Land subsidence susceptibility mapping at Kinta Valley (Malaysia) using the evidential belief function model in GIS
Land subsidence is one of the frequent geological hazards worldwide. Urban areas and agricultural industries are the entities most affected by the consequences of land subsidence. The main objective of this study was to estimate the land subsidence (sinkhole) hazards at the Kinta Valley of Perak, Ma...
Saved in:
Main Authors: | , , , |
---|---|
Format: | Article |
Published: |
Springer
2014
|
Online Access: | http://psasir.upm.edu.my/id/eprint/34347/ http://link.springer.com/article/10.1007%2Fs11069-014-1128-1 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Summary: | Land subsidence is one of the frequent geological hazards worldwide. Urban areas and agricultural industries are the entities most affected by the consequences of land subsidence. The main objective of this study was to estimate the land subsidence (sinkhole) hazards at the Kinta Valley of Perak, Malaysia, using geographic information system and remote sensing techniques. To start, land subsidence locations were observed by surveying measurements using GPS and using the tabular data, which were produced as coordinates of each sinkhole incident. Various land subsidence conditioning factors were used such as altitude, slope, aspect, lithology, distance from the fault, distance from the river, normalized difference vegetation index, soil type, stream power index, topographic wetness index, and land use/cover. In this article, a data-driven technique of an evidential belief function (EBF), which is in the category of multivariate statistical analysis, was used to map the land subsidence-prone areas. The frequency ratio (FR) was performed as an efficient bivariate statistical analysis method in order compare it with the acquired results from the EBF analysis. The probability maps were acquired and the results of the analysis validated by the area under the (ROC) curve using the testing land subsidence locations. The results indicated that the FR model could produce a 71.16 % prediction rate, while the EBF showed better prediction accuracy with a rate of 73.63 %. Furthermore, the success rate was measured and accuracies of 75.30 and 79.45 % achieved for FR and EBF, respectively. These results can produce an understanding of the nature of land subsidence as well as promulgate public awareness of such geo-hazards to decrease human and economic losses. |
---|