The relationship between atmospheric chemistry and concentrations in ice cores

Project leader: Eric WOLFF

1. Resumé

Ice cores are used to make statements about past changes, of global significance, in the atmosphere. However, fundamental processes that determine the relationship between ice core chemistry and Antarctic atmospheric chemistry, and between Antarctic and global atmospheric chemistry, are not yet understood or documented. This project aims to understand these processes, so that the relationships can be used with confidence.

Some work will be aimed at understanding the present atmospheric chemistry of Antarctica. Simple aerosol collections will be used to define the seasonal pattern of Antarctic atmospheric concentrations. An automated air sampler will increase knowledge of the spatial distribution of chemistry. Such measurements, and trajectory analyses where possible, will be interpreted in terms of sources and transport, with particular emphasis on a few species that are important in ice cores (sea salt species, NO3-, SO42-, MSA). Atmospheric modelling will be used to build measurements into a larger-scale picture of the atmosphere. Opportunities will be sought to extend measurements in the vertical direction to obtain a three-dimensional picture of the atmosphere.

Aerosol measurements will be combined with surface snow and ice core data to make an initial definition of which processes are important in determining the relationship between air and ice core concentrations. Based on this, specific processes will be studied in the field and the laboratory, so that quantitative relationships can be determined, again concentrating on crucial ice core species.

2. Scientific background

The purpose of ice coring is to discover the state and composition of the atmosphere in the past, and to see how and why it has evolved. In order to do this, we need to understand the processes that created the ice core record.

Firstly, it is necessary to understand the chemistry of the present-day Antarctic atmosphere, and to set it in a regional and global context. At present, both the spatial distribution of chemical concentrations within the Antarctic troposphere, and the reactions taking place within it are very poorly known (Bodhaine and others 1992). The IGBP IGAC (International Global Atmospheric Chemistry) Project, under its Polar Focus, made it a major objective "to understand the role of polar tropospheric chemistry in global change" (IGBP 1994).

A second requirement is to understand how the chemical record becomes incorporated into the ice sheet. The concentrations seen in ice are controlled not just by the sources, but by transport to the deposition site, by a range of deposition processes (in-cloud, below cloud, dry deposition), and by post-depositional losses and movement. Our ability to extract paleoatmospheric information from the ice core data is severely hampered by our lack of knowledge of these processes. Their importance in determining the concentrations seen, for at least some sites and species, has become increasingly clear (White 1989, Neftel 1991, Jaffrezo and Davidson 1993, Wolff 1994). As a result, a number of international initiatives have begun to address this problem. The IGAC Project's second major objective in the Polar Focus is "to establish the relationship between atmospheric chemical composition and that of glacier snow and ice" (IGBP 1994).

This project directly addresses the concerns of IGAC. A first step is to understand the present-day atmosphere and air/snow relationship. However, in order to extend these into the past for ice core studies, it is necessary also to understand the controlling processes. Concentrating on crucial species measured in ice cores, we will carry out studies to improve our understanding of aspects of Antarctic atmospheric chemistry. Combining chemical and meteorological information, we will determine transport pathways of chemicals to ice core sites. We will then document the relationship between atmospheric concentrations and ice core concentrations, and choose important processes controlling this relationship to study in greater detail.

3. Research and methodology

We will combine an observational and analysis programme with a theoretical and modelling effort. Much of our work will be concentrated on regions where future ice coring may take place, and at Halley, where support for more complex programmes is available. Although the chemistry of all species is linked, we will concentrate particularly on a small number of species that are of particular importance in ice core studies (eg sea salt species, NO3-, SO42-, MSA, although this list may evolve during the five years).

Atmospheric chemistry and transport

Our first aim will be to learn more about the seasonal and spatial pattern of chemical concentrations in the Antarctic atmosphere. Year-round samples for aerosol species and nitric acid have already been collected at Halley; these will be analysed. An automated year-round sampler is being constructed, and will be used at the Berkner drilling site, and subsequently at other sites. Funding will be sought (in collaboration with other nations' researchers) to deploy samplers at several sites.

