| The landslide disaster is the most important geological disaster in the Loess Plateau,which has seriously affected the economic and social development of the Loess Plateau.In Heifangtai,Yongjing County,Gansu Province,because of agricultural irrigation,the groundwater level of the Taiyuan has risen continuously,and hundreds of loess landslides have been induced.It is imperative to carry out early warning and forecast of landslides in this area.There are many early warning criterion models for landslide monitoring.Which model can effectively warn landslides is the focus of research now,and whether the time prediction of landslides is also the main concern of people in the affected areas.In this paper,the ?natural landslide test field? of Heifangtai is used as the research area.Monitoring instruments are deployed to obtain monitoring data,and a process early warning model of the ground displacement and deformation of Heifangtai landslide and a comprehensive early warning model of ground tilt deformation are established.The way that the Heifangtai landslide is affected by groundwater,and the prediction method of landslide occurrence time was studied,and the short-term time prediction method of landslide was constructed.The following research results have been achieved:(1)The process early warning model based on surface deformation can effectively early warning the Heifangtai loess landslide,and successfully warned six landslides.For strong sudden landslides that cannot be effectively pre-warned,it is a feasible solution to adopt long-term early warning based on the critical water level loess landslide early warning method.(2)There is a deformation process of the surface dip angle before the failure of the slip-collapse loess landslide.According to the characteristics of the deformation curve,a threshold criterion based on the inclination deformation rate is determined,and a comprehensive early warning model based on the ground inclination deformation is established based on the deformation rate increment.The three-level rate thresholds are ±0.05°/d,±0.075°/d and ±0.1°/d.The inclination of the ground is reversed from the inclined free surface to the slope body,indicating that the cracks of the sliding-collapse type landslide have closed,and the landslide has entered a dangerous state.(3)The increase in pore water pressure caused by groundwater is the main cause of landslide instability.CSD experiments show that when the pore water pressure increases to about 0.26 times the axial stress,it will cause the soil deformation to increase rapidly and destroy,and there is a critical pore water pressure.The results of the field survey indicate that the loess infiltration height of the loess area in the landslide area accounts for an average of 0.44 of the loess thickness,and the groundwater level in the slope body accounts for approximately 0.54 of the loess thickness.Based on the height of the infiltration line and the groundwater level,a warning method for loess landslides based on the critical water level is established.(4)The laboratory test of electroosmosis drainage method shows that the addition of electrochemical slurry can effectively discharge the water in the remolded loess sample,and the bearing capacity of the soil sample is significantly improved.(5)Using the inverse velocity method for landslide time prediction,there are two monitoring value evaluation methods for the cumulative time window and sliding time window,which can be subdivided into 9 evaluation methods in total.The study found that using the real-time cumulative median of the prediction result data set for landslide time prediction has a higher reliability and is superior to the prediction result using the SLO method. |
PDF Full Text Request