본문 바로가기
  • Home

Classification of Landslide Occurrence using Statistic Analysis of Rainfall Data

  • Crisisonomy
  • Abbr : KRCEM
  • 2010, 6(3), pp.103-112
  • Publisher : Crisis and Emergency Management: Theory and Praxis
  • Research Area : Social Science > Public Policy > Public Policy in general

Jun Byong Hee 1 Namgyun Kim 1

1강원대학교

Candidate

ABSTRACT

Climatic factors are an important subject for a better understanding of hydrological response of landslide activity and are essential for developing warning system. In this paper, we proposed a method that incorporates antecedent rainfall accumulation and rainfall intensity to develop a landslide initiation threshold for the Korea. Hydroclimatic data were hathered for 18 rainfalls that triggered landslide and 17 storms that did not. Discriminant analysis is a technique for classifying a set of observations into predefined classes. Discriminant functions were developed for the landslide triggering and non landslide triggering storms with the antecedent rainfall accumulation variables and rainfall intensity variables. Discriminant function analysis classified the landslide triggering rainfall from those storms did not trigger landslide and classification results of 97.1% for the 38 cases were obtained with a satisfactory success rate. For the validation of suggested discriminant function, a heavy rainfall affected the Jecheon-si on July 14, 2009 occuring landslide and a large storm 'USAGI' on August 5, 2007 did not document landslide-triggering were applied to the discriminant function, and the calculated discriminant score shows good matching for the landslide occurring. These results show the possibility of prediction tool for the landslide occurring.

Citation status

* References for papers published after 2023 are currently being built.