Abstract: The aim of this study is to elucidate the correlation between landslide crack displacement and rainfall and to reveal the connection between attenuation coefficient and landslide displacement. With the rainfall-induced Xiaoyantou landslide as a case study, we employed the Pearson method for the correlation analysis between landslide crack displacement and daily, initial, and effective rainfalls, followed by the application of a linear fitting Logistic regression model to fit the stepwise period dataset of the Xiaoyantou landslide, and finally obtained the regression equations of slope displacement-correlation coefficient for different attenuation coefficients. The findings indicate that 1) Effective rainfall period of landslide spans six days to the most, and the stepwise phenomenon occurs when effective rainfall reaches 116.9 mm. 2) The correlation coefficient r between displacement and rainfall at different stages augments with an increasing rainfall attenuation coefficient α_i. When the landslide’s attenuation coefficient is 0.7, the most significant shift in displacement with varying attenuation coefficient can be observed. 3) Notably, an abrupt change in the correlation coefficient at the maximum inflection point of the slope infers whether the landslide is transitioning into a period of accelerated deformation or failure. This inference is based on fluctuations in both the attenuation and correlation coefficients, hence enhancing the precision of rainfall monitoring and early warning models.

Key words: Xiaoyantou landslide, effective rainfall, landslide displacement, attenuation coefficient, correlation analysis