Antarctic Peninsula Precipitation and Mass Balance

1. Resumé

The project seeks to answer the question: is the ice cover over the Antarctic Peninsula in equilibrium, and what is its likely response to a change in the climate. The Antarctic as a whole has been estimated by the Intergovernmental Panel on Climate Change (IPCC) to contribute -0.3 0.3 mm a-1 K-1 to the future rise in sea level, i.e. an increase in temperature produces an increase in ice volume (Oerlemans, 1981). In contrast, the relatively warm climate of the Antarctic Peninsula may produce a negative mass balance with increasing temperature if mass losses exceed increased accumulation. Further, its complex mountainous terrain, and smaller, more rapidly responding ice sheet is more likely to reveal an early response to climate change than the main Antarctic ice sheet, making this region particularly sensitive to global warming. There is already evidence for the break-up of the ice shelves (such as the Wordie and Larsen Ice Shelves) and a warming signal has been noted in meteorological records (Doake and Vaughan, 1991; Skvarca, 1993).

The goal of the project is to define the normal year-on-year variability of accumulation across the Peninsula region, and to develop a model of the likely change in mass balance if precipitation or mass loss should increase outside this range.

2. Scientific background

Climate models have predicted an increase in global mean temperature of about 0.3C per decade (with a wide band of uncertainty), assuming greenhouse gas emissions remain as today (the IPCC "Business-as-Usual" scenario). This suggests an increase in global mean temperature of 1C by 2025, and 3C by the end of the next century, faster than seen at any time over the past 10000 years. The change in temperature will lead to a rise in sea level, estimated at about 6 cm per decade over the next century, due mainly to the thermal expansion of the oceans, but in part to the melting of some land ice. The IPCC put a wide band of uncertainty on the predicted rise in sea level and state that this is in part a consequence of our incomplete understanding of polar ice sheets and their effect on sea level rise.

The effect of the Antarctic as a whole is to take water out of the global system (Connolley and King, 1993). This assumption is based on the observation that the change in accumulation is proportional to a change in the saturation mixing ratio above the surface inversion layer, and saturation mixing ratio has a linear relationship with the mean annual surface temperature. Hence for a warmer surface temperature, there will be an increase in the moisture above the inversion layer, leading to a higher rate of precipitation. A further assumption is that the main Antarctic ice sheet is in equilibrium, and that the dynamic response to climate change is slow, and can be ignored on the 100 year time scale. The implication is therefore that the main ice sheet will thicken slightly in response to increased precipitation, taking water out of the global system.

The hypothesis of slow dynamic response to climate change are unlikely to hold true for the Antarctic Peninsula (Hindmarsh, 1992). Here the climate is warmer, and the ice sheet considerably smaller, at lower elevation but with steeply mountainous terrain, and closer to temperate at the margins. The implication is that the dynamic response of the Peninsula ice sheet to climate change may be more rapid than the main central Antarctic ice sheet. Synoptic records from the relatively well studied Peninsula have demonstrated an increasing trend in temperature for several decades along with an increase in the frequency of reported precipitation. Marginal ice shelves have already shown instability, with the breakup of some of the smaller ones already underway. The challenge is to determine the balance between precipitation and dynamic response, and the effect on the mass balance of the region in order to estimate the contribution of the ice sheet to sea level rise.

3. Research and methodology

The goal of the project will be met by:

Assessment of present mass balance:

Surface elevation changes - level-line survey, satellite altimetry, photogrammetry

Accumulation rate: characterise present-day spatial and temporal variability - ice cores, snow-level sensors, remote sensing (microwave imagery) and synoptic observations

Ice shelves: characterise and document wastage - satellite imagery

Investigation of processes controlling mass balance

Cloud and precipitation processes

Synoptic processes

Modelling mechanisms and responses (leading to capability for prediction)

Climate modelling

Ice-sheet modelling

Variation in surface elevation of the Antarctic Peninsula Ice Sheet

The change in mean annual air temperature noted on weather records from the Antarctic Peninsula has already produced considerable changes in the Antarctic Peninsula Ice Sheet. Wordie Ice Shelf disintegrated from 2000 km2 during the 1960s to 700 km2 in 1989 (Vaughan, 1993). The northern portion of Larsen Ice Shelf showed a similarly dramatic collapse during the 1980s. Several smaller sites have also shown a considerable reduction. The Antarctic Peninsula thus provides a unique opportunity to study the processes of ice sheet retreat in action. We aim to: investigate the mechanisms linking the state of the ice sheet to climate in this region, evaluate the direct effect of changes in the ice sheet on sea-level, and use our observations to establish a sound foundation for the modelling of changes in the continental ice sheet.

