Evaluation of Radar Images for Updating 
Geo-Forestry Information 
Abstract 
As part of the: annual program of the forestry accounting 
service of the Department of Forestry of Quebec, several 
hundred thousand dollars are allocated to the acquisition of 
air photos for the management of forest inventories. With a 
view to reducing the costs of this operation, alternative 
solutions are being explored. To this end, the use of 
remotely sensed satellite images could lead to significant 
savings in data acquisition. 
Currently, optical "images (Landsat-TM or SPOT) are used 
in operations for certain aspects in the processing of geo- 
forestry information. However, optical images are subject to 
the weather conditions prevailing from year to year, such 
that the managers of the program cannot be assured of good 
images every year for all parts of Quebec. This is a major 
limitation, as the managers can hardly tolerate random 
absences of images in the process of updating forest 
inventories. 
Recent research in the field of RADAR images suggests a 
potentially advantageous application in the updating of geo- 
forestry information. RADAR is able to penetrate cloud cover 
and is unaffected by poor weather conditions. We know that 
airborne SAR images allow rapid detection of major 
disturbances (burns, clear cuts and strip cuts). However, 
there are still a number of unanswered questions about the 
potential for detecting partial disturbances in the forest 
cover (such as diameter-limit cutting and selection cutting). 
The objectives of this study are (1) to evaluate the 
potential of airborne SAR images (SAR C/X of CCRS in C-band) 
for the annual updating of geo-forestry information; (2) to 
evaluate the potential of hybridizing airborne SAR images 
with satellite optical images in the forest environment; and 
(3) to evaluate the potential of certain satellite SAR images 
of the ERS-1 and. J-ERS-1 types. The airborne SAR images and 
SPOT and Landsat-TM images used in this project are ortho- 
images resulting from rigorous geometric correction developed 
at the Canada Centre for Remote Sensing. 
A first interpretation of the images reveals that the 
Scenes taken in winter provide a clearer distinction of 
forest cut boundaries than do images taken when snow is 
absent. The snow cover smooths and masks regeneration in 
cutovers, making them easier to distinguish from nearby 
forest cover. 
446 
ABSTR/ 
A techr 
Synthet 
process 
data we 
features 
Much of 
cover w 
weather 
operatio 
updating 
optical « 
KEY WC 
1. INTR 
The Gec 
Division 
and ope 
Geograr 
includin: 
largest 1 
ARC/INI 
1:20,00 
of apprc 
mapbas 
hardcop