Nguyen Dinh, Duong 
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
i- Based on the research undertaken in the framework of NASDA RA pre-launch algorithm development for land cover 
1S mapping by ADEOS-II GLI data, the author proposes the following land cover classification scheme. In this system 
i land cover is divided into categories with different leaf coverage and water component percentages. Each class is coded 
d by two digits, the first one indicates the dynamic component and the second one is linked to the static component. The 
d dynamic component is ranked into 9 groups so that total number of classes for single date data analysis will be 100. In 
le | combination with multi-temporal data, the final land cover map can have 255 codes for different categories. Table 2 
le shows a proposal of the classification system for single date remote sensing data. 
s Class code Dynamic component Static component 
2 91 Vegetation with 70 — 100% Broad leaf forest 
92 caverage Forest plantation 
93 Needle leaf forest 
94 Mangrove forest 
or 95 Cropland 
m | 96 Other grass type vegetation 
es | 81 Vegetation with 50 — 70% Broad leaf forest 
er 82 coverage Forest plantation 
m | 83 Needle leaf forest 
9 | 84 Mangrove forest 
7] 85 Cropland 
ZZ 86 Other grass type vegetation 
8 | 71 Vegetation with 30 — 50% Broad leaf forest 
72 caverace Forest plantation 
S 73 Needle leaf forest 
74 Mangrove forest 
T 75 Cropland 
of 76 Other grass type vegetation 
T, 61 Vegetation with 10 — 30% Shrub land 
62 coverage Woodland 
lis 63 Wetland Shrub 
ita 64 Cropland 
65 Other grass type vegetation 
31 Vegetation with 0 — 10% Grassland 
52 Coverage Wetland Shrub 
53 Cropland 
41 Non-biotic cover Rock 
42 Sand 
| 43 Cloud 
44 Construction 
| 45 Dry soil 
46 Cloud or topographic shadow 
  
  
  
  
  
  
International Archives of Photogrammetry and Remote Sensing. Vol. XXXIII, Part B7. Amsterdam 2000. 987