nbul 2004 
>w and or 
Figure 4 
centre of 
ast to the 
ngst other 
and new 
orth of the 
  
and 1995 
2003 (red) 
important 
he area of 
ogoda, La 
nunication 
  
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B1. Istanbul 2004 
system is easily seen by the blue tones, whereas the new 
transversal east-west road stands out, contributing to the 
enhanced connectivity within the wider metropolitan area. 
As mentioned previously, the classification exercise for 
determining the land cover classes for the 2003 images has, at 
the time of writing, only been completed for one part of the 
SPOT Scene, covering just 172 km? of the MAB. The full 
results will be presented the ISPRS Conference in Istanbul in 
July 2004, however it is worth mentioning that some nine land 
cover classes have been isolated resulting from the supervised 
classification methodology used and described under Section 
2.3 of this paper: shadow; 3 industrial classes; residential; soil; 
green; shallow water; and occupied area. 
Figure 6 illustrates a detail of the final map resulting from the 
aforementioned. supervised classification process, focusing on 
the industrial areas of the Zona Franca, in Barcelona, lying to 
the north of the Llobregat River and the adjoining Logistic 
activities zone (ZAL), within the municipality of El Prat de 
Llobregat, lying to the south. 
      
       
X 
Figure 6: Example of the resulting supervised classification 
© SPOT Image Copyright 2003, CNES 
4. DISCUSSION AND CONCLUSIONS 
4.1 Monitoring urban sprawl 
Official figures relating to the growth in population of the 
Metropolitan Area of Barcelona for the five-year period 1996- 
2001 indicate an overall increase in population of some 
0.295%. This demographic change has by no means been 
uniform, as is highlighted by Table 3. While the population of 
the City of Barcelona has decreased by some 0.326%, the 
remainder of the Metropolitan Area under the jurisdiction of the 
MMAMB saw an overall increase in population of just over 
1%. 
This level of growth is more modest than that suggested by 
Table 2, but is probably nearer the truth for a number of 
reasons. On the one hand the spatial extent of the MAB is just 
one part of the much wider territory subjected to the satellite 
image interpretation (remote sensing) comparison between 1995 
and 2003. On the other hand this comparison identified just 
three very broad classes of land cover: occupied areas (implying 
some form of human intervention), green areas and soil. 
Furthermore the two images used for this comparative exercise 
are in raster data format. For greater accuracy they would need 
to be in vector data format enabling the delimitation of 
polygons and the determination of exact areas. Once this 
conversion procedure has been carried out upon the SPOT 
Images used in this project, the research team will then be in a 
position to quantify with precision the level of urban growth 
and development experienced between 1995 and 2003 within 
the limits of the MAB, and make the judgement as to what 
extent this growth can be constituted as urban sprawl. 
  
  
  
  
  
  
  
  
  
1996 2001 % change 
Barcelona 1,508,805 1,503,884 -0.326 
Rest of MAB 1,328,735 1,342,041 +1.001 
Total MAB 2,837,540 2,845,925 0.295 
  
Table 3: Change in population for the Metropolitan Area of 
Barcelona 1995-2003 
As yet, the research project referred to here is not sufficiently 
advanced to offer conclusive results concerning the 
quantification and qualification of urban sprawl or urban 
growth within and on the edge of Barcelona’s metropolitan 
area. However the emerging results clearly corroborate the 
hypothesis of increased land consumption beyond the central 
core of the MAB. 
In this sense the morphological approximation to the 
delimitation of urban and/or metropolitan systems through the 
interpretation and analysis of satellite images, particularly 
incorporating a dynamic approach by comparing two points in 
time, would appear to offer a clear advantage over more 
functional approaches. By simply superimposing the carlier and 
later of two images, as carried out in Section 3.1 and illustrated 
by Figures 4 and 5, the professionals working in the realm of 
urban and regional planning are provided with a wealth of 
information. This enables them to determine at a glance the 
arcas where development pressures have been most strongly 
experienced. 
4.2 Conclusions 
Part of the reasoning behind the invitation extended to the 
UPC/MMAMB from the CNES and SPOT Image to participate 
in the SAVP exercise was to gain feedback on the potential 
applications of Spot 5 data for the purposes of urban planning. 
With this in mind, although the final results of the project are 
still to be reached, it seems appropriate to offer some comments 
concerning the enhanced characteristics of the Spot 5 Images, 
resulting from the increased resolution. 
The difference between the 2003 and the 1995 images, in terms 
of increased visual clarity and quality, is patently obvious from 
Figure 7. This illustrates the built-up fabric of Barcelona, 
incorporating the city’s historic core to the north-east adjoining 
the former Port Area, today with a leisure boat marina and other 
recreational/leisure uses, and the grid-ironed 19" Century 
extension to the city to the north-west. The clarity of the urban 
fabric in the southern portion of the figure contrasts sharply 
with the more blurred nature of the northern section. 
Figure 7 clearly provides very real and local evidence of the 
potential for a deeper understanding of the dynamics of urban