Few of the faults known to exist in the test area could be detected on ERTS 
images. Again, lack of definition causes disappearance of most of the subtle 
features which otherwise would make detection possible. Exceptions are large 
faults associated with features having pronounced physiographic expression, like 
straight alignments of larger-order streams or straight boundaries of major geo 
logical units. A prominent lineament NE of Aliaga, which was observed on ERTS 
images, does not appear on existing maps. There is no field evidence to support 
the presence of a fault so that the actual nature of this feature remains obscure. 
Joint-patterns are known to occur in several rock units of the test area. Though 
visible on 1:30, 000 scale aerial photographs, these patterns could not be detected 
on ERTS images. 
COMPARISONS 
To establish the accuracy of the interpretation an evaluation of the classification 
results had to be made. Basically this was done by comparing the final map pro 
duct with the existing 1:100, 000 Geological Map of the Montalban Region, and 
with information obtained from 1:30, 000 scale aerial photographs. 
On the aerial photographs at least 20 rock stratigraphic units can be distinguished 
and mapped. On the photogeological map of the Montalban Region [l ] these 20 
units have been grouped into 11 time stratigraphic units. The 1:250, 000 geologi 
cal map prepared from ERTS imagery shows 7 geological units. The two classi 
fication schemes are compared in Table L 
It is seen that the main deficiency of the ERTS-classification is that it fails in 
sub-dividing the Trias sic and the Lower Cretaceous. 
In order to investigate the accuracy of boundaries, the photogeological map was 
reduced to 1:250, 000 scale and the number of geological units was reduced to the 
number of units present on the ERTS- interpretation. There was a good general 
agreement between the two maps. Summarizing the results of the comparison it 
can be stated that the number of mappable units is reduced considerably if the 
ERTS classification scheme is compared with the scheme derived from aerial 
photographs, but that consistently mappable boundaries on ERTS imagery show 
a good correlation with air photo data.