Research on early-warning prediction model of critical slide of creep landslide based on the MACD index

The forecast of landslide’s early-warning prediction is always a hot and difficult issue. The inverse velocity method is wildly used for the landslide forecast, while the accuracy of the prediction is directly affected by the appropriate starting point of the fitting curve. This paper introduces the moving average convergence divergence (MACD) index, which is a mature forecasting method in the financial filed, to quickly search the start point of fitting curve (that is, onset of acceleration during the stage of accelerated deformation). The accuracy of the model is evaluated by the case of the Qubuga landslide in Huize County, Yunnan Province. The study results show that when the MACD index of deformation rate is in the range of −1 to 1, the deformation trend of landslide does not change. When the MACD index exceeds the range of −1 to 1, the deformation tendency radically changes and the point at which the short-term average line crosses the long-term average line and the MACD index changes from negative to positive can be regarded as the onset of acceleration. The starting point of the fitting curve is quickly and effectively found by the MACD index, and the monitoring data processed by the exponential smoothing function (ESF) has a higher certainty coefficient and lower rate of deviation, less than 2%. With the update of the data, the prediction results of the model approach the actual time of failure and has higher accuracy.