Master thesis defence by Johanne Aagaard
Measurements of total air content in glacier ice
Abstract
Total air content in ice sheets is directly related to the elevation of the ice sheet surface at the time the air is enclosed in the ice. This is one of the only ways to determine the elevation of the ice sheets in the past, making the study of total air content of the ice important.
This thesis is an attempt to further develop and improve an apparatus to measure total air content build by Sebastian Bender (2012). The apparatus is based on the barometric method described by Lipenkov et al. (1995). The basic idea in the experiment is to extract the trapped air from an ice sample by melting and refreezing the ice under vacuum in a known volume, and measuring the pressure and temperature.
In the earlier experiment not all air was extracted from the ice. Therefore the apparatus was expanded with an air trap (a filter with HayeSep polymers). With this expansion it is possible to perform several melting/refreezing cycles and trap the air between cycles. It is concluded that only two cycles are necessary to extract all air from the ice sample.
The extraction chamber is no longer a part of the measuring area which has some advantages: a temperature gradient is no longer present in the measuring area, the volume of this area no longer changes due to volume change of the ice samples, and the duration of the experiment is shortened. A water trap, installed just before the air trap to dry the air before entrapment, should decrease the water vapor pressure to be neglectable, but it is not working optimally, resulting in a small indeterminable error.
11 samples from EUROCORE, Greenland, are measures. The total air content is corrected for the error of the bubbles that are cut on the surface of the sample, which is found to be 5.42%. The total air content results lie between 0.0835 - 0.0922 cm^3/g and have a mean value of 0.0881 cm^3/g. The error of the measurements is 2.18%. The results of the total air content are close to results obtained in other studies from the same site. The results also show seasonal variations that fit the variations of the delta^{18}O record from the same depth.
It is concluded that the experiment is working. The expansion with an air trap is successful and the obtained results of the total air content are good though they are a little low.
Supervisors
Thomas Blunier, Professor, Centre for Ice and Climate
Bo M. Vinther, Associate professor, Centre for Ice and Climate