The effect of variations in climate on the volume of ice in different areas of the Antarctic Peninsula will be monitored by measurement of surface elevation profiles. Nine sites north of 72S have already been established. Data collected over the last 20 years show that in regions where the mean annual temperature is less than 11C changes in surface elevation mirror changes in accumulation rate. However, in warmer regions, summer melting influences the surface mass balance and hence surface elevation. The network of sites will be completed by establishing level lines in the Behrendt Mountains and Haag Nunataks. Remeasurement of all sites will be undertaken towards the end of the second quinquennium. We will continue to maintain a yearly investigation of the elevation changes of Rothera ramp. The value of these studies will be extended by regional monitoring of the ice sheets, using a combination of satellite data (visible, SAR and passive microwave) (Vaughan et al, 1993; Van der Veen, 1993).

Mathematical modelling of the densification of snow under changing climatic conditions will assist in the interpretation of surface elevation changes in terms of the mass balance of the ice sheet. The techniques developed for analysis of ground-based measurements elevation will also be applicable to satellite altimetry data but these are likely to be sparse for the region as the instrument loses lock over steep mountain slopes, and most of the Antarctic Peninsula is very mountainous.

Collaborative research with Mullard Space Sciences Laboratory will aim to improve the interpretation of satellite altimeter return signals by modelling the interaction of radiation with snow. A CASE studentship for the period 1995/98 will be sought.

Ice core evidence for accumulation rate change

A systematic effort will be made to determine year-on-year trends in snow accumulation rate in the Antarctic Peninsula and adjacent areas of the Antarctic ice sheet. A continuing investigation of the trends in mass balance along the Antarctic Peninsula will monitor changes in the surface elevation of reference sections across the ice sheet. Shallow ice cores will be used to compute the recent snow accumulation at representative sites, and to apply a correction to the derived estimates of mass balance changes that arise from variations in snow accumulation rate between successive survey observation of surface elevation.

The progress of accumulation on an event-by-event basis controls the representativeness of the ice core record for sampling "climate". Changes in the seasonal balance of accumulation can potentially distort the derived annual average parameters. It is planned to develop and install snow accumulation logging devices (with cm resolution) at potential drilling sites, and to evaluate shallow pit/ice core isotopic and chemical profiles from these sites alongside the records of snow accumulation. The aim will be to determine how far such distortion needs to be accounted for in interpreting deep cores. Such field investigations may be combined with efforts to provide ground-truth for the acquisition of satellite-derived data for snow accumulation based on infrared/visible imagery, opening up the possibility of extrapolating the data over a wide geographical area.

Precipitation processes

It is essential to understand cloud properties and precipitation formation mechanisms in the Antarctic if numerical models are to represent these processes correctly and give realistic cloud distribution and accumulation totals. A programme of in-situ measurements of cloud properties and precipitation formation will be undertaken to gain insight into cloud droplet distribution and the nature of Antarctic precipitation. This will be carried out using surface-based instrumentation on exposed high ground and from tethered balloons. Such information will also be of great value in interpreting visible, infra-red and microwave imagery and in extracting cloud and precipitation information from such data.

Passive microwave imagery can now provide data on precipitation distribution and intensity for the tropics and mid-latitude regions and algorithms will be developed to produce such data for the high southern latitudes. Data on precipitation distribution over the ice-free ocean will be produced which will be related to synoptic-scale activity. Model fields will also be used to determine the source regions of precipitating air-masses.

Collaborations will be carried out with other workers in the field to improve the retrievals of annual accumulation rates from passive microwave data. This will be done by incorporating the output from snow models into the retrieval and by trying to develop a radiative transfer model for microwave radiation within the snow pack.

The topography of the Peninsula has a significant blocking effect on the eastward progression of weather systems producing markedly different climatic regimes on the eastern and western sides. Numerical models currently have great difficulty in representing this barrier correctly and therefore producing realistic precipitation distributions. Combined observational and modelling studies will be carried out to gain detailed insight into how the Peninsula interacts with synoptic disturbances with the aim of improving the model representation. This work forms part of a proposal for the EC Framework 4 Programme.