This basic information at a few sites will be augmented by study of near-surface snow concentrations. Although these are not a direct measure of atmospheric concentrations, they can give clues about the transport and evolution of chemical species, for instance as air masses move inland. We will make our own collections, and review those of others (also relevant to ICD 1).

In order to fully realise this work, a three-dimensional picture of atmospheric concentrations is needed. We will take advantage of any opportunities for balloon, kite or aircraft work as they arise, although this is likely to be possible only later in the five year period.

The transport of chemical species from source regions to the deposition site is controlled mainly by the large-scale atmospheric flow. For the first time, analyses for the FROST special observing periods open up the possibility of making accurate back trajectory analyses for selected cases. This, along with storm track information determined from historical analysis and more recent satellite imagery will be used to determine the source regions and average track for tracers.

During the FROST special observing periods, cases will be selected for times when good measurements are available in deposition areas. These will be studied in detail, and back-trajectories calculated. This should enable the environmental conditions during transport to be found, and the source regions to be identified. In more general cases, satellite data will be used to track the positions of depression centres as they approach the Antarctic continent. These two types of information will form input for some chemical modelling studies, and will enable Antarctic chemical measurements to be set in a larger-scale context. Back trajectories will also be calculated from the Met Office numerical model by following passive tracers introduced into the model.

Data from the atmospheric measurement programme, and features of ice core data, will be interpreted using theoretical study of atmospheric chemical reactions, and suitable atmospheric models. In the early part of the five-year period, we will be concentrating some effort into interpreting stratospheric data (both from new BAS instruments and other sources). One aim of this work will be to gain a better understanding of winter chemistry. The opportunity will be taken to consider the possibility that some stratospheric signals may be visible in ice core data (eg from nitrate). The chemistry of the ice age stratosphere (see ICD 3) will also be assessed.

As the project progresses, particular aspects of tropospheric chemistry will also be modelled. As this is a new area for us, the first task will be to consider what type of model would be required to answer the various questions posed by the atmospheric and ice core data (some examples might be sulphur chemistry, nitrogen chemistry, deficits and excess of sea salt elements). We will then acquire (by collaboration) the necessary tools to carry out such work. As already stated, studies will mainly be based around species that are of particular importance for ice core interpretation.

Air/snow exchange

We will aim to study the problem of air/snow chemical exchange both by gross comparisons of air and snow concentrations at the same site, and, later in the five-year period, by detailed process studies of some individual mechanisms. These studies will be set firmly in the context of international initiatives, in which BAS has a major role.

We will start by documenting the nature of the relationship between air and snow concentrations throughout the year at Halley station. We have already collected snow surface and aerosol samples daily at Halley over a two-year period. Weekly nylon filter samples (to allow total nitrate to be measured) are also available. Shallow snow cores covering the period of these samples will complete the picture. The relationship between concentrations in air, fresh snow, and buried firn will be interpreted in the light of detailed meteorological information. This will inform the choice of individual processes for more detailed study.

It is clear that drifting snow plays a substantial role in moving snow and chemicals around in Antarctica. Sublimation of blowing snow, and from the snow surface, could raise chemical concentrations in snow. Wind pumping of chemicals into sastrugi has also been implicated as a factor in raising concentrations. Specific collections to study the role of drifting snow have been made, and these will be analysed. Further study of drifted snow and sastrugi is anticipated, with the aim of deciding over what area they may be important in changing chemical concentrations.

Automated air samplers are to be deployed at future ice core sites. We will consider possible designs of automated snowfall collectors to allow the air/snow relationship to be studied at remote sites.

Specific processes will be studied during the five-year period. These are likely to include:

mechanisms controlling post-depositional losses of nitrate

post-depositional movement of MSA from summer to winter snow layers (see also ICD 1)

fractionation of sea salt elements, both during and after deposition.

