416 
1974 imagery near Fort Good Hope and the Ramparts (Figure 4). 
These plots represent our best example of the positive ability 
of "Quick-Look" imagery to monitor the break-up, Imagery for this 
area is available for May 21st and 22nd during the 1974 break-up. 
The fit of shorelines and islands permits comparison of 
the changes in ice cover and open water. The growth of open water 
areas is visible at localities marked A, B and C on the centre 
plot. These are of sufficient size and agree well enough with 
the base map to permit interpretation of a recognizable break-up 
condition (stages 3 and 4) .Areas labelled D and E indicate areas 
where the interpretation has changed from the first plot to that 
of the following day. Area E illustrates this type of problem 
very well, a problem which was encountered in many other localities. 
As well as the loss of detail, the "reappearance" of an island 
known to exist, on the plot of May 22nd, points out a flaw in the 
interpretation of the previous day. 
In assessing this ERTS study of the break-up of the 
Mackenzie River, the results should be judged in terms of the 
anticipated use of the data. If monitoring the progress of break-up 
is to be an aid to field operations, the basic requirement is to 
provide this information within six to seven days from the date 
of the satellite pass. This restricts the user to "Quick-Look" 
imagery with its reduced resolution, as delivery schedules for 
standard ERTS imagery do not permit quasi-real-time use. The 
fundamental problems of the eighteen-day repeat cycle and the risk 
of cloud cover rule out reliance on ERTS imagery. During some 
break-up seasons a complete ice cover may exist during one 
satellite pass, while the cycle eighteen days later shows a 
completely ice-free condition. The use of an aid to interpretation 
such as the Zoom Transfer Scope tends to remove the interpretation 
from staff in the field to others in the office. 
Because of the problems here outlined, and because of 
other problems linked to conditions in the complete drainage basin, 
it seems that other sources of information (pilot reports, radio 
links with settlements along the river) are more reliable for the 
short-term planning of field operations. Our experience leads us 
to believe that ERTS imagery presently available cannot replace 
our current methods of recording the break-up and is of limited 
usefulness as a supplement. Further work should be carried out 
using ERTS imagery from the late summer low flow conditions. We 
will continue to study the difficulties of shoreline and island 
interpretation, by using new maps as they appear, and all the 
available air photography. Hopefully ERTS imagery at a time of 
low flow may assist in eliminating or explaining the difficulties 
we experienced, thereby helping to establish the use of ERTS imagery 
in hydrological problems of this type.