275 
Ice Monitoring 
The initial use of the STAR-1 system was an ice reconnaissance 
program for Canmar Drilling’s fleet of drillships in the Beaufort 
Sea, as part of their ‘Ice Early Warning System’ (Mercer and 
Routledge 1983). Rapid, repetitive large-area coverage was 
provided of the region within which moving ice could affect 
drillship operations. The imagery (e.g., Figure 2) was used to 
detect and identify ice types in the context of their potential threat 
to the drilling operations. The requirement for near real-time 
imagery was one of the major reasons for development of the 
STAR-1 on-board real-time processor, as well as of a broad-band 
downlink system for transmission of digital image data to a base 
station or vessel (equipped with appropriate receiver, recorder 
and film writer). Typically, film strips were coarsely mosaicked 
to provide 1:250.000 scale coverage over areas of several 
hundred km on a side. 
Subsequently STAR-1 (and STAR-2) were used to support 
Canarctic’s MV Arctic during its voyages into the Northwest 
Passage. The MV Arctic has been used as an experimental 
platform for a new ice navigation system (Canarctic 1985) with 
the objective of improving ship performance in ice. The need for 
near real-time radar image downlink was again one of the driving 
requirements. 
The STAR-2 system has been providing operational support for 
the CCG’s icebreaker fleet since January 1990, as described 
above. The system is able to downlink the radar imagery in real 
time to the receiver vessels. It can also compress and burst the 
data from a whole mission to several AES ground stations, from 
which it is retransmitted via satellite to the AES Ice Centre in 
Ottawa. There the data are digitally displayed, analysed and used 
in preparation of daily ice reports and forecasts for general 
shipping. This system has acquired over 2 million square km of 
imagery throughout most of the Canadian Arctic and East Coast 
regions. 
Forest Cover and Depletion Monitoring 
The use of S AR, and INTERA ’s STAR-1 system in particular, for 
forest cover mapping and forest depletion monitoring over the 
past six years has been described elsewhere (Thompson and 
Dams 1990), and is summarized as follows. SAR has received 
increasing attention for its use in forest management over the past 
several years, since being recognized as a valuable tool for 
monitoring forest resources in cloud-covered tropical areas. 
Most of the research to date has been carried out in temperate 
forest regions, and show that certain forest cover types can be 
identified relatively accurately using SAR data, with the best 
results using seasonal data in temperate forests. The use of SAR 
for forest cover parameters has not been extensively researched, 
but work on stand age, biomass and forest disturbance (clearcuts, 
wildfire, regeneration) shows some promising results (e.g., Sader 
1987; Drieman, Ahem and Corns 1989; Dams, Hall and Ahem 
1989; Leckie 1984; Moulton and Peddle 1989). 
Research programs in tropical forest regions have been more 
limited. As it was recognized that temperate forest research 
results may not necessarily be extrapolated to the tropical rain 
forest environment (due to differences in species, species diver 
sity, biomass, crown density, crown geometry, climate and 
moisture, stand structure, management practices, etc.), some 
recent studies, all based on STAR system data, have been carried 
out using SAR in tropical regions: northern Australia (Lowry, 
Van Eck and Dams 1986); Costa Rica (Dams et al. 1987); West 
Malaysia (Ahmad et al. 1988); Congo (unpublished report, 
INTERA Technologies Ltd.); and northeast Colombia (de Mo 
lina and Molina 1989). 
The results of these studies (as reported in Thompson and Dams 
1990) show that STAR data can be used for forest cover type 
mapping; mapping/monitoring of logged/cutover areas; planta 
tion mapping; forest management planning. Tropical forest 
cover can be separated into primary and secondary forest, and a 
range of forest types (e.g., mangrove, peat swamp forest, hill 
dipterocarp forest, high scrub forest, eucalyptus forest), plus 
some individual species. Cutover area and perimeter location can 
be identifed relatively accurately in temperate forests. In tropical 
areas, clearcuts and three levels of selective logging were iden 
tified, as were clearings as small as 0.25 ha. Plantations of pine, 
eucaluptus, oil palm, coconut palm and rubber were mapped 
using STAR data. Finally, the STAR studies showed that forest 
management information from SAR data includes: forest distur 
bance; road, rail and canal location; settlement, permanent and 
shifting agriculture location; physiographic units; forest clear 
ings; recent forest burned areas. An example of STAR-1 data 
from southeast Asia, showing forest clearcutting and plantation 
development, small settlements, roads and other features related 
to forest management in the region, is provided in Figure 3. 
Agricultural Land Use/Soil Capability Mapping 
Several studies have been carried out using STAR data for the 
purpose of agricultural land use mapping and monitoring, or soil 
capability mapping. Most were confidential client reports, and as 
such have not been published. However, the general results of the 
work can be reported. 
All of these studies have been carried out in tropical areas, and 
frequently in conjunction with other work (e.g., Dams et al. 1987 
in Costa Rica). Permanent and shifting agricultural land use is 
readily identified using SAR data, although there may be confu 
sion as to active cultivation versus regenerating plots in shifting 
cultivation areas. Some individual crops may even be identified: 
banana and indian cane in Costa Rica (Dams et al. 1987); tapioca, 
cocoa, rice and pepper in southeast Asia (unpublished report). In 
an urban/rural study in Colombia, the rural/urban interface, 
pasture, wetland, cropland, agricultural settlements, drainage 
and reservoirs, and various urban classes were mapped using 
STAR data (unpublished). Other unpublished agricultural land 
use studies have been carried out in Alberta, Canada; in East 
Malaysia; and in Indonesia. 
In Indonesia, standard resolution STAR data has been acquired 
for much of the country. For the island of Sumatra, where 
standard aerial photographs were of inadequate quality for 
mapping, and satellite images were unavailable or did not pro 
vide adequate detail for the mapping (Buurman 1989), these data