Improved interpretation of stable water isotope signals in snow and ice from northern Greenland

This thesis deals with the analysis and interpretation of ice core data from northern Greenland, one of the least studied areas in the Arctic. The focus of the thesis is the study of delta18O values which can be used as a proxy for temperature. The 13 ice cores from the North Greenland Traverse (NGT 93/95) are between 100 and 175 m long and were cored between 1993 and 1995. These cores were dated in this thesis and the highresolution values of delta18O and accumulation rate averaged to annual mean values. The time series spans the last 500 - 1000 years. The topography of the ice sheet is confirmed as the main influencing factor on the long-term annual mean values of delta18O and snow accumulation rate. The predominant southwest/west winds ensure that the northeast of Greenland is located in the precipitation shadow area of the main ice divide, which divides northern Greenland into east and west regions. The lowest delta18O ratios are found east of the divide. In addition, the eastern ice cores are characterized by small variations in delta18O and thickness of accumulation horizons. The correlation between the time series is low due to the large distance between the ice cores and the influence of stratigraphic local noise. By stacking all individual records the signal-to-noise ratio was improved. The stack is representative for a large area in northern Greenland and allows for conclusions regarding the climate during the last thousand years (ending 1994). Globally-known results, such as the Medieval Climate Anomaly that was followed by the Little Ice Age, could be detected in the averaged delta18O time series. The time series is also sensitive to decadal variations such as the warm signal between 1920 and 1930, which cannot be broken down in climate models. Another warm period around 1420 A.D. was observed, which until now has not been seen in any other data set. It is suspected that changes in the local climate dynamic as a result of changed sea ice coverage in the Arctic Sea are possible reasons. Because there was no clear trend in the delta18O NGT 93/95 data in recent years, additional ice cores were drilled in 2011 and 2012 in northern Greenland at comparable positions, in order to extend the series into more recent times. Five cores, each about 30 m long, were analyzed at the AWI Bremerhaven for this purpose. The cores were yearly dated, and the high-resolution delta18O values were averaged for each year. An overlap of about 100 years made it possible to use the time series as an extension of the NGT 93/95 time series. For the first time, an estimation of the actual air temperatures in Greenland over the last thousand years was possible with only one proxy. Since 1980, there has been a constant warming trend. The gradient of the warming is not unusual when compared with other values from the last thousand years, but the level of the values is unusual high.