Temporal Patterns of Nonseismically Triggered Landslides in Shaanxi Province, China

Haijun Qiu,Yifei Cui,Yanqian Pei,Dongdong Yang,Sheng Hu,Xingang Wang,Shuyue Ma

DOI: https://doi.org/10.1016/j.catena.2019.104356

IF: 6.367

2019-01-01

CATENA

Abstract:Examination of the temporal patterns of landslide events provides valuable insights into the baseline information used to determine landslide activity and perform risk assessment in a given area. We collected a catalog of historical nonseismically triggered landslides that occurred over 22 years in Shaanxi Province, China. We found that the annual number of slides was significantly related to the annual number of falls. The average annual numbers of slides and falls were approximately 17 and 10, respectively. The active and nonactive periods of landslides alternated within the time series of the annual number of landslides. An empirical power-law correlation exists between the complementary cumulative frequency and the annual number of landslides. The monthly distribution of landslide events is significantly associated with monthly rainfall. Most landslide events occurred in the rainy season between July and October. The average time intervals of falls and slides from July to October were approximately 12 days and 8 days, respectively. Moreover, the temporal distribution of landslide events is clustered owing to the impact of nonuniformly distributed rainfall activities. Most of the landslides concentrated in one or two months of a year. Furthermore, the nonzero values in the landslide time series are nonuniformly spaced. The complementary cumulative frequency distribution of the time intervals between landslide events can be adequately fitted by an exponential function. Based on these equations, the temporal probabilities of landslide events can be predicted. In addition, most of the nonseismically triggered landslides in Shaanxi Province were triggered by long-term antecedent rainfall and high-intensity intraday rainfall.