Master Thesis Defense by Johanna Katharina von Drachenfels

Title: Continuous Flow Analysis of Firn Core Densities

Abstract:
The availability of snow accumulation data is a critical factor in accurate Surface Mass Balance predictions of the Greenland Ice Sheet, which forms the base for predictions of sea level rise in a warming climate. Currently, this data remains deficient due to incomplete geographical coverage and poor temporal resolution. The Lightweight In Situ Analysis (LISA) box is a portable system built for fast and straightforward accumulation measurements on firn cores directly in the field. The sample cores are melted from one end, and Continuous Flow Analysis (CFA) of seasonally varying chemical impurities and conductivity in the melt water reveals annual layers in the firn. With additional firn density measurements with a typical depth resolution of 0.5−1m, the accumulation rate of the years that are represented in the core can be reconstructed to annual resolution.
This thesis works towards a high-resolution density measurement method for the LISA box to achieve sub-annual accumulation rate resolution. An ultrasonic distance sensor now automates the melt speed measurement with high resolution. Furthermore, a liquid flow meter continuously measures the flow rate of melt water produced, which is proportional to the density of the melted sample section. The flow meter has an absolute accuracy of 10% for a flow rate of 100mL/min. Theoretically, a firn density measurement with an uncertainty of 15 − 17% could be achieved, with the low uncertainty for cores that are low density, have a large radius, and maintain a regular shape. The uncertainty could be further reduced by choosing a flow meter with higher accuracy. In practice however, to prevent water loss by overflow, the water is drained at a generously set rate. This induces air bubbles to the melt water that alter the flow rate measurements. Detailed sensitivity tests have shown that the effect can be corrected with a linear calibration for water-to-air ratios of 25 − 100%, but the results have to be smoothed to a depth resolution of several centimetres in order to remove the variability in the flow rate generated by the bubbles. Further tests for water-to-air ratios below 25% should d be conducted.
Nevertheless, the method seems to a promising approach to conducting highresolution density measurements. The effect of the bubbles can potentially be removed, e.g. by using a mass flow meter and lowering the melt water drainage rate. The limit on the depth resolution is then set by the amount of mixing of water from different layers through percolation into the sample and in the volumes of the instruments and tubing down to a centimetre scale, which is a great improvement.

Supervisor: Helle Astrid Kjær
Censor: David Basler Harder