Sudden Loess Landslide Monitoring And Early Warning Research

The loess is widespread in China and covers around 10% of the total land area.Landslide is the major geo-hazard in loess plateau,causing massive property damage and affecting the economic development.Numerous researches on the loess landslides triggered by irregation have been studied.Accordingly,early-warning models based on the basic characteristics and formation mechanisms of loess landslids are build in use to reduce the risk of landslisde failure.Here,research based on Heifangtai in Gansu province is used as a case study to investigate the basic characteristics and process evolution of irrigation induced sudden loess landslides.A monitoring and predictive methodology is proposed according to the performance of physical simulation and triaxial tests,includes early-warning model,time forecast method and sliding distance prediction models.The major achievements in this paper are as followed:(1)The basic characteristics of loess landslides in Heifangtai are investigated.The loess landslides perform the features of sudden and progressive regression in the effectunder the work of underwater.The evolution process and formation conditions of loess landlisdes in Heifangtai are analyzed for this type of sudden loess landslides.(2)A self-developed adaptive automatic displacement meter(AADM)is adopted to monitor surface deformation for the sudden loess landslide,which overcome the limitation that monitoring frequency and power and storage capability during the long-term monitoring process.According to characteristic of progressive regression,the AADM is located/placed on the area where the landslide failure easily occurs in order toto capture the comprehensive deformation curve of the sudden loess landslide.Meanwhile,the water level gauge,temperature and humidity barometer are conducted to acquire/get the primary environmental inforamtion and monitor the surrounding nature.The results show that the deformation curve of the this type of landlside goes slowly with a constant deformation rate at first,thenand then suddenly increases to fasten the deformation process.Also,the humiture,atmospheric pressure and precipitation of this area have no direct influence on landslides deformation.(3)The physical simulation tests are conducted to study the deformation destroy process,through the widely spread sensors to get micro vibration and infrasound before the landslide failure.The correlation between porewater pressure and slope displacement is studied and it is found that the slightly fluctuation of pressure leads to landslide unstability until the porewater pressure increases to a certain level.The destroyed process of the sudden loess landslide is vertified experimentally by physical simulation tests.What's more,ICU and CSD are used to investigate the formation mechanism of sudden loess landlsides.The performances show the characteristics of strain softening and liquidation for the loess in Heifangtai.The soil axial strain suddenly increases and then causes instability failure when the porewater pressure reaches to a certain level,which explain the sudden instability process of lanslides causing by underwater induced porewater pressure.(4)A threshold of deformation rate early-warning model is proposed to monitor loess landslide deformation,in which threshold of rate is used to estimate the deformation rate of landslides and then combined with rate increament to predict the development trend of landslide deformation.It reveals that tthe…..The transfer from a stage of uniform deformation to next stage of increasing deformation is mainly due to the critical vertical deformation.According to the influence of porewater pressure and vertical deformation,an early-warning model of critical underground water and critical vertical deformation for loess landslides in Heifangtai is carried out as well.(5)The study of landslide deformation curve,underground water changes and vertical deformation are carried out to build a comprehensive early-warning model for loess landslides in Heifangai.The inverse velocity model is adopted to quantificationally analyze sudden loess landslides deformation curve in Heifangtai.Based on the deformation characteristics of this type of landslides,short-term and medium-term time prediction methods are suggested through fitting the landslides deformation rate curve and determing vertical deformation in uniform deformation stage,respectively.And the newly proposed long-term quantitative time prediction method intergrates underground water changes with seepage characteristics at front of slope.(6)The movement and accumulation characteristics of loess landslides in Heifangtai are investigated and the main factors affecting sliding distance are determined based on photogrammetry and statistic analysis.Furthermore,a simplified and effective sliding distance prediction model is poposed through considering only two parameters of landslide back distance and average gradient of front slope for landslide sliding distance calculation.