On prediction of slope failure time with the inverse velocity method

The inverse velocity (INV) method is widely used for predicting the slope failure time. When applying the INV method, the inverse velocity can be assumed to be a linear and non-linear function of time, respectively, which are called linear and non-linear INV methods in this paper, respectively. Very few guidance is available in the literatures on the use of the two types of INV methods. In this paper, the performances of the linear and non-linear INV methods are assessed using a landslide database with 55 case histories. It is found that, two types of pitfalls may be encountered when applying the non-linear INV method, i.e. the saddle point and the ill-conditioned Hessian matrix. For the landslides examined in this paper, the linear INV method is free from the two pitfalls. When these pitfalls are encountered, the failure time predicted based on the non-linear INV methods may be significantly different from the actual slope failure time. For the landslides examined in this paper, the linear INV method is not only more stable, but also more accurate than the non-linear INV method. It is suggested that the linear INV method should be preferred over the non-linear INV method in future applications.

Automated summary

This paper assesses the performances of linear and non-linear inverse velocity (INV) methods in predicting slope failure time. It is found that non-linear INV methods may encounter pitfalls such as saddle points and ill-conditioned Hessian matrices. On the other hand, linear INV methods are more stable and accurate, making them a preferred choice in future applications.