the past extent and deglaciation history of the western Ross Sea is reasonably constrained, the deglaciation history of the eastern Ross, Bellingshausen, and Amundsen Seas is virtually unknown. The nature and coarse timing of the retreat in the western Ross Sea have provided crucial indicators of the retreat's character and the possible condition of the till at retreat initiation. Lateral continuation of the geological and geophysical database into the eastern Ross Sea and farther eastward to the Bellingshausen and Amundsen Seas will permit evaluation of individual ice-stream histories. Core transects parallel to ice-stream flow will allow assessment of timing of deglaciation and rate of retreat in each paleo-ice stream, and hopefully will clarify which, if any, ice stream(s) acted as a "weak link(s)" to initiate retreat of the ice sheet occupying the Ross Sea. Finer resolution dating throughout the Ross Sea would address questions concerning linkages between deglaciations in the Northern Hemisphere and the Southern Hemisphere.

Additional swath bathymetry, which provides a footprint image of the seafloor, seismic data, longer cores and increased core coverage are required. The current knowledge of Ross Sea deglacial history permits selection of sites that will yield high-quality and extensive records of past ice activity. Higher resolution geological and geophysical tools, currently available, must be employed to further define the sedimentary boundaries between glacial environments.

H3* What is the ice-surface elevation history revealed in the mountains within West Antarctica?

Relatively little work has been devoted to establishing former levels and ages of the inland ice as measured on the nunataks and coastal mountains of West Antarctica. These data are essential for the numerical modeling of the ice sheet from its maximum to its present configuration. Prior work has been limited by the availability of datable material, but the development of cosmogenic radionucleide dating methods is opening new areas to study. The most useful and promising geographic areas for these studies presently appear to the Reedy Glacier-ice Stream A area of the north-facing Transantarctic Mountains; the Whitmore, Theil and Horlick Mountains where the East and West Antarctic ice sheets join; and the coastal mountains and nunataks of the Pine Island Bay-Thwaites Glacier area, including the Hudson Mountains and Mt. Murphy; and the near coastal mountains, including the Kohler, Ames and Flood ranges, along the Marie Byrd Land coast.

Each area is an independent science package, and as such can be worked in any order based upon available logistics. The research in each area will require a party of 4-6 people working from a tent camp and using a combination of snowmobiles and a small helicopter. The inland camps can be put in by a C-130 aircraft. The coastal area could be researched either from a research ship/icebreaker, by helicopter, or by a tent camp equipped with snowmobiles and a small helicopter deployed from the ship, or with the same camp configuration put in by a C-130.

H4* What is the contemporary rate of relative sea-level rise and does it help constrain the deglaciation history of West Antarctica?

Detailed histories of isostatic rebound for large portions of the Laurentide and Fennoscandian post-glacial uplifts have placed constraints on Northern Hemisphere ice-sheet volume history and Earth rheology. In contrast, neither the history of uplift, nor contemporary rates of uplift, are known for West Antarctica. Relative sea-level (RSL) change inferences and GPS-determined vertical velocities are needed to enable detailed modeling of ice-sheet collapse and ensuing isostatic rebound to place important constraints on the timing, location, and magnitude of West Antarctic ice-sheet mass change.

RSL observations include mapping and dating of relict shoreline features. These observations are limited to ice-free regions, but are essential for establishing a chronology of uplift in crucial areas such as the zone of grounding-line retreat along the western margin of the Ross Sea. In addition, lake coring that is undertaken primarily to determine past environmental conditions can also establish the transition of a lake basin from marine to fresh-water or hyper-saline conditions, and hence the time at which the basin became isolated from the ocean.

GPS-determined veritical velocities can be obtained in the interior of Antarctica, where RSL methods are not possible. GPS observations in the Transantarctic Mountains would test the hypothesis of a large, late collapse in the Ross Embayment and would complement RSL inferences from Victoria Land, which are sensitive both to past ice mass changes in the Ross Embayment and in the adjacent regions of East Antarctica. GPS observations could also address tectonic questions, such as whether there is ongoing extension of West Antarctica, although the glacial rebound horizontal response is potentially significant, and would need to be considered.

Elevation data and fossil collection for deriving RSL profiles from emerged beaches, along with sediment cores in marine-transitional lakes, should be undertaken at various sites in the Ross Sea (Victoria Land and islands) and ranging eastward to the Antarctic Peninsula. Costs should be equally divided between strandline surveys and lake coring. Continuous GPS receiver operation over most or all of an Antarctic field season on an annual basis is needed to provide vertical rate accuracies sufficient to usefully discriminate between different deglaciation scenarios in a short (2-4 year) period of time. With semi-autonomous (solar- or wind-powered) GPS systems and data links for ongoing data and system monitoring, field requirements would consist of an initial deployment phase requiring several days' occupation of each site and subsequent visits for troubleshooting on an as-needed basis. Annual deployment and collection of the instrument packages may be required, depending on package design. Transportation by helicopter or fixed-wing aircraft is necessary, as appropriate. For five sites in the TAM, operating over 3 years (plus 1 year start-up) the cost is estimated to be $200k per year.

