The Mediterranean (Med) region will be warmer and drier with enhanced seasonal contrast in the 21st century. Warming and reduced river discharge have antagonistic effects on the Med surface water density. The future circulation is estimated to be weaker although the degree of modification is model-dependant.
The objective of MedSens project is to evaluate the Med Sea circulation sensitivity to hydrological/thermal perturbation under warm and strong seasonality condition. To tackle this issue, we will combine (1) the reconstruction of Med Sea state during the past perturbation events of strong amplitude using a series of proxies with (2) numerical simulation based on highly-resolved regional (1/8º) proxy-enabled models that can simulate localized convection in the Med Sea. Our target is organic-rich layers called sapropel that were deposited in stagnant circulation state, in particular sapropel S5 formed during the last interglacial period, the penultimate warm period comparable with near future. This strategy is supported by the success of our previous works on the reconstruction of the Med circulation during the Holocene sapropel S1 with neodymium isotopic composition recorded in the authigenic phases as well as the development of regional modelling platform that simulate the past Med circulation. Major innovative and original aspects of the MedSens project are as follows. First of all, the combination of proxy reconstruction and proxy-enabled (neodymium isotopic composition and oxygen isotopic composition) regional modelling to study the past Med circulation is unique on the international scale. It will allow direct data-model comparison to evaluate physical processes affecting proxy variability and the performance of model that will be used for future projection. Secondly, we will apply multi-proxy approach (geochemistry and micropaleontology, including new potential proxies) to a series of sediment cores along a large zonal transect in the Med Sea. This approach will obtain robust reconstruction of climate variables by minimising the effects of proxy-specific bias. Thirdly, recently developed automated tool MiSo (Microfossil Sorter prototype, within “Foraminifera Image Recognition and Sorting Tool” http://first.cerege.fr, patent pending) will permit foraminiferal sample preparation and automated micropaleontology, yielding high-throughput and high resolution micropaleontological and geochemical records.