1.2 million years of climate change, globally and in the Mediterranean

Abstract

In this thesis we make a detailed reconstruction of climate changes based on materials from the Mediterranean Sea. Not only does this provide new insights in climate changes in the Mediterranean region, the aim is to improve our understanding of global climate changes as well. We created a single record from the data of Ocean Drilling Program Sites 967 and 968 in the eastern Mediterranean. Then, we created a detailed age model for the sediments of these sites, by tuning the fingerprint of the North African Monsoon activity to the known variations in solar irradiation caused by Milankovic variables. The monsoon activity is recorded by the elemental ratio of titanium (reflecting dust input from a dry Sahara) over aluminum (reflecting clay input from the river Nile). This is a novel way to date the sediment, which provides an independent age estimate for the climate changes recorded in these cores. With this independent method of dating we review and revise or comment on several aspects of climatic variability, both regional and global, that were measured on these cores. I comment on the ages of sapropels, which are episodic events in the Mediterranean where the deep water becomes devoid of oxygen, and thick layers of organic material are deposited on the sea floor. Also, I measured planktic (surface ocean) and benthic (seafloor) oxygen isotopes, which reflect local temperature and global ice volume. Benthic isotopes are a common method to study the behavior of ice ages. This is the first high-resolution, long term benthic isotope record from the Mediterranean to be published, and this, in combination with its unique age model, puts us in a good position to comment on the relation of the Mediterranean and other reconstructions from the ‘open ocean’. Our studies show that, while there is no systematic discernible difference between the Mediterranean response and the open ocean per se, our method of dating the sediment does arrive at different ages than existing literature for certain deglaciations, most notably that of Marine Isotope Stage 16. Our findings suggest that there is much more variability in the global climate response to solar irradiation than generally assumed. The planktic oxygen isotope record from our Sites is integrated in the existing body of planktic isotope data for the Mediterranean surface ocean, filling several gaps in the record. We revise the age model of MedStack according to our Ti/Al based estimates, and comment on the surface/deep ocean contrast. Also included is a climate modelling study using results from Climber-2, a low resolution climate model with great potential for paleoclimate research. We use the model run (representing 650,000 years of climate change) to simulate the existence of wetlands on the Eurasian continent versus the monsoon areas. We simulate the production of natural CH4 – an important greenhouse gas – in these wetlands, and compare them to the data of atmospheric CH4 concentrations found in air bubbles of the Antarctic ice cores.