17 
being determined by the—highly specialised—site adaptation. In most cases, 
however, the degree of correspondence between site adaptation and activity 
varies from fairly high to low. Where there is a low degree of correspondence, 
reliability of aerial photographic interpretation is affected negatively. Impor- 
tant additional factors in this context, however, are the cultural and socio- 
economic background, and the location of the activity (Lo 1971). 
Reliability of photo-interpretation of activitiesis variable, and more 
errors are made or uncertainties exist in city centres than in heavy manufac- 
turing areas (Bashour 1970; Dodt 1974). In areas of a high degree of complex- 
ity of activities and/or where site adaptation is not a consistent indicator of 
activities, reliability will be low. Research should pay more attention to in- 
vestigation of the reliability of photo-interpretation, as few publications give 
clear information on this point. A favourable exception is the studies by 
Collins and El-Beik (1971) as well as Collins and Bush (1969). Moreover, 
studies are concerned mainly with urban land use in industrialised countries, 
whereas the specific situation of developing countries is neglected. The question 
should especially be raised of an “optimal” classification of land use: optimal 
from the point of view of the urban planners’ information requirements as 
well as the capability of the photo image (and interpretation technique) as a 
data source of acceptable quality (Collins and El-Beik 1971). Here the 
efficiency (least cost in time and money for the result obtained) of a land use 
survey is paramount. 
If recent aerial photography is available, a land use survey with photo-inter- 
pretation, complemented by a field check, can produce information quickly 
and economically; in potential it competes favourably with conventional 
methods (mainly ground survey) and existing data sources. Insufficient 
information, however, is available of what is actually done in planning practice, 
since photo-derived data are not a standard input. Discussion should there- 
fore be stimulated to provide urban planners with a better insight into the 
potential of aerial photos, and to define in a more precise way the planners’ 
information needs. Only then can the efficiency problem be approached 
realistically. In developing countries, where land use data are frequently not 
up to date, and are soldom available or reliable, where decision time for a 
land use policy is limited, where human and financial resources in urban 
planning are scarce, and organisational tradition is not yet a barrier to inno-