Deforming tendency prediction study on typical accumulation landslide with step-like displacements in the Three Gorges Reservoir, China

Abstract

Lots of accumulation landslides with step-like displacements occurred in the Three Gorges Reservoir due to periodic reservoir operation. This type of landslide has continued to deform for a long time, with large cumulative deformation, which may influence the land use in the reservoir and landslide control risk judgment. The accumulation landslides with four typical step-like displacement-time curves during more than 10 years, namely Baijiabao landslide, Laoshewo landslide, Heishiban landslide, and Shuping landslide, are demonstrated. The cumulative displacement-time curve of Laoshewo landslide obviously shows that its deforming tendency is slowing down, while that of Heishiban landslide clearly indicates the increasingly serious deformation. When the step-like displacement increments get close to each other, it is harder to predict the landslide tendency based on displacement, velocity, inverse velocity, or acceleration in a whole continuous period. A step-like displacement-time curve can be considered as a curve of a piecewise function, varying as time goes. Therefore, the piecewise function is proposed to analyze the tendency by different time periods. As the environment of rapid deforming period is similar, the deforming tendency of a landslide with step-like displacements can be better demonstrated by collecting data in the rapid deforming period to analyze its tendency. For Shuping landslide, the fluctuating decline in velocity and acceleration in the rapid deforming period shows that the landslide tends to become more and more stable. This well matches the status quo of Shuping landslide, as it has been treated during August 2014 to March 2015, which also verifies the effectiveness of this proposed method. For Baijiabao landslide, the inverse velocity-time curve in the rapid deforming period features fluctuating decline, indicating that the landslide tends to be increasingly unstable. Meanwhile, the discussion on failure prediction found that the watch window for critical prediction of Baijiabao landslide may open in the flood season of several years’ future. The method proposed is effective for predicting the deforming tendency of landslides with step-like displacements, provides a time window for critical prediction, and offers technical support for landslide risk control, alarm, and forecasting.