Landslide susceptibility mapping using frequency ratio and binary logistic regression

Citation

Kalimuthu, Hemalatha (2019) Landslide susceptibility mapping using frequency ratio and binary logistic regression. Masters thesis, Multimedia University.

Full text not available from this repository.
Official URL: http://erep.mmu.edu.my/

Abstract

This research explores the landslide detection and susceptibility analysis on areas highly susceptible to landslides in Cameron Highlands, Malaysia. Cameron Highlands is a main tourist attraction in Malaysia among all the other highlands due to its beautiful landscape on its tea plantation, farms and cooling atmosphere. Unfortunately, this highland is an area prone to soil erosion which includes deep failure of slopes, rockfalls and shallow debris fall. There were 89 landslides found within the area of interest through ground survey and historical database. 63 landslides were used as the training data while the remaining 26 were used for testing. Two methods were explored to produce the map on the area susceptible to landslides namely frequency ratio method and logistic regression method. The finalized independent variables used as the causative factors are elevation, slope, aspect, land use, average rainfall, distance to roads and distance to rivers. The DEM (Digital Elevation Model) data was obtained from ASTER GDEM (Advanced Spaceborne Thermal Emission and Reflection Radiometer Global Digital Elevation Model) project by the Japan Space Systems (JSS). Land use, road and river data was obtained from Malaysian Centre for Geospatial Data Infrastructure (MaCGDI) while the average rainfall data was obtained from National Climate Centre, Malaysian Meteorological Department. The interpolation studies on lower resolution DEM shows that ASTER GDEM data is able to produce highly reliable and accurate higher resolution DEM model using the universal kriging interpolation method. The test on correlation coefficient among the variables shows that the independent variables are not highly correlated or dependent among each other with values less than 0.7 and are ideal to be used in this research. Logistic regression method shows higher accuracy of 92.31% compared to frequency ratio method that produced 86.15%, thus it is concluded to be a better predictive modelling method in constructing landslide susceptibility map in Cameron Highlands.

Item Type: Thesis (Masters)
Additional Information: Call No.: TK6565.D4 H46 2019
Uncontrolled Keywords: Frequency modulation detectors
Subjects: T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK5101-6720 Telecommunication. Including telegraphy, telephone, radio, radar, television
Divisions: Faculty of Engineering (FOE)
Depositing User: Ms Nurul Iqtiani Ahmad
Date Deposited: 26 Sep 2024 04:05
Last Modified: 26 Sep 2024 04:06
URII: http://shdl.mmu.edu.my/id/eprint/12986

Downloads

Downloads per month over past year

View ItemEdit (login required)