DOI: 10.1007/s11069-020-03997-8
论文题名: Assessment of susceptibility to landslides through geographic information systems and the logistic regression model
作者: Riegel R.P. ; Alves D.D. ; Schmidt B.C. ; de Oliveira G.G. ; Haetinger C. ; Osório D.M.M. ; Rodrigues M.A.S. ; de Quevedo D.M.
刊名: Natural Hazards
ISSN: 0921030X
出版年: 2020
卷: 103, 期: 1 起始页码: 497
结束页码: 511
语种: 英语
中文关键词: Geoprocessing
; Landslides
; Logistic regression
; Natural disasters
英文关键词: efficiency measurement
; geological survey
; GIS
; hazard assessment
; identification method
; landslide
; logistics
; mass movement
; model test
; model validation
; pixel
; probability
; regression analysis
; slope dynamics
; spatial analysis
; Brazil
; Rio Grande do Sul
英文摘要: The increase in the frequency of natural disasters in recent years and its consequent social, economic and environmental impacts make it possible to prioritize areas of risk as an essential measure in order to maximize harm reduction. This case study, developed in the city of Novo Hamburgo, Rio Grande do Sul state, Brazil, aims to identify and evaluate areas susceptible to mass movements, through the development of a model based on logistic regression, associated to Geographic Information System (GIS). The construction of the model was based on the use of only four independent variables (slope, geological aspects, pedological aspects and land use and coverage) and a binary variable, which refers to the occurrence of mass movements. In total, 123,308 pixels were used as samples for the logistic regression modeling in SPSS software. As a result, we have the spatialization of a mass movement probability map with 87.3% of the correctly sorted pixels. A validation with the landslide susceptibility map built by the Brazilian Geological Survey was also performed using the receiver operating characteristic (ROC) curve, indicating a prediction accuracy of 82.5%. This research showed the efficiency of the integrated use of GIS and logistic regression, with emphasis on the relative simplicity of the model, speed of application and good ability to identify areas susceptible to landslides. The proposed model allowed the determination of the probability of occurrence of landslides with good predictive capacity, surpassing the usual model used by the Geological Survey of Brazil (CPRM). © 2020, Springer Nature B.V. Citation statistics:
资源类型: 期刊论文
标识符: http://119.78.100.158/handle/2HF3EXSE/168701
Appears in Collections: 气候变化与战略
There are no files associated with this item.
作者单位: Feevale University, ERS 239, 2755, Novo Hamburgo, Rio Grande Do Sul 93525-075, Brazil; Federal University of Rio Grande Do Sul, Av. Bento Gonçalves, 9500 (prédio 44202), Porto Alegre, RS 90501-970, Brazil; Vale do Taquari University – Univates, Av. Avelino Tallini, 171, Lajeado, RS 95914-014, Brazil
Recommended Citation:
Riegel R.P.,Alves D.D.,Schmidt B.C.,et al. Assessment of susceptibility to landslides through geographic information systems and the logistic regression model[J]. Natural Hazards,2020-01-01,103(1)