e Digital 
idly, the 
ry nature 
handling 
| process 
'chniques 
preciation 
ly, spatial 
dating of 
isiderably 
tography. 
| of the 
ore timely 
icular will 
his is an 
AacMillan 
ed by the 
typical of 
PA 
ill allow 
w spatial 
y manner 
xcessarily 
will be a 
inted and 
cess. As 
ves, data 
cification 
ted over 
Charlotte 
The data 
Aperture 
d satellite 
lata), and 
bands of 
satellite. 
'esolution 
ach HRV 
the 10m 
de (PLA) 
"3 um; 2) 
operating 
ution and 
|, 0.67110 
For the 
present study, PLA data was acquired May 
27, 1993 and MLA data on August 21, 
1992. 
The SAR data set includes three flight lines 
from the CCRS airborne C/X SAR, one 
satellite SAR image from the European ERS-1 
system, and two simulated RADARSAT Fine 
Resolution Mode images. 
Narrow mode airborne C/X SAR data were 
acquired on July 26, 1993, in C-band (5.66 
cm wavelength) and X-band (3.24 cm 
wavelength), dual polarization (HH and VV), 
with a nominal resolution of 6 m x 6 m and 
incidence angles ranging from 35? in near 
range to 76? in far range. Two of the flight 
lines were planned to correspond with 
RADARSAT orbital characteristics and look 
directions. 
One ERS-1 SAR image was collected April 
10, 1993, during the two week Roll-Tilt 
Mode (RTM) acquisition period. The ERS-1 
SAR operates exclusively in C-band, VV 
polarization. The Roll-Tilt orbit phase 
permitted data collection with an incidence 
angle of 35? (mid-swath) and a nominal 
resolution of 30 m (The ERS-1 standard SAR 
mode operates at 23? incidence angle). The 
imaging geometry of the RTM data set is 
close to RADARSAT's Fine Resolution Mode 
(Figure 1). 
RADARSAT Fine Resolution Mode simulation 
images were produced with C/X SAR HH 
polarized images as input data. The airborne 
data were modified using the INTERA 
SARPAC software package (Parashar and 
Wessels, 1989). These data were resampled 
to 9 x 11 m resolution and radiometrically 
altered to simulate the expected radiometric 
characteristics of RADARSAT imagery. 
The base map information used is the BC 
Provincial TRIM data set. It consists of a 
digital elevation model (DEM) and digital 
planimetric information at 1:20,000 scale in 
the NAD83 coordinate system. 
Figure (3a) shows the ortho-corrected SPOT 
PLA, (3b) one flight line of the airborne C-HH 
C/X SAR data (registered to the SPOT PLA), 
449 
(3c) the TRIM DEM, (3d) and a composite 
image of the SPOT PLA (red), the C-HH C/X 
SAR (green), and the TRIM DEM (blue). 
4. METHODOLOGY 
The project developed separately for the 
optical and radar segments. The refinement of 
methodologies for the use of SPOT data was 
considered as the operational demonstration 
segment. The testing and development of 
techniques for the use of satellite SAR data 
was designated as the applied research 
segment. 
Spatial resolution was the main criteria for 
data selection. The SPOT PLA band and the 
RADARSAT Fine Resolution Mode (simulation) 
were selected due to their 10 m and 11-9 x 9 
m resolution respectively. Given new 
software developments for satellite ortho- 
rectification, it was felt that these sensors 
could reach acceptable mapping accuracies. 
4.1 The SPOT Segment 
The technique employed with the SPOT PLA 
data included: 
A) Ortho-rectification and enhancement 
of the contrast between forest stands 
and cutblocks; 
B) Integration of the ortho-corrected 
satellite data files with the GIS 
environment; 
C) Interpretation of SPOT data for the 
extraction and mapping of new roads 
and  cutblocks within the GIS 
environment. 
4.2 The SAR Segment 
Two main questions were investigated: 1) 
Can SAR effectively distinguish between 
forest stands, cutblocks and roads? 2) Can 
SAR meet the accuracies required for 
mapping these features in mountainous 
terrain? 
The first question is being investigated using 
the airborne SAR and the simulated 
RADARSAT Fine Resolution data. The