Conditions of landslides occurrences in the past 
in the WAJIMA region are 
(1) the angle of inclination is greater than 12.5 
degrees 
(2) the elevation is-under 300m 
(3) main geology types are sandstone, mudstone 
and conglomerates 
Landslide area corresponding to these conditions 
were 72%. This indicated that landslides may oc- 
cur in regions where these conditions are satisfied. 
4.3 NVI data in landslide regions 
We are present studying the NVI data calculated 
from the satellite data to understand the relation 
to landslides. We are investigating the distribu- 
tion of frequencies from two groups of NVI data 
in the NOTO region. The data from one group 
is data from the areas landslide occurred in the 
past and pointed. dangerous zone. The other is 
data from non-landslide areas. The average value 
of each NVI is shown in Table 2. 
Table.2 NVI data of each Categories 
  
  
Items Total count Average values 
number of NVI 
Dangerous 5631 80.6 
Occurred 14818 79.5 
Dangerous--occurred 20449 79.8 
Non-landslide 129330 75.1 
  
As seen from Table 2, the value of NVI data in 
non-landslide areas is relatively small compare to 
dangerous area and landslides areas. One reason 
to the NVI values of non-slide area is smaller than 
those of landslide areas is that the non-landslide 
areas include urban areas. However, it has been 
said that the occurrence of landslides is consid- 
ered to be closely related to the content of under- 
ground water [3],[5]. The content of underground 
water is considered also to be closely related to 
ground surface vegetation. 
Therefore, we need to investigate further to un- 
derstand the relation between value of NVI data 
and landslide areas. 
5. CONCLUSIONS 
An extraction of landslides areas using NVI val- 
ues which were transformed from MSS data and 
Thematic maps has been performed. This analy- 
sis was mainly made by the Image-Map database 
system. As a result, it was shown that some char- 
acteristic features in landslide areas can be eas- 
ily extracted by using the Image-Map database 
system which included satellite data (NVI), land- 
slides area data, geological data, elevation data 
and inclination angle data. Moreover, it was found 
that the average value of NVI data in landslide 
area was larger than non-landslide area. 
References 
[1] D.J.Sauchyn and N.R.Trench, Landsat Ap- 
plied to Landslide Mapping, Photogrammet- 
ric Eng. and Remote Sensing, Vol.44, No.6, 
pp.735-741, 1978. 
[2] Geographic survey Institute of Japan, Man- 
ual of Digital National Land Information, 
1986. 
[3] K.Kawamura and S.Suzuki, Landslide Pre- 
diction Method Considerd with Characteris- 
tics of Run-off in Surrounding Valleys, Proc 
of 36th Symp on Soil Mechanics and Foun- 
dation Engineering. 
[4] M.Shikada, T.Kusaka and K.Miyakita, An 
Analysis of Landslide Areas Using an Image- 
Map Database System, Proc of 12th Asian 
Conf. on Remote Sensing, Vol.II, pp.1-6-1 to 
I-6-6, 1991. 
[5] T.Kusaka, M.Shikada and Y.Kawata, Ex- 
traction of Landslide Areas Using Spatial 
Features of Topographic Basins(submitted), 
Proc. of IGARSS'92, 1992. 
388