Medieval climate change and settlement in Iceland

Abstract

A key issue in studies of Norse settlement in Iceland is the role that climate has played in shaping the history of the island. The thesis has two main objectives: (1) to constrain the timing and nature of any significant variations in climate during the Medieval period, and around the time of Norse settlement in Iceland (c.500-1500AD). (2) to evaluate likely impacts post-settlement of such changes upon those living in Iceland. To do this, the project uses a multidisciplinary approach, combining new empirical data, existing empirical and documentary data, and a new climate model for Iceland.

Medieval glacier retreat was examined using lacustrine sediment sequences collected from two localities, that from SkeiSsvatn in Trollaskagi being notable as of being at least 1350 years in length. At Skeibsvatn, lacustrine evidence indicates the onset of Late Holocene glaciation around A.D. 650, and that glaciation of the catchment has continued uninterrupted to the present day. This constrains the warmth of the 'Medieval Warm Period', while terminal moraines constrain both 'Little Ice Age' and earlier Neoglacial advances. This is the first such lacustrine record of glacier retreat from Iceland.

Two models are presented in the thesis, the most important being an original massbalance and vegetation cover model. The first modelling approach used is to apply the threedimensional ice sheet model GLIMMER to test the sensitivity of selected small glaciers to growth and disappearance. The second modelling approach was the construction of a new model of climate and vegetation for Iceland.

The Iceland climate model is Java-based, and includes well over 12,000 lines of original code. It shows that it is possible to model monthly and annual vegetation limits, snowlines and snowcover based on inputs of primarily topography, temperature and precipitation, and constrain the modelling using glacier mass balance. This model has been used to evaluate spatial and temporal environmental responses to changes in temperature and precipitation of known magnitude, and to test the responses to long timeseries of temperature data.

Integrated mass balance and ecological modelling combined with geomorphic data can be used to assess settlement changes in southern Iceland. Soil erosion in Thorsmork was not directly driven by climatic change, and the degradation has a land management origin. The solution, as shown in the preservation of woodlands in Thorsmork, was a land management decision, and not deterministically driven by a drop in temperature. Birch woodland in Thorsmork is at risk of removal during sustained colder spells, but the climatic impact on soil erosion is limited.

Modelling of the Myvatn region indicates increased vulnerability of the landscape to environmental changes on the order of ±1°C, which are likely to trigger large changes in tree birch cover or good quality grazing land area. Growing season length is shortened by c.25%, and late-lying snowcover is an additional challenge in colder years. Response to warming is one of substantially reduced snowcover at lower elevations, and an inland shift of vegetation boundaries.

Aeolian sediment sequences from Geithellnadalur, in combination with modelling data gives further insight into the relative importance of climate and human land management practices, and this can be used to identify threshold events. Geithellnadalur shows that environmental change in Iceland is not always directly related to climate, but that climate may have an indirect influence on landscape changes. Within the sediment accumulation record, evidence of both the impact of settlement and the impact of the Little Ice Age are visible.

The new modelling approach, combined with the gathered empirical data, has provided key insights into the many different ways in which climate and environment interact over a varied topography, with consequently diverse effects upon settlement.