Statistical analysis of seasonal displacements at the Nordnes rockslide, northern Norway

The Nordnes rockslide in northern Norway poses a threat to local settlements along the nearby fjord due to its potential of generating tsunamis. Therefore, a monitoring program was initiated in 2007. Evaluation of the resulting monitoring data is expected to provide important contributions to the understanding of the sliding mechanisms.This paper focuses on statistical analyses of continuous laser and crackmeter measurements at the Nordnes rockslide during a period of 16 months. Annual linear displacements and seasonal fluctuations were estimated from time series of 3 lasers and 10 crackmeters.Results from the analyses show that the north-westernmost part of the area has the largest movement of more than 5 cm per year, which makes this part the most critical in terms of generation of a rapid rockslide. The amplitudes of the seasonal fluctuations estimated from crackmeter time series were approximately 0.5 mm. The largest displacements clearly occur in autumn and early winter with a stagnation or retreat phase in spring and summer. Thus, the movements are not increased during snowmelt which is a normal seasonal characteristic elsewhere. Although the temperature changes have a significant effect on the observed displacements, the seasonal variations could not fully be modelled with temperature terms in the regression models suggesting that there are other additional controlling factors. The rockslide is localized in arctic and periglacial conditions, and the documented seasonal variations are interpreted to be linked to effects of deformations caused by seasonal frost and permafrost.Prediction intervals for future displacements were also derived from the current time series. These prediction intervals are considered useful for the evaluation of future measurements and may serve as basis for defining alert thresholds for possible future early warning systems.