Evaluation of landslide reactivation: A modified rainfall threshold model based on historical records of rainfall and landslides

This study proposes a modification of the conventional threshold model for assessing the probability of rainfall-induced landslide reactivation. The modification is based on the consideration that exceedance of a pre-determined rainfall threshold is a necessary but not sufficient condition to reactivate a landslide. The proposed method calculates the probability of reactivation as a function of the probability of exceedance of a pre-determined rainfall threshold, as well as the probability of occurrence of a landslide after such exceedance. The data for the calculation were obtained from historical records of landslides and rainfall.The method was applied to two complex landslides (“San Donato” and “La Salsa”) involving fine-grained debris in the southern section of the Apennine foredeep. The minimum rainfall threshold triggering landslide reactivation on the two slopes was determined by examining rainfall patterns during the 180 days preceding the slide events. For the San Donato and La Salsa landslides, the minimum triggering threshold consists of rainfall events lasting 15 days, with cumulated rainfall exceeding 150 and 180 mm, respectively. Based on hydrological and statistical analyses, the annual probabilities of exceeding the thresholds were estimated to be 0.38 and 0.25, respectively. During the period from 1950 to 1987, the minimum threshold was exceeded 14 times, and four reactivations occurred at San Donato; whereas, the threshold was exceeded 10 times and three reactivations occurred at La Salsa. Hence, the probabilities of landsliding after exceedance of the minimum rainfall threshold are 4/14 and 3/10, respectively. Finally, annual reactivation probabilities were calculated to be 0.11 and 0.08, respectively. The reliability of the minimum rainfall threshold was tested by: i) simulating variations in the stress–strain behavior of the slopes as a result of fluctuations in the water table from normal to extreme values; and ii) analyzing the results of continuous multi-year monitoring of pore pressure and rainfall variations on a slope composed of dominantly fine-grained debris.