Ice core measurements will define the problem, and specific laboratory and field experiments will test hypotheses about the mechanisms. In-cloud processes will be considered, although experiments are not envisaged. Cloud microphysics studies (see ICD 2) will assist this work.

Processes within the snow as it compresses to form firn and ice can be investigated using distributed models for snow. In addition to the experimental investigations described above, we shall attempt to extend the distributed models for physical processes in snow to include chemical processes (cf Morris 1987), in order to examine the reactions and dispersions of chemical species within the snow, and thus to determine appropriate transfer functions. One example would be studies of deuterium excess, the parameter which specifies the relationship between oxygen and hydrogen isotope ratios in ice. This parameter is used to determine the source temperature of the water vapour from which ice was formed, and is of great interest in paleoclimate research (Dansgaard and Oeschger 1989). The method could be improved by a closer analysis of isotopic processes within the snow, to determine how well the deuterium excess values are preserved.

Finally, possible processes at greater depth in the ice sheet will be considered. These could include movement of impurities (particularly in liquid layers at grain boundaries), diffusion, and reactions that might affect concentrations of some species (eg CO2, HCHO). Such processes are particularly important when studying old ice in deep cores. Our principal tool will be the scanning electron microscope. We have access to firn and shallow core samples, and to GRIP core samples with a wide range of chemistries.

4. Wider implications

Millions of pounds are being spent on ice core studies intended to improve our understanding of the causes of past environmental change. Such understanding will allow sensible decisions to be made about policies to avoid, or mitigate the effects of, future climate change, so that quality of life can be maintained. The studies in this project will allow that investment in ice core studies to be fully realised, by adding confidence to the interpretations of simple parameters, and giving paleoclimatic meaning to many new parameters that allow causes as well as effects to be successfully resolved. Because there exist already huge volumes of ice core data that cannot be properly interpreted, the potential increase in useable ice core data as a result of this project (within the context of international efforts) is at least as large as that to be gained by drilling new cores.

5. References

Bodhaine, B.A., L.A. Barrie, R.C. Schnell, G.E. Shaw and J.K. McKie. 1992. Symposium on the tropospheric chemistry of the Antarctic region. Tellus, 44B, 250-251.

Dansgaard, W. and H. Oeschger. 1989. Past environmental long-term records from the Arctic. In Oeschger, H. and C.C. Langway, Jr. eds. The environmental record in glaciers and ice sheets. Chichester, John Wiley, 287-318.

IGBP 1994. IGBP in action: Work plan 1994-1998. Global Change Report No. 28. Stockholm: IGBP, 51-52.

Jaffrezo, J.L. and C.I. Davidson. 1993. The Dye 3 Gas and Aerosol Sampling Program (DGASP): An overview. Atmospheric Environment, 27, 2703-2708.

Morris, E.M. 1987. Modelling of water flow through snowpacks. In Jones, H.G. and W.J. Orville-Thomas, eds. Seasonal snowcovers: physics, chemistry, hydrology. Dordrecht, Reidel, 179-208.

Neftel, A. 1991. Use of snow and firn analysis to reconstruct past atmospheric composition. In Davies, T.D., Tranter, M. and Jones, H.G. eds. Seasonal snowpacks: Processes of compositional change. Berlin Heidelberg, Springer-Verlag, 385-415.

Oeschger, H. and C.C. Langway, Jr. eds. The environmental record in glaciers and ice sheets. Chichester, John Wiley.

White, J.W.C. and others 1989. How do glaciers record environmental processes and preserve information? In Oeschger, H. and C.C. Langway, Jr. eds. The environmental record in glaciers and ice sheets. Chichester, John Wiley, 85-98.

Wolff, E.W. (In Press) Nitrate in Polar Ice. In Global Biogeochemical Cycles in Polar Ice. Berlin, Springer-Verlag.