Snow avalanches are a natural hazard in Iceland as well as in Norway. I present an outline of recent work on estimation of avalanche risk, carried out in cooperation with the Icelandic Meteorological Office and initiated in reaction to two avalances that fell on fishing villages in northwest Iceland in 1995, killing in total 34 people.
Data on long avalanches in a given path is usually too scant to base risk estimation on that alone. Thus a key problem is to transfer data from other paths in a meaningful way. The central idea is to do this with the help of numerical models that simulate the runout of the avalanches.