The formation of causative factors and the prediction of outbreak for repeated landslides on the same slope

以1999年台湾集集地震引发的两次大型滑坡-丘芬山滑坡和草岭滑坡为例，研究了滑坡破坏后的地理特征变化和滑坡的恢复。此外，本项目期间发生的新泻县中部地震和野户-汉头地震也是本次研究的对象。对于150年来有滑坡记录的草岭滑坡，估计的最大加速度与滑坡的发生没有明确的相关性。滑坡的发生不能仅用地震来解释，即提出了诱发因素的影响。通过一个简单模型的试算表明，存在重力蠕变产生屈曲褶皱的可能性，且屈曲褶皱的位置在斜坡的下部。这两个滑坡的下坡的屈曲褶皱和侵蚀都进行了一年的…更多是在一年后。草岭滑坡GPS观测结果表明，坡体中部水平位移约为5 cm/年，坡体下部水平位移约为1 cm/年，隆起1~2 cm/年。该边坡逐渐失稳。进行了简单的地震勘探，以研究应力释放对地基强度的影响。发现了弹性波速以每年400-600m/S的速度递减的趋势。在地理信息系统的支持下，分析了新泻县中部地震引起的滑坡分布。滑坡地貌中出现的滑坡面积率为5.26%，非滑坡地貌中出现的滑坡面积率为0.16%。这一结果表明，地震滑坡的发生受地表历史的强烈控制，在尚未发生滑坡的斜坡上不容易发生地震滑坡。这一现象被认为与上荷载去除较少的斜坡的重新失稳有关。

Changes in geographical features of a slope after failure and re-instabilization of the slope by those were investigated for two large-scale landslides -- the Chiu-fen-erh-shan and the Tsaoling -- triggered by the Chi-Chi earthquake in 1999 in Taiwan. Furthermore, the mid Niigata prefecture earthquake and the Noto-Hantou earthquake which occurred during the term of this project were targeted for this study, too.1. For the Tsaoling landslide which has the record of landslides in the past 150 years, the clear correlation between estimated maximum acceleration and occurrence of landslide wasn't admitted. The occurrence of landslide can't be explained only by earthquake, that is, the influence of causative factors is suggested.2. Trial calculations by a simple model show that there is a possibility of buckling fold by gravitational creep and the position of buckling fold is on the lower part of slope.3. The buckling fold and erosion on the lower slope of both the landslides progress year a … More fter year. The GPS survey at the Tsaoling landslide showed the horizontal displacement of about 5cm/year on the central part of slope and 1cm/year on the lower part of slope and the upheaval of 1-2cm/year. The slope is instabilizing gradually.4. Simple seismic exploration was carried out to investigate decrease in the ground strength by stress release. The tendency that an elastic wave speed was decreasing at the rate of 400-600m/s a year was found.5. The distribution of landslides caused by the mid Niigata prefecture earthquake was analyzed by GIS. The area rate of landslides which appeared in the landslide topography was 5.26% and that on the non-landslide topography was 0.16%. This result shows that the occurrence of landslide by an earthquake is strongly controlled by the history of the ground and a landslide by an earthquake doesn't happen easily in a slope where a landslide hasn't occurred yet. This phenomenon is thought to concern the re-instabilization of a slope from which upper load is removed Less