Mechanism And Early Warning Of Landslide For Unstable Slope Of Loess Mountain Ridge In Baoji

The loess landslides in the Loess Plateau of northwest China have high frequency of occurrence,and the landslide hazards are highly concerned events due to its serious menace to human.Geoenvironmental factors such as hydrological and geological conditions,hydrogeological parameters,rainfall and infiltration and recharge characteristics of groundwater have huge influce of shallow loess slope stability.Therefore,learning exactly the properties of shallow loess has significanct guiding and practical value for the prevention and control of shallow loess landslide.The study area is located in the south of Wei River and the loess hilly region in front of Qinling Mountains in Baoji,Shaanxi.The control factors of loess slope stability were investigated in this study area,the hydrogeological parameters and rainfall infiltration on slope stability were mainly focused.The groundwater flow model was established to obtain the hydrogeological parameter partition of the whole study area,combined with other hydraulic parameters and the engineering properties of the material parameters,the landslide early warning model of loess slope is established and the stability on different rainfall conditions of loess slope is predicted.The main achievements of this study are as follows:(1)The steep modern terrain of the study area provides the free face for the loess landslide,the loess provides the medium condition for the landslide.The sand and mudstone of Sanmen Group under the loess which the slope is very steep is the sliding layer,rapid rising of groundwater level caused by heavily rainfall intensity,and long cumulative duration or high frequency of rainfall is the induced factor of landslide.The landslide is a palaeo-terrain-controled and rainfall-induced shallow loess landslide.(2)The vertical hydrological conductivity of the aquifer is about 0.15 m / d,while the horizontal hydrological conductivity is about 0.014 m / d,and the hydrological conductivity has obvious anisotropy.The numerical model of groundwater flow in the study area was established,and the distributed hydrogeological parameters were obtained.(3)Considering rainfall is the main factor for inducing the landslide in the study area,TRIGRS-A Fortran Program for Transient Rainfall Infiltration and Grid-Based Regional Slope-Stability Model designed by Baum et al was selected for the landslide early warning model.In order to make the early warning model accurate and reliable,based on the landslide induced by rainfall in September 2014,the model is calibrated and calibrated.(4)The early warning landslide model calculated results show that the landslide does not occur on condition that the rainfall is 50 mm and occurs when the rainfall is 100 mm and with the increasing of the rainfall intensity,the the safety factor tends to decline.The occurrence areas of the landslide are located in loess escarpment of the landslide and the edge of the landslide,which consistent with the observed occurrence areas of the landslide history.