| Summary: | 碩士 === 國立成功大學 === 水利及海洋工程學系碩博士班 === 101 === In this research, we use ArcGIS by NDVI (Normalized Difference Vegetation Index) to
judge landslides in 4 periods of Formosa II satellite images. By using 5-meter DEM (Digital
Elevation Model) which is drawn by Soil and Water Conservation Bureau in 2003 and the
rock data which is established by National Cheng Kung University Disaster Prevention
Research Center (DPRC) in 1995, we can get slope, aspect, elevation and rock data.
Individually discussed this four factors which their effect of collapse.
For researching the data of collapse during the first period (during 2009 August to 2011
Feb.) to the fourth period (during 2012 March to 2012 July). Making the first period data of
collapse be the datum and make the other three periods subtract the first period. Thus, we
can get the benchmarking of this three periods. By discussing this benchmarking, we get a
result that slope and elevation have a tendency of direct ratio to collapse ratio. Using
dangerous value method to calculate potential collapse. By the factor of geography, we can
calculate collapse index and weight. Combining the collapse index and weight, we can get
the empirical formula of potential collapse in research area. Input the data of research area
to empirical formula, we can get potential collapse of this area. Integrating potential collapse
of these three periods. Thus, we can get high potential collapse of these area. In researching,
we use this method to mark higher risk of collapse in the research area and have a result:
elevation bigger than 2500 meter, rock which is Pilushan Formation, slope bigger than 30
degree and aspect is south and south-west have more possibility to occur collapse.
|
|---|
