The neotectonic movement in the middle reaches of the Jinsha River is active and the earthquakes occur frequently. Lacustrine sediments are commonly distributed on both sides of the river with stable sedimentary environment, good horizontal continuity and relatively developed stratification, which are good carriers for recording paleo-seismic events. In this study, a large number of soft sedimentary deformation structures are found in the riverside lacustrine sediments in the Taoyuan Town area in the middle reaches of Jinsha River, with strong deformation and large scale. We focus on the comprehensive analysis of four soft-sedimentary deformation profiles. In which the profiled strata are mainly medium-fine sand and clay. And the soft sedimentary deformation structures mainly include sand liquefaction, rootless faults, clay lumps and folds.
Causes analysis: In the profiles of soft sedimentary deformation structures, there are medium and fine sand layers whose thickness is from thick to super thick. Sedimentary bedding has not been observed in the sand layer; and a large number of clay debris or lumps are involved in the sand layer, which are often filled between the adjacent clay lumps; and there are quicksand channels in the sand layer. All the features indicate that the sand layer in the study profiles has been liquefied. In the study profile, we found that the soft sedimentary deformation structure has the following characteristics: The faults found in the study profile extend downward and terminate in the lower liquefied sand layer and a large number of clay lumps. There are clay lumps in the place where the clay fold structure develops, and a large number of liquefied sand bodies are filled between the fold structures. The deformation structures in the profiles are not contrastive in terms of extension, chaotic deformation characteristics and obvious stress direction. Based on the characteristics of sand liquefaction and clay deformation in the above profile, it is inferred that the deformation structure in the profile is mainly due to sand liquefaction. The liquefaction strength of sand layer determines the deformation degree of clay layer.
Trigger factors analysis: There are many factors that can trigger the liquefaction deformation of the unconsolidated sediment, such as flood, freeze-thaw, collapse and earthquake, which can cause the liquefaction deformation of the sediment under certain conditions. In this paper, the possible trigger factors are analyzed based on the combination of the structural characteristics of soft sedimentary deformation, sedimentary environment and geological background of the area. First the stratigraphic characteristics also reflect the hydrostatic sedimentary environment at that time. The soft sedimentary deformation on such a large scale could not be mainly caused by the disturbance of lake waves. The research profiles are located at a sheltered bay with weak hydrodynamics, and no alluvial strata have been found in the upper part of the soft sedimentary deformation stratum. Moreover, the soft sedimentary deformation structure caused by flooding is often a small-scale curly layered structure, which has a large difference with the deformation structure and scale in the study profiles. This suggests that alluvial and diluvial events are not the main triggering factors of the deformation. Although the landslide is likely to occur near the study area, no trace of bedrock landslide is found near the study profiles. Therefore, the invasion of bedrock landslide into the sedimentary layer cannot be the triggering factor. Moreover, the occurrence of lacustrine sedimentary layer is nearly horizontal, which is a relatively stable sedimentary state, and it is impossible to form such a large-scale slump structure due to its own gravity effect. And we don't find any sliding surface in the profiles. Therefore, the collapse is ruled out. According to the geological background and geological survey of the study area, this area does not have the conditions triggered by volcanism, glaciation and freeze-thaw. Because of the active neotectonic movement and frequent earthquakes in the study area, and seismic actions are the main trigger factors for liquefaction. So it is considered that seismic action may be the main trigger factor for the strong liquefaction deformation in the study area. According to the previous studies, the relationship between the soft sedimentary deformation structure, the liquefaction thickness and the seismic strength is discussed, the magnitude of this ancient seismic event probably reached 7 or higher.
There are sand layers in the section of “soft sedimentary deformation structure” caused by earthquake, the lower stratum is sand layer and the upper stratum is clay layer. The thickness and deformation strength of the lower sand layer determine the strength of the deformation structure of the overlying clay layer. The upper and lower surface of the sand layer are undulating, and there are clay lumps in the sand layer. The deformation structure of clay layer is complex and there is no obvious deformation rule.

There are carbonate rock, limestone and caves in the reservoir head area of Xiluodu Reservoir, which is the third largest reservoir in the world. After the impoundment, the water level has risen to about 140 meters, and consequently, more than 6 000 micro-earthquakes occurred on the reservoir head region, with magnitude of the vast majority being less than 1 and the maximum magnitude M_L3. These micro-earthquakes concentrated within an area of 10km in width from the reservoir banks, 5km in depth, and 40km in length along the reservoir basin. These earthquakes did not affect the safety of the reservoir and dam. We inverted 700 focal mechanisms by using the waveforms recorded by the reservoir's digital seismic network before and after the impoundment, and further inverted the stress field of the whole reservoir head region and the sub-regions. The results show a complex orientation of focal mechanism, different rupture types, and uneven and unstable stress state, which is not in consistency with other regional stress fields obtained by a lot of natural earthquakes, indicating the reservoir induced seismicity is not strictly controlled by the regional stress field. According to the analysis, the reservoir water flows into caves, penetrating into cracks and joints, leading to increase of pore pressure, reducing the friction and fracture strength of rocks, and generating elastic deformation caused by the increased load of reservoir water. The joint actions of these may be the cause of the earthquakes. The accumulated regional stress and local stress were released first, then, the additional stress produced by the reservoir water loading was dominating. There are no major active faults in the reservoir head area. Reservoir water level will rise again by tens of meters in 2014. With the penetration of cracks, the adjustment of stress field, and the backflow of water which will inundate the upstream region of the reservoir basin, the possibility of occurrence of moderate earthquakes cannot be ruled out. The seismic fortification criteria are high for the dam of Xiluodu Reservoir, so these earthquakes will not cause safety problems. We suggest carrying out detailed hydro-geological, geophysical explorations during the continuous active period of the reservoir-induced seismicity to obtain accurate scientific data for determining the causes of induced seismicity and searching for the technical approaches for controlling the induced seismicity. These measurements will mitigate the impact of emergencies and play an exemplary role for the other similar reservoirs.