Abstract

Future sea level change is one of the most serious, and yet poorly constrained, impacts of anthropogenic climate change. One of the issues is the lack of observational data of very slow, glacial processes, which makes it difficult to parameterise them accurately in ice sheet models. Some attempts have therefore been made to “tune” ice sheet models using palaeoclimate records, and to use these tuned models to project future sea level rise. However, these projections are somewhat controversial, largely because of the uncertainties associated with the past sea level reconstructions. Here I discuss some approaches to improve our reconstructions of past ice sheet variability using geochemical techniques. I also present some intriguing geochemical data and modelling results from the Oligocene-Miocene Transition (~23 million years ago), which casts some doubt on the traditional way we view the evolution of Earth’s Cryosphere as a simple progression from greenhouse to unipolar icehouse, to bipolar icehouse conditions.