Ice cores are ice samples obtained by drilling in the ice sheet and glaciers, it is composed of layers which result from the accumulation of successive layers of snow year after year, more recent ice being at the top and the oldest at the bottom of the core. Historically, the Vostok station in East Antarctica provided the deepest ice core ever collected to a depth of 3,623m before reaching a subglacial lake and extended the ice record to the last four glacial-interglacial cycles, which provide a paleoclimate record for the last 420 000 years, while the oldest continuous ice core records dated back to more than two million years was drilled Allan Hills Blue Ice Area (Antarctica) (Yan et al., 2019). These ice cores contain a myriad of proxies that offer a range of information on past atmospheric composition and past climatic parameters, these proxies are impurities (such as dust), isotopes (Oxygen and Hydrogen), and fossil air enclosed by ice (bubbles) that allows to reconstruct past changes in atmospheric gas composition.
The dating of ice cores is fundamental to the interpretation of the different proxies, it consists of establishing a relationship between the age of the ice and the depth. One method consists of counting the annual layers, the same way it is done for tree rings and many other paleoenvironmental records, counting the annual layers is maybe the most intuitive solution. Another method is to try to spot reference horizon, it can be any significant volcanic eruptions that have emitted ashes and can be spotted in the ice cores, in such way that it can be used as a reference horizon.
Atmospheric gases
The polar ice caps contain a multitude of air bubbles, this property comes from the fact that the deep snow
However, difficulties exist, in particular when it comes to extracting the gas from the bubbles without altering its composition, another difficulties arise when we try to date these bubble, as the air spreads rapidly within the neve, the air and ice collected at the same depth are not necessarily contemporary, Barnola et al. (1991) showed that the age difference between the air trapped in bubbles and the ice can be as high as 6000 years in Vostok (Antarctica), it appears that more the rate of snow accumulation is low more the age differences will be significant. These sources of difficulties raise considerable uncertainty that must be taken into account when analysing the information to reconstruct climatic conditions from these bubbles.
Isotopes
The isotopic analysis of snow and ice samples from ice cores in Greenland and Antarctica holds a special place in the history of climate studies. An Isotopes can be defined as the different types of atoms of the same element which are distinguished only by their number of neutrons. Because isotopes of the same elements have the same numbers of protons and electrons they have the same physical and chemical properties. Thus, Oxygen has three stable isotopes, 16O, 17O, and 18O and hydrogen two, 1H and 2H (or D), from these Oxygen and Hydrogen isotopes we can found several water molecule isotopes, the most abundant are, H216O, H218O and HD18O. These isotopes are widely used in paleoclimate reconstruction, based on the principle that when seawater evaporates the light (H216O) molecule passes more quickly into the vapour phase than the heavy isotope (H218O). Consequently, when the vapour water condensate, the H218O isotope condense preferentially and the water vapour then becomes poorer in H218O.
During its migration to the poles, the air mass undergoes successive condensations process which depletes it in heavy isotope (H218O). As we know that a relationship between the proportion of water Isotopes in ice and the temperature exists (Dansgaard, 1964), we can infer the temperature at the time of the precipitation deposition based on the isotope proportion, usually the colder the temperature is the less the precipitation will contain heavy isotopes. This is explained by the increasingly impoverished depletion in heavy isotopes of the air mass by successive condensations at increasingly colder temperatures as explained above.
However this method is affected by several parameters, first, this method lay on the analysis on the changes in isotopic composition, but these changes are only recorded in the ice if there is snowfall, thus changes in the seasonality of precipitation could lead to bias in the interpretation, this issue was first raised by Krinner et al. (1997), who have shown that a decrease in winter precipitation was likely to distort the temperature recording.
Dust
Dust can be from natural sources and come from the erosion of rocks and sediments, volcanoes, fires, marine aerosols and cosmic origin. The dust varying deposition rates of dust on ice tell us about the chemistry of the past atmosphere, the changes in atmospheric circulation, and reflect climate changes, particle size and dust mineralogy, Dust can be transported over long distances by the wind to various distances and some dust particles even reached Greenland and Antarctica.
The size of dust particles inform us about the climate, for instance, large particles cannot travel over long distances which mean that it comes from bare land nearby implying a small ice extent and a warm climate, besides since that the uplift and transport of dust is mainly dependant on the concentration of dust in the atmosphere, the winds, the wind and the soil moisture/precipitation(Lawrence & Neff, 2009), the glacial period is commonly associated with an increase in dust fluxes up to 25 times greater than in interglacial period) (Delmonte et al., 2007) due to dryer climate and a decrease in sea level exposing a larger continental area from where dust will be uplift by the wind.
However, the interpretation of dust in the ice core is not straightforward. The sources of the dust are different for Antarctica, and Greenland and do not allow for inter-hemispheric correlation. The dust found in Antarctica are mainly from Patagonia (South America) (Basile et al., 1997), while the origin of Greenland dust appears to be more Asian (Gobi Desert) (Svensson et al., 2000).