Yes / Precise chronologies underpin all aspects of archaeological interpretation and, in addition to improvements in scientific dating methods themselves, one of the most exciting recent developments has been the use of Bayesian statistical analysis to reinterpret existing information. Such approaches allow the integration of scientific dates, stratigraphy and typological data to provide chronologies with improved precision. Settlement period sites in Iceland offer excellent opportunities to explore this approach, as many benefit from dated tephra layers and AMS radiocarbon dates. Whilst tephrochronology is widely used and can provide excellent chronological control, this method has limitations; the time span between tephra layers can be large and they are not always present. In order to investigate the improved precision available by integrating the scientific dates with the associated archaeological stratigraphy within a Bayesian framework, this research reanalyses the dating evidence from three recent large scale excavations of key Viking Age and medieval sites in Iceland; Aðalstræti, Hofstaðir and Sveigakot. The approach provides improved chronological precision for the dating of significant events within these sites, allowing a more nuanced understanding of occupation and abandonment. It also demonstrates the potential of incorporating dated typologies into chronological models and the use of models to propose sequences of activities where stratigraphic relationships are missing. Such outcomes have considerable potential in interpreting the archaeology of Iceland and can be applied more widely to sites with similar chronological constraints. / British Academy (MD120020) awarded to C. Batt. Rannís PhD funding for M.Schmid. / The full text was made available at the end of the publisher's embargo.

Grazing by domestic livestock is one of the primary ways by which humans have modified landscapes. At low stocking rates livestock grazing can modify vegetation community composition, but at high stocking rates grazing can also reduce vegetation productivity and initiate soil erosion, leading to land degradation. The country of Iceland has undergone severe land degradation over the past 1100 years, with over half of the former vegetation cover being lost, and the remainder having depleted productivity. This work focuses upon the role that grazing by domestic livestock played in this degradation, and how the interactions between farm management, vegetation cover and climate affected grazing patterns in space and time. The aims of the research were achieved by constructing an environmental simulation model, called Búmodel, which allowed a cross-disciplinary approach that integrated landscape ecology, environmental archaeology and historical analysis. Búmodel was loosely coupled with GIS so that spatially based model inputs and outputs could be displayed and analysed in map form. The purpose of Búmodel was to predict spatial and temporal patterns of vegetation biomass production and utilisation (through grazing and hay-making) with a view to commenting on vegetation degradation in the pre-modern period (pre-1900 AD). The model was parameterised using contemporary and historical Icelandic agricultural data. Model validation was undertaken using sensitivity tests and comparison with data from an independent grazing experiment in the north of Iceland. Búmodel was then applied to two contrasting study areas: Vestur- Eyjafjallahreppur, a farming community on the south coast of Iceland, and Hofstaðir, a farm estate in the north east of the country, situated inland by Lake Mývatn. These applications demonstrated the importance of farm management in avoiding land degradation and in ameliorating the impact of climate. They also established the usefulness of Búmodel as a tool for the investigation of human and environmental interactions in Iceland.