Talk by H.C. Steen-Larsen - NB! at hrs. 11.00 – Niels Bohr Institute - University of Copenhagen

Forward this page to a friend Resize Print kalender-ikon Bookmark and Share

Niels Bohr Institute > Calendar > Activities 2013 > Talk by H.C. Steen-Lar...

Talk by H.C. Steen-Larsen - NB! at hrs. 11.00

Observing water vapor isotopes in the North Atlantic

- implications for ice core isotope records and isotope-enabled General Circulation Models

Water vapor stable isotope observations from Bermuda, Iceland and Greenland collected over the last two years point to weaknesses in the representation of the atmospheric hydrological cycle in isotope-enabled General Circulation Models (GCMiso). It is possible with the three stations to directly observe and compare the water vapor isotopic composition in the source region, en route to Greenland, and in the sink area on top of the Greenland Ice Sheet where snow, which constitute the ice cores, accumulate. In the region around Bermuda it is found that GCMiso have good correlation with the observe isotope and d-excess variability, albeit not with the correct amplitude pointing to issues with the parameterization of evaporation from the ocean in the models. In the region around Iceland poor agreement between the observed variability in the observed isotopes compared to GCMiso outputs are found indicating issues with the treatment of recycling of moisture i.e. the ratio between advected and local evaporated moisture. In Greenland good correlations between the observed large scale synoptic variability in the isotopes/d-excess and the model outputs from GCMiso are found, albeit with a constant offset and amplitude bias. This discrepancy between observations and model outputs from GCMiso are likely a result of the sum of weaknesses in the representation of the hydrological cycle from source to sink. If we want to use GCMiso to correctly to interpret past climatic changes as observed in ice core isotope records, it is therefore important to improve the representation of the hydrological cycle in GCMiso.