The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part B7. Beijing 2008 
. 
61 s “» 64*““ 66"“" (>7»» 69°“" 70 s “» 
arwR vr&n. ar* mnt ivwt-. 
M»|> Sc*te 
H I 98 7. Urban Area 
1997, Urban Area 
2007, Urban Area 
Figure 5. Urban boundaries. 
Urban boundaries of Istanbul for every year are shown in Figure 
5. The Figure shows urban areas expension. Red area presents 
the 1987 urban boundary, blue area symbolizes the 1997 urban 
boundary, and the white area represents the 2007 urban area. 
3.1.2 Thermal Channel Classifications 
If one wants to detect UHI with satellite images, satellite 
image’s thermal band must be used. Landsat 5 TM satellites 
sensor’s thermal band is used for classification of the images. 
Density slice algorithm is used for classification of the image 
which is Landsat satellite’s sensor’s thermal channel. 
When the satellite image has warm colors like red, orange or 
yellow after the classification, these colors present warm areas 
in the image. If the image has cool colors like blue or green, 
they represent the cold areas of surface. 
1997 Landsat 5 TM Termal Bant 
35Twu W4snt mi mjB 
Figure 7. Thermal band classification of 1997 image. 
2007 Landsat 5 TM Termal Bant 
antra armi m< mm mcu 
70**" 
64*** 66*** 
eœœesi::™ 
70*™’ 
antra 
UHI in Istanbul is shown in figure 6, figure 7 and figure 8 
which is the classification of the thermal channel. The 2007 
Istanbul image in figure 8 represents more definitively urban 
heat islands than other images. 
Figure 8. Thermal band classification of 2007 image 
The boundary of the urban area is added to the thermal 
classification of Istanbul image in Figures 9, 10 and 11 
Figure 6. Thermal band classification of 1987 image 
Figure 9. Istanbul image’s thermal band is classified and urban 
boundary is added for 1987.