Sources and proxy potential of long-chain diols in marine and continental settings

Understanding past climate variability is of key importance to better predict future climatic evolution. As instrumental records are limited in time, geological archives such as sedimentary records are used. So-called proxies, obtained by the analyses of these sedimentary records, are used to reconstruct past environmental settings and climate change. One type of proxies are biomarkers and this thesis focus on one class and their potential as proxy: the long-chain diols. These compounds are resistant to degradation and can be found in sediments up to 65 million years old. These long-chain diols are useful to reconstruct sea temperature, though this reconstruction is hampered by riverine long-chain diols which bias the original signal. However, the riverine diols can be used to trace past riverine input in coastal margins. In lakes and rivers, diols are produced seasonally by unicellular algae in low-flow areas of river systems. The results described in this thesis allowed for the development of a new diol-based proxy for river input, which, in combination with other proxies, enables a much better understanding of riverine production and transport, and of soil erosion processes. New paleo-reconstructions in both tropical and polar regions have been generated, enabling improved reconstructions of past climatic changes.