The installation of the ARIES satellite receiver imagery at Rothera has provided frequent, high resolution imagery showing the synoptic and mesoscale weather systems over the BAT area. This information will be used in conjunction with scatterometer surface winds to investigate the inter-annual variability of the synoptic activity in the region of the Peninsula and to examine the major depression tracks and cyclogenesis regions. Information on cyclone activity in numerical analyses will also be examined to determine the capabilities of models to represent correctly the synoptic environment of the Peninsula area.

Climate modelling

Climate model prediction of processes in the region of the Antarctic Peninsula are currently hampered by the low resolution of the models in this mountainous region (Connolley and Cattle, 1994; Simmonds, 1990). Higher resolution models are just becoming available and will be studied to determine whether acceptable resolution has been achieved. We shall also compare them with current models to discover the effects of this increased resolution, eg on the movement of depressions across or around the Peninsula. Insofaras precipitation is controlled by sea surface temperature (SST) and sea ice variations it may be possible to use GCM runs which prescribe these (over the past century) to provide information on large-scale accumulation which could be used to interpret local (eg ice-core) data.

Ice-Sheet modelling appropriate for mass-balance studies

Short term modelling of the dynamics of non-marine ice sheets can be carried out using a simpler set of techniques than those needed for computations through the full glacial cycles. Linearisation approaches are particularly attractive, because they are as accurate as the non-linear calculations over these time scales, and are very much more suited for the combination of statistical and inverse techniques with modelling techniques. Linearisation methods are particularly appropriate for the construction and solution of stochastic differential equations which predict the evolution of probability distributions for the configuration of the ice sheets. The Peninsula is probably the best site for carrying out such modelling, partly because it appears that ice sheet dynamics are primarily a response to climatic forcing rather then through internal mechanisms, and partly because the Peninsula will provide a suitable test-bed for these techniques before they are used in other areas where the non-linear equations must be solved in order for modelling to be carried out successfully.

4. Wider implications

A considerable effort has been made in recent years to model the global climate and predict the future trend in climate change. A consensus has now developed that the global temperature and sea level may rise slightly in the next century, partly as a response to emissions of greenhouse gases. Governments will need to respond to this scenario in planning for both reduction in emissions and the consequences of climate change. This will entail massive spending on programmes, and governments will want to be quite sure that the models predicting change are in fact accurate. This project aims to determine the strengths and failings of numerical models around the Peninsula and develop an improved representation of the topography of the area. This should result in better prediction of precipitation distribution and intensity. It is hoped that this work will also lead to a better understanding of the general problem of modelling topographic barriers which will be applicable in other parts of the world. The Antarctic Peninsula is one of the areas of the globe that the models predict will respond early to changes in global temperatures: monitoring the mass balance of the Peninsula could give a first indication of major climatic change as the result of Man's activity.

5. References

Hindmarsh, R. 1992. Estimating ice sheet response to climate change. In The Contribution of the Antarctic Pensinsula to Sea Level Rise. Ed. Morris, E.M.

Oerlemans, J. 1981. Effect of irregular fluctuations in Antarctic precipitation on global sea level. Nature 290, 770-772.

Van der Veen, C.J. 1993. Interpretation of short term ice sheet elevation changes inferred from satellite altimetry. Climatic Change 23, 383-405

Connolley, W.M. and H. Cattle. 1994. The Antarctic climate of the UKMO unified model. Ant. Sci. 6(1), 115-122.

Connolley, W.M. and J.C. King. 1993. Atmospheric water vapour transport to Antarctica inferred from radiosondes. Q. J. R. Meteorol. Soc. 119, 325-342.

Simmonds, I. 1990. Improvements in general circulation model performance in simulating Antarctic climate. Ant. Sci. 2(4), 287-300.

Doake, C.S.M. and D.G. Vaughan. 1991. Rapid disintegration of the Wordie Ice Shelf in response to atmospheric warming. Nature, 350(6316), 328-330.

Skvarca, P. 1993. Fast recession of northern Larsen Ice Shelf monitored by space images. Ann. Glaciol., 17, 317-321.

Vaughan, D.G. 1993. Implications of the break-up of Wordie Ice Shelf, Antarctica for sea level. Ant. Sci., 5(4), 403-408.

Vaughan, D.G., D.R. Mantripp, J. Sievers and C.S.M. Doake. 1993. A synthesis of remote sensing data on Wilkins Ice Shelf, Antarctica. Ann. Glaciol., 17, 211-218.