The temperature distribution and evolution in steep rock was largely unknown until now, even though, permafrost thaw in rock faces maybe equally consequential (in terms of natural hazards and geotechnical consequences for infrastructure) as that in debris-covered slopes. Additionally, the thermal response of rock faces to individual extreme events is fast compared to debris slopes that are often insulated by blocky surfaces [Harris and Pedersen, 1998] and have a high ice-content.

The modeled thaw of 2003 exceeds the maximum of all previous years (one dimensional energy balance model verified with the data measured at the 14 sites). The summer thaw of 2003 (modelised) is 10–50 cm deeper than the maximum of the 21 previous years for the active permafrost zone. The immediate response of the permafrost active layer to increase of temperature has been observed during the 2003 summer.

In northern slopes, the depth of thaw is mainly controlled by the influence of air temperature (mostly via long-wave radiation) on surface temperatures, whereas southern slopes additionally receive high amounts of shortwave radiation. As a consequence, southern slopes exhibit greater inter-annual variability of thaw depth, larger pre- 2003 maxima and, therefore, a smaller 2003 anomaly.

The 2003 thaw depth is likely to exceed previous maxima even on time scales of centuries, considering the pronounced recent global and hemispheric temperature rise inferred from instrumental records and proxy data and its influence on Alpine ground temperatures

Most of the rock falls took place between June and August [2003] when the depth of thaw was not at is maximum but when the heat flux in the ground at somewhat shallower depths was greatest.

The increased thaw during the summer of 2003 far outweighs the direct effect that gradually rising temperatures have on rock wall stability in the uppermost meters. The extreme frequency of rockfall in the Alps during 2003 [Keller, 2003; Schirmeier 2003] corroborates this finding.

The observed domination of events in northern slopes can be explained by the strong effect of 2003 as well as the greater extent of perennially frozen northern slopes.

After the immediate response, such as the one observed during the 2003 summer, a delayed response would take place. The temperature profile within the permafrost would become disturbed and the lower boundary of the permafrost layer will rise (final response), both possibly causing large and deep-seated instabilities delayed by decades or centuries.

Therefore, following the projected rise in mean annual and summer temperatures during the 21st century the locations, magnitudes and frequencies of rock wall instabilities are likely to develop beyond the ranges of historic variability.