365 
égrer la 
infrared scanner was developed several years ago principally for forestry use 
by Computing Devices of Canada Ltd. 
heur des 
aïeuls ne 
nts observés 
ent d’un 
le. Ces 
qui pourront 
forêt. 
These IR systems have been developed primarily to suit the fire 
detection and fire mapping needs of agencies responsible for fire control. 
The fire researcher's needs and constraints are somewhat different from those 
of a fire control agency with respect to forest fire surveillance. I shall 
discuss an approach to forest fire behavior research currently being used in 
various regions of Canada by the Canadian Forestry Service. 
THE RESEARCH PROBLEM 
rikmaessig 
rt, um 
Richtlinien 
Data is now being gathered by the Canadian Forestry Service on the 
behavior of forest fires during their initial growth stages in many kinds of 
fuels so that predictive fire behavior guidelines can be developed. Such 
guides are required to augment the uses already being made by Canadian fire 
control agencies of the Canadian Forest Fire Weather Index (Anon. 1970), a 
relatively new system of forest fire danger rating. 
t und wird 
ntensivitaet 
ithodenroehre- 
L. 
Because of the logistical problems of transporting fire researchers 
to most fires and of measuring such fire behavior parameters as rate of spread 
and intensity on wildfires whose occurrences are to some extent unpredictable 
in time and space, progress in this field has been rather slow. It was felt 
.gkeit und 
•enden 
»reitung 
ilative 
len, mit 
ttegrieren 
that an IR imagery system mounted in an aircraft could overcome both the 
problem of getting researchers of many fires on the ground and the problem 
of instrumenting fires with suitable sensors for measuring fire spread and 
residence time. 
Besides the need to monitor many wildfires burning simultaneously 
within a radius of 250 km of a dispatch base, prescribed fires in logging 
: dem 
Lehe 
)as 
re bestimmt 
îlle von 
lie 
residues and deliberately ignited research test fires in wildland fuels must 
be monitored. These latter fire situations differ from the wildfires in that 
they can be more easily instrumented with ground sensors for fire behavior 
measurements but IR imagery is proving to be a very useful adjunct to other 
means of data collection in which high costs and risk of equipment breakdown 
and data loss always exists. 
Our fire research requirements of an IR system differed from a fire 
control agency's in the following ways. An instrument designed specifically 
for fire detection was not required, eliminating the need for a wide total 
angle of view, which usually is about 120 for such systems. The ability to 
use a much narrower total field of view enabled us to use an imaging system 
:elligence 
, 1968) 
1962 into 
ascribed a 
îtion of a 
. (1972) 
using IR 
apper" 
in which the IR sensor sees the total field at once instead of having to use 
the scanning principle to achieve the large total view angles demanded by fire 
detection systems. One common problem with IR scanning systems which this 
avoided was the difficulty in matching scanning rate and aircraft ground speed 
to avoid image distortion. Our problem was to be confined to mapping fires 
generally less than 40 ha in size, so we had to choose a system which could 
provide this imagery from workable flying heights.