Fig.3 Imaged elevation data by 30m mesh size 
(Monzen district) 
This data was made by us on handiwork 
  
Elevation Data 
' 
River 
| 
Watershed |<” 
  
  
  
  
  
  
  
  
    
      
    
    
Inclination Angle 
Gelogy 
Shape Factor 
TM Band6 
NVI 
  
  
  
  
Fig.5 Flow of study 
data) has the information of temperature. It is 
shown that TM band 6 data represent temperature 
of ground surface. Landslide closely relates un- 
der ground water. Generally speaking, water con- 
tents in landslide area has higher than that of 
non-landslide areas.In other words,temperature 
of ground surface in landslide areas has lower 
than that of non-landslide areas. TM band 6 data 
was used to estimate indirectly landslide areas. 
3.3 NVI DATA 
It is well known that the NVI data shows vitality 
of vegetation. Calculation of NVI is as follows. 
NVI=(NIR-VIS)/(NIR+VIS) (1) 
In this equation , NIR is the near infrared band 
data and VIS is visible band data . The reason 
why we use NIR band data is because it reacts to 
the amount of chlorophyll in vegetation. Gener- 
ally speaking, as NVI increases , mass of vegeta- 
tion or the vitality of that vegetation also in- 
788 
  
Fig.4 Imaged elvation data by 50m mesh size 
(Monzen district) . Elevation data was 
made from DEM. 
DEM was establised by Geographic 
Survey Institute of Japan. 
crease. There are many index to represent the 
condition of vegetation by using satellite data. 
We adopted equation (1) to analyze the relation 
between the vegetation index and occurrence of 
landslides. We used band 4 as NIR and band 3 as 
VIS in the equation (1). NVI data calculated 
from the above equation is greater than or equal 
to -1 and less than or equal to +1. Data of equa- 
tion (1) needs to be changed to an integer. Equa- 
tion (2) was used to change a decimal NVI value 
to an integer value. 
NVII={(NIR-VIS)/(NIR+VIS)}*100+100.5 
(2) 
4. METHOD AND TECHNIQUES 
4.1 FLOW OF ANALYSIS 
Figure 5 shows the outline of analysis.There are 
five major part in this study. They are generation 
of elevation data (in this paper we adopted DEM 
data which was made by Geographic Survey In- 
stitute of Japan),generation of virtual river from 
elevation data , generation of watershed from vir- 
tual river and extraction of characteristics of 
landslide areas in each watershed. In this paper 
we describe the characteristics of inclination 
angle , geology, shape factor F ,TM thermal band 
data and NVI data within watershed. 
Major data of the research is the elevation data. 
GIS gives us virtual rivers from elevation data. 
We used GIS software of ARC/INFO ,however, 
algorithm to extract rivers from elevation data 
does not open to users. Our object is not evalua- 
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B4. Vienna 1996