anbul 2004 
nakomai 
  
a 144 
valuate the 
green, blue 
| the nadir- 
red dots in 
rve around 
ometer can 
descending 
i is 614 km, 
We first 
-P area for 
liometer is 
ere moved 
cated in a 
rban areas. 
gap of one 
ime time. 
m with a 5 
iCP. The 
) km with a 
|-frequency 
) km with a 
ment. One 
can be used 
to validate 
y, pointing 
91 second, 
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B1. Istanbul 2004 
  
Photo 2. Overview of GEONET GPS receiving station 
3.4 Along-Path GCP Using Automatic GPS Stations 
Using the along-path GCP in Figure 8, there are only three 
observations during the 46 day recurrent ALOS orbit because 
the along-path GCP is only covered by three paths (RSP 67 — 
69). That is not enough to evaluate above items, especially in 
the initial calibration phase. Therefore, we examined the 
automatic GPS receiving stations for use as GCPs. 
In the Japan area, GSI maintains the GPS Earth Observation 
Network (GEONET). This network was basically established 
to monitor crustal deformation. Currently over 1,200 automatic 
GPS receiving stations have been installed and are being 
continuously monitored. The average distance between each 
station is 10 km, so these stations can be used as GCPs for 
addition, this data is available worldwide via the Internet. 
Available real time data with image acquisition of PRISM after 
launch of the ALOS will be distributed worldwide. 
Photo 2 is an overview of the GEONET stations with a height 
of about 5 m. Figure 9 depicts the locations of all GEONET 
GPS stations in Japan. We tried to identify the 133 stations as 
the test data using aerial photographs, although some could not 
be clearly identified. 
The red dots in Figure 9 indicate the locations of selected the 
133 points, which are along ALOS paths i.e, RSP 67-69, 73-75, 
and 82-84. From the results, 25% (33 points) could be 
identified in aerial photographs, 6896 (91 points) were unclear, 
and 7% (9 points) could not be identified. This result is 
dependent on the scale of the utilized aerial photographs. 
: PRISM. Their absolute accuracy is very high and stable. In 
17 
    
Figure 9. Locations of GEONET station and selected points 
  
Figure 10. Example of the identified GEONET in aerial photo 
Therefore, the success rate will be increased using high-scale 
aerial photographs. Figure 10 is an example of a station 
identified in a 1:4,000 scale aerial photograph. 
3.5 Along-Path GCP to Evaluate Thermal Conditions 
Figure 11 displays ALOS paths RSP 67 to 69, from north to 
south latitudes around the East Asia and Oceania regions, and 
the locations of the prepared GCPs. An along-path GCP must 
be required for over 91 sec to evaluate the effects of variability 
of thermal conditions, especially the thermal distortion of CCDs 
that is a function of latitude. It would be convenient to use 
identical orbits to reduce other factors. Therefore, we estab- 
lished a GCP around southwest Australia. 
Figure 12 indicates the GCP prepared using a DGPS in 
southwest Australia with ALOS paths RSP 67 to 69. Three 
GCP areas were established between the cities of Kalgoorlie 
and Esperance. Each area was composed of only one line from 
west to east with less than a 10 km step. These areas could not 
be covered by nadir observation of RSP 67 because they were