Modelling and interpretation
In order to evaluate the nonlinearities of the ostracod isotope transfer functions to precipitation isotopic composition, we will further extend the methodology developed by Danis et al (2003), using a coupled thermal lake model/drainage basin model, including mass and water stable isotope transfers. The methodology requires a tuning of model parameterisations specific to each lake system and transient sensitivity studies to a variety of reasonable climate and environmental (e.g. vegetation types) scenarios.
Numerical simulations with the ECHAM atmospheric model exist already for our principal target periods (last millennium, 8.2ky event). These simulations have been carried through in a coupled mode, i.e. the atmospheric model was coupled to different ocean models to obtain a realistic representation of the dynamic of the entire climate system. The results of these simulations are however only available in a version without the embedded isotope module and have therefore to be repeated to obtain the appropriate forcing field for the high-resolution model REMO. Specific periods in the order of 100 years will then be selected for our downscaling approach with REMO. We focus here in particular on non-Dansgaard effects for the water isotopes, i.e. atmospheric circulation patterns that might partly decouple the isotopic composition of precipitation and local condensation temperatures over either Greenland or Europe. For example, lakes situated north and south of the Alps might come under the influence of different source areas (Mediterranean versus North Atlantic) in function of the large scale atmospheric flow pattern.