CONTACT US: Contact info
|
Citation |
Tian S, Zhang S, Tang Q, Fan X, Han P. J. Natural Disasters 2019; 2019(6): 137-145. |
|
Copyright |
(Copyright © 2019, 中国灾害防御协会;中国地震局工程力学研究所, Publisher Science Press) |
|
DOI |
|
PMID |
unavailable |
|
Abstract |
In order to explore the influence of different evaluation units on the accuracy of the evaluation of landslide vulnerability,the area was divided into 12949 slope units and 27017398 grid units. The paper selected eight factors including altitudes, slopes, aspects, slope shapes, fault distance, engineering geological rock group, water system distance and highway system distance, and evaluated the vulnerability of the upper reaches of Minjiang River based on the information value model.Mathematical statistics and ROC curve accuracy test were conducted on the distribution of 2413 landslide at different levels of susceptibility. The results showed that the proportion of disasters in extremely high prone areas and advanced easy areas, grid units was 76.92% and the slope units were 78.82%, the proportion of disastrous points in extremely high prone areas is 10% lower than that in sloped areas; the density ratio of slope units in extremely vulnerable areas is higher than that in grid units, the slope units in extremely low and low prone areas are lower than that in grid units, and the evaluate result of the slope units are more reasonable. The AUC value of slope units are higher than that of grid units, and the slope units can contain a larger number of disasters in smaller areas, indicating that the evaluation results of slope units are more reasonable and accurate than that of grid units. Language: zh |