^a Forestry and Forest Products Research Institute (FFPRI), P.O. Box 16, Tsukuba Norin, Ibaraki, 305-8687, Japan

^b Norwegian University of Science and Technology (NTNU), Norway

The two landslides that were monitored showed definite secondary and tertiary creep stages before they collapsed. One of the landslides moved from the secondary stage to the tertiary creep stage when another landslide occurred nearby. The tertiary stage of this landslide showed reconstruction of short primary, secondary, and tertiary creep stages. These phenomena suggested that (1) the stress at the end of the landslide mass was released during the nearby landslide, and (2) a new stress distribution was formed in the landslide mass. The critical strain differed for 14 times between the two landslide masses we monitored. The difference was likely attributable to the difference in the contents of quick clay, which shows small critical stress against slope failure, as well as topological factors.

Our analyses of the effects of hydrological and meteorological factors on landslides showed that the precipitation of 3 and 10 days before six slope failures as the final stages of the landslides that had occurred in the research area was no different from the mean precipitation of periods that showed no slope failure, suggesting that precipitation had no direct effects on the collapse of the landslide masses. On the other hand, the traveling velocities of the landslide masses during the secondary creep stage, which was prior to their collapse, were affected by the water content of the soil and precipitation (and the amount of snowmelt water), but was little correlated with the pore-water pressure of the quick clay layer. We also found that the presence of snow cover scarcely affected landslide movements.