From precipitation to ice cores: an isotopic comparison at Summit, Greenland

The observed deuterium excess (d-excess) in ice cores from Summit, Greenland has high summer values and low winter values, which is opposite of the seasonal variations of most northern hemisphere locations. The interpretation of this d-excess seasonality in the context of moisture source changes is made more complicated by possible post-depositional modifications. We investigate potential post-depositional modifications within 3-4 years after precipitation events by collecting precipitation samples and comparing them with snow pit profiles at Summit. Precipitation was sampled on a storm-by-storm basis from July 2011 to September 2014. To assess the effect of wind blown snow on cross-storm contamination, we sampled at three heights (1, 2, and 4 m). Snow pits were sampled in the summers of 2013 and 2015 to span the entirety of our precipitation record. All samples were analyzed for δD and δ18O and d-excess was calculated. Mixing of snow between different storms was identified only for samples collected at the lowest height. We thus use the samples collected at the top height for interpretation. The annual cycle of precipitation isotopes follow the established seasonal relationship with the average summer enrichment of -217 and -29‰, and winter depletion of -317 and -40‰ for δD and δ18O, respectively. The d-excess shows an average summer maximum of 16‰ and winter minimum of 3‰. In the snow pit, the seasonal amplitude and phase of both oxygen and hydrogen isotopic ratios as well as the d-excess compare remarkably well with those of the precipitation. The profile appeared to be devoid of major post depositional effects except for a thin layer that changed during a melt event in 2012. However, this type of event is extremely rare at Summit, and should not significantly compromise the interpretation of precipitation isotopes in ice cores, except perhaps during climatic warm period summers. The precipitation d-excess seasonality is typically interpreted as resulting from changing moisture sources, but this does not explain the positive relationship between d-excess and d18O at Summit, Greenland. We propose that moisture sublimated from the snow surface, which typically has high d-excess values, may be an important moisture source captured in the isotope record.