The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Voi. XXXVII. Part Bl. Beijing 2008
259
amplitude dispersion index of the point targets in a time series
of SAR images. Image coregistration was generated prior to
identifying the PS candidate. Each pair of interferogram was
processed with respect to the master image. Shuttle Radar
Topography Mission (SRTM) Digital Elevation Model (DEM)
which has three arc-second resolution was used to remove the
topographic phase and generate the initial differential
interferograms. Finally, displacement of PS candidate was
estimated. It involved phase unwrapping as well as correction
for linear deformation, DEM error, baseline error and
atmospheric noise. The PSInSAR results were interpreted and
compared to investigate with the effect of groundwater over
extraction to urban deformation. The GIS software is used to
interpret the PSInSAR result. The combined methods between
PSInSAR and GIS allow an integration of information from
various sources and hence improve the efficiency for
interpreting the data.
Figure 5. SAR image at test site
To compare with DInSAR methods, which is normally used to
ground subsidence monitoring, we were used CRC-DInSAR
software. Theoretically, differential radar interferometry can
survey deformation to high level (up to 1cm) of accuracy over
large spatial extents with high spatial resolution. [Ge et al,
2007]. However, source of DInSAR has limitation of temporal
baseline and perpendicular baseline. In our research, test field is
located in the city area and deformation of ground is small per
year. So it is too difficult to detect the deformation in the city
area. Figure 6 is the interferogram of test area using DInSAR.
Figure 6. Interferogram of test site using DInSAR
From the interferogram shown in figure 6, it is not able to
obtain much useful information of land deformation around test
region. For long term deformation monitoring, PSInSAR
technique is more powerful, precise and fully operational tool
than DInSAR method [Kampes 2006].
Photogrammetry used two pairs (10/1994, 10/2004) of aerial
photography for investigation of small area deformations which
are detected from PSInSAR results. Each projects of aerial
photo are processed and detected using SOCETSET software.
Ground Control Point (GCP) was achieved 10mm resolution
from GPS surveying for external orientation. Building
extraction of subsidence area was conducted by feature
extraction method in SCETSET. The ArcGIS software was
used to interpret the building deformation result. The aerial
photo information and image are shown in table 2 and figure 7.
DATE
ID
Frames
Focal Length
1994/10
NSW 4224
161 & 162
152.76
2004/10
NSW 4877
21&22
152.76
Table 2. Information of aerial photo
Figure 7. Aerial photo of test site
161 161 161 151 151 161 161 161
151 161 161 151 151 161 161 151
Figure 8. Deformation result using PSInSAR