H5* What do internal layers detected by radar reveal about the ice-sheet history?

The stratigraphic layers of ice sheets provide the only known markers that can be used to examine the past flow history of the ice. Radar provides the means to map shapes of internal layers over wide areas. In areas of ice sheets that have not been disturbed by motion of ice streams, internal layering conforms to the surface and bed except possibly near the bed. The smooth layering found in Siple Dome between ice streams C and D shows that this inter-ice-stream ridge has been stable since at least the beginning of the Holocene.

Even in this stable environment, evidence for changes in the ice sheet can be detected. For example, migration of the flow divide of Siple Dome over the last several thousand years is evident in the sub-surface shapes of the layers. Where observations have been made in ice streams the layers are severely warped, probably as a result of deformation at the entry and within the ice stream. Prior disturbance by ice stream activity in presently stable areas should be detectable in the layering at depth. This concept has already been used to verify that the configuration of ice streams has changed over the last several thousand years.

A crucial question is whether the interior of WAIS inside the present onset zones of ice streams has been affected by ice stream motion in the past. If so, the present head-ward encroachment of ice streams into WAIS could be argued to be a cyclic process that will reverse. If not, the concern for major collapse of WAIS is heightened. It is important that the interpretations of the layers be based on observed geometry, rather than just the presence or absence of observed layers, which may be affected by the capabilities of the instrumentation, attenuation in the ice and surface characteristics.

Ridges (idge A/B; ridge B/C; and ridge D/E), and onset areas (particularly the onsets of B, D and E) need to be investigated with surface-based radar. The work would entail a small surface camp for one long or two short field seasons in each area.

8.3. FEASIBILITY

None of the critical research for WAIS requires major technical advances, although improvements on many fronts such as cosmogenic dating, finer resolution sonar, and more efficient drilling, obviously would benefit the research. The necessary capabilities are currently available in the United States. With the recent demonstration of ice-drilling capability for both recovery of core and bed access, as well as the emphasis on new aerogeophysical platforms, any part of the West Antarctic can be investigated either directly or remotely. Satellite and airborne remote-sensor data collected from current or planned instruments will expedite the collection of reconnaissance data and add to the analysis of field data. The laboratory capabilities also exist, as do the computers and modeling expertise to utilize them.

8.4. COORDINATION

WAIS attempts to make efficient use of logistic support through coordination of its diverse field activities. Activities by different investigators, but in the same regions, will be executed together to the extent possible. Such coordination provides greater return for expensive logistic support such as remote field bases, Twin Otter aircraft or helicopters in regions far from McMurdo Station. This coordination is encouraged by the WAIS Working Group, but the Science Support Section of the National Science Foundation/Office of Polar Programs will ultimately determine the specific logistic support. OPP Program Managers and individual scientists will work in concert with the Science Support Section to realize as much logistic efficiency as possible in the execution of WAIS investigations.

8.5. OTHER RELATED SUPPORT

The goal and objectives of WAIS fall well within the research interests of many other agencies; e.g., NASA, NOAA, DOE and USGS, which are active in global change and Antarctic studies. NASA already funds some WAIS-related research through its Radarsat and EOS programs. The WAIS Working Group will work with NSF/OPP to enlist the interest of these agencies in WAIS and to garner additional funding.

Further, many of the data sets are collected by the remote-sensing programs of other agencies. NASA supports the Landsat program while NOAA supports the AVHRR instruments. With the exception of the most recent Landsat data, these data are available at the cost of duplication.

Finally, the NSF/OPP maintains the Automatic Weather Station (AWS), Support Office for Aerogeophysical Research (SOAR), National Ice Core Laboratory (NICL) and Polar Ice Coring Office (PICO) facilities under separate funding, as well as the collection of AVHRR data at McMurdo and Palmer Stations. All of these facilities will be utilized in WAIS.

8.6. OPPORTUNITIES FOR NEW RESEARCHERS

WAIS involves many investigators, large field campaigns, and substantial funding. In relevant disciplines where the population of currently active Antarctic investigators is small, there is opportunity for other interested scientists to become involved in an Antarctic research project with direct relevance to global concerns. Scientists currently involved in Antarctic research are encouraged to involve colleagues with relevant expertise in collaborative proposals.