280 
SYMPOSIUM PHOTO INTERPRETATION, DELFT 1962 
and mosaics are used to give an overall assessment. This method is primarily 
used in urban analysis [44]. In these cases no interpretation is attempted 
other than preliminary mental recognition of the major surface elements and 
their juxtaposition. 
Stereoscopic interpretation of dynamic forms has resulted in differing tech 
niques. Occasionally, the classification that has been used is so broad (3 to 5 
units) that despite the scale, errors due to recent change do not significantly 
invalidate type boundaries [31, 32, 37, 39]. Another technique is to map with 
a more detailed classification, and correct as far as possible from fieldwork [29, 
32, 35, 36, 39]. This method is sometimes used to the extreme, where all land 
use data is from fieldwork, and the air photographs are used only to outline 
field boundaries and forest areas [29, 54]. The most desirable method is to map 
from recent photographs at large scales (1 : 20,000 to 1 : 10,000). This has 
been found to allow the use of the World Land Use Classification with a 
minimum of fieldwork and with a maximum of accuracy [29, 54, 59]. Fre 
quently it also allows use of a more critical sub-division, such as that between 
orchards and vineyards. 
This leads directly to sequential studies. As outlined earlier, Canada is 
fortunate in this regard and sometimes it has been possible to obtain two 
coverages of air-photographs on which significant changes to the landscape 
have occurred. A typical study was carried out in the Niagara District [21] 
where the encroachments of urban sprawl on high grade farm land between 
1934 and 1954 were plotted. The interpretation, coupled with census data, 
used the easily recognisable road grid as a base unit. The results were expressed 
as percentages per unit area and tabulated on punch cards for electronic 
computation. 
The immense value of sequential study has become apparent in other areas 
of dynamic photo interpretation. Fluctuations in glacier volume, firn line, and 
crevasse patterns over 10 and 20 year periods are being plotted [54]. The 
effects of shoreline erosion and deposition on sea [12], river and lake shores [60] 
have also been traced. The main difficulty in these studies is the contrast in 
scales and season between the two sets of available photography. In the case 
of the glacier fluctuation it includes comparison between obliques taken from 
20,000 feet above sea level with recent verticals taken from 35,000 feet above 
sea level with a negative scale of 1 mile : 1 inch. The difficulties are resolved 
by use of a perspective grid, and optical scale reductions until static points 
coincide, when finally glacier fluctuations can be measured. Such studies have 
brought an increasing realization of the advantages of a planned sequential 
study. Very few of these have been carried out in Canada, though large scale 
planned sequential land use interpretation is projected for the Peace River 
area [54] where power developments are expected to alter land use patterns 
markedly. A small scale planned sequential study is proposed in the vicinity of the 
new Saskatchewan Dam [54], where increased irrigation possibilities and newly 
available power are expected to have considerable impact on present land use.