DOI: 10.1007/s11069-020-04498-4
论文题名: Exploring the impact of introducing a physical model into statistical methods on the evaluation of regional scale debris flow susceptibility
作者: Sun J. ; Qin S. ; Qiao S. ; Chen Y. ; Su G. ; Cheng Q. ; Zhang Y. ; Guo X.
刊名: Natural Hazards
ISSN: 0921030X
出版年: 2021
卷: 106, 期: 1 起始页码: 881
结束页码: 912
语种: 英语
中文关键词: Debris flow susceptibility mapping
; Integrating model
; Physical model
; Statistical model
; Yongji County
英文关键词: debris flow
; integrated approach
; mapping
; microtopography
; statistical analysis
; China
; Jilin
英文摘要: Regional scale debris flow susceptibility is widely evaluated by statistical methods. However, the initiation mechanism of debris flow is not considered, which leads to the neglect of the microtopographic characteristics. To address this problem, we established three novel combined models by introducing the physical model into statistical methods. The integrating models consists of two parts, the statistical models and the TRIGRS model. The eventual results obtained with the integrating model consider both the prediction result of the statistical method for debris flow susceptibility and the mechanism of debris flow initiation. To test the feasibility of the integrating model, three representative statistical models, the analytic hierarchy process (AHP), Shannon entropy (Entropy) and support vector machine (SVM) were selected to evaluate debris flow susceptibility in Yongji County of Jilin Province, China. The results demonstrate that the performance of the integrated models is significantly better than that of the single statistical model, especially in the local areas. The integrating models (AHP-TR, Entropy-TR, SVM-TR) can generate higher quality debris flow susceptibility maps (DFSMs) than the single model, which clearly reflect the scope and boundaries of the areas which are most prone to debris flow and identify the flat land and valleys between adjacent high-prone areas. It also reduces the overprediction generated by the physical model. In general, combining the statistical methods with the TRIGRS model can maximize the strengths of these models and avoid their weaknesses and obtain the effect of 1 + 1 > 2. © 2021, The Author(s), under exclusive licence to Springer Nature B.V. part of Springer Nature.
Citation statistics:
资源类型: 期刊论文
标识符: http://119.78.100.158/handle/2HF3EXSE/169387
Appears in Collections: 气候变化与战略
There are no files associated with this item.
作者单位: Jilin University, Changchun, China
Recommended Citation:
Sun J.,Qin S.,Qiao S.,et al. Exploring the impact of introducing a physical model into statistical methods on the evaluation of regional scale debris flow susceptibility[J]. Natural Hazards,2021-01-01,106(1)