Title: Modelling of present and Eemian stable water isotopes in precipitation.
A thesis submitted November30, 2009 for the degree of Doctor of Philosophy and defended Friday February 26, 2010.
The Niels Bohr Institute, Centre for Ice and Climate, The Graduate School of Science,
Faculty of Science
University of Copenhagen
Sigfús J. Johnsen, Centre for Ice and Climate
Laboratoire des Science du Climat et de l'Environnement
Modelling of present and Eemian stable water isotopes in precipitation.
The subject of this thesis is the modeling of the isotopic temperature proxies δ18O, δD and deuterium excess in precipitation. Two modeling studies were carried out, one using the regional climate model, and one using a global climate model. In the regional study the model was run for the period 1959 to 2001 using meteorological data and a domain including Greenland and the surrounding North Atlantic. The model was found to reproduce the observed seasonal variability of temperature and precipitation well. In comparison with ice core data from Greenland and observations from coastal stations the model captured a significant part of the winter δ18O signal from most sites. A signature of the North Atlantic Oscillation on the patterns of temperature, δ18O and precipitation was found in agreement with ice core data and observations. For inland Greenland the modeled deuterium excess level is overestimated, particularly during winter. This is probably related to the representation of micro-physics during snow formation. For coastal areas the annual cycle of deuterium excess was captured well by the model.The global climate model was used for three time slice experiments for the warm Eemian interglacial. Present day boundary conditions were used except for the insolation and the SST patterns. The modeled summer temperatures for the Northern Hemisphere were found to match proxy data well, with the large summer insolation anomalies causing warmer summers than for present day. The peak summer anomalies are +6oC for central Greenland. However, the temperature anomalies are considerably smaller and only marginally significant in Antarctica. The modeled δ18O for Greenland follows the tendency of the ice core data in the different time slices, but with underestimated amplitude. For Antarctica the modeled isotopes do not agree with ice core data. The discrepancy between the model output and the ice core data is attributed to the boundary conditions, where changes in ice sheets and vegetation have not been accounted for.