The basis of the glacial history of Svalbard lies in stratigraphic studies of terrestrial and marine geological records. Classic litho- and biostratigraphy give us local relative chronologies of events but to compare with other areas and records we are dependent on absolute age control.

This presentation will focus on the Middle to Late Quaternary terrestrial record to which a range of different absolute dating techniques have been applied; these techniques all have their pros and cons regarding materials, resolution, age range, etc. Amino acid chronology, although not a strictly absolute dating method, may provide some age inferences. Radiocarbon dating of organic material such as shell and bone is still the most common technique and one that is generally considered reliable. It is however limited by its age range to the last ~50 ka, and our interest goes further back than so. Various versions of luminescence dating, particularly optically stimulated luminescence (OSL), have therefore become strong contenders during the recent decade, mainly due to their longer age range and applicability to a wider range of deposits than radiocarbon dating. Luminescence dating nevertheless has its drawbacks too, for Svalbard these are mainly related to the risk of incomplete bleaching (leading to age overestimation) and poor resolution (leading to large uncertainties for final ages). Other techniques that have been used on fossil material include electron spin resonance (ESR) and uranium-thorium (U/Th) dating. In common for most techniques is that we cannot date glacial events, but the intervening non-glacial (interstadial, interglacial) episodes. Cosmogenic exposure dating is an exception, which is also able to provide information on the thickness of ice sheets through time.

From a geologist’s point of view we will discuss these different techniques, their usefulness and what we can expect of them. Do the dating methods meet our wishes regarding resolution? Are results from different methods comparable? The discussion will be based on our experience from the ongoing SciencePub project (www.ngu.no/sciencepub) and on literature and we will show examples of method applications and comparisons.