Table 2. Available sensors that can be used for burnt area mapping at high and medium spatial resolutions 
BLUE GREEN RED 
Medium Resolutio 
MSS 0.45-0.52 0.5 - 0.6 0.6 - 0.7 
MSU-SK 0.5 - 0.6 0.6 - 0.7 
WIFS 0.62 - 0.68 
High Resolution 
MSU-E 0.5 - 0.6 0.6 - 0.7 
LISS-3 0.52 - 0.59 | 0.62 - 0.68 
HRV 0.52-0.59 | 0.62-0.68 
TM 0.45-0.52 | 0.52- 0.6 | 0.63 - 0.69 
NIR NIR NIR MIR TIR 
0.7 - 0.8 0.8 - 1.1 
0.7 - 0.8 0.8 - 1.1 
0.77 - 0.86 
0.8 - 0.9 
0.77 - 0.86 1.55-1.70 (na) 
0.77-0.86 
0.76 - 0.90 : : 10.04-12.5 
    
(na) Not available. NIR (Near infrared); MIR (mid-infrared); TIR ( Thermal infrared) 
Results presented by San Miguel er al. (1998) show an 
improvement in using the BI for burnt area discrimination 
in comparison to other transformations such as the 
Tasseled Cap (Kauth and Thomas, 1976), or the Principal 
Component Analysis. 
Figure 2. Land-cover polygons from CORINE overlaid on 
the burnt mask. (a) Whole image (b) Close up of the fire 
area. 
II RESULTS AND CONCLUSIONS 
The total forested area burnt in this fire was estimated in 
667.72 ha, which corresponds to 90% of the official 
estimates estimates, i.e. 744 ha including forests and 
shrubs. Underestimation of burnt grasslands occurred 
because of the late acquisition date of the post-fire image. 
Autumn rains permitted grassland regeneration after the 
fire which restrain the discrimination of burned and 
unburned grasslands. 
The results of the intersection of the area mapped as burnt 
with the CORINE layer are presented in Table 3. 
  
Table 3. Results of the intersection of the burnt mask with 
the CORINE land-cover layer 
  
  
  
  
  
  
Land Cover CORINE code Burnt Area (ha) 
Broadleaf forests 311 362.32 
Sclerophyllous 323 305.40 
vegetation 
Natural grassland 32] 20.29 
Discontinuous 112 : 5.93 
Urban Fabric 
  
  
  
  
172 Intemational Archives of Photogrammetry and Remote Sensing. Vol. XXXII, Part 7, Budapest, 1998 
Fig 
on | 
Ho 
maj 
that 
det: 
are: 
Is p 
In 
pro 
ima 
amc 
will 
the 
dan 
  
Fig 
affe