The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part B7. Beijing 2008 
109 
information from the unfilled cycle conditions is used. If the 
ambiguity search improved the solution, it is used as the 
starting point for a new iteration. 
However, due to a number of measurements excluded during 
the check process, it may happen that a scene is totally 
excluded, i.e. all interferograms containing a particular scene 
are excluded, 
making it impossible to compute the deformation for the 
particular scene for the point. This must be counted for when 
creating the adjustment matrices. 
Unfortunately, in-situ measured data are not yet available for 
comparison. Currently, we only try to reach the spatial 
continuousness of the results, which would mean that 
unwrapping errors, which are large in some areas, are corrected 
by the iterative adjustment. The spatial continuousness is partial, 
with some points having a deformation jump. 
Two statistical tests are implemented: Kolmogorov-Smimov 
test for checking if the data set has normal distribution (it tests 
the standard deviation as well as the normality of the residues) 
and Snedecorov-Fischer test for testing the standard deviation. 
5. RESULTS 
So far we do not have a proper way to display the results. We 
intend to use a GlS-software to easily visualise both spatial 
deformation between two time points as well as deformation of 
a single point in a time series. 
Figure 2. Deformations in Ervenice corridor computed for time 
interval between 1997-03-03 and 1997-09-29 
In figure 2 deformations in the area of Ervenice corridor are 
displayed and figure 3 contains visualisation of deformations in 
Kost’any area. 
The first figure of the triple displays the deformation of all 
processed points, the second one contains points validated using 
the Kolmogorov-Smimov test and the last one contains points 
validated using the Fischer test. The scale is imaged in figure 4, 
with -150 mm on the left and 150 mm on the right. 
Figure 3. Deformations in Kost’any area computed for time 
interval between 1997-07-21 and 1998-04-27 
Figure 4. Colour scale, -150 mm on the left border, +150 mm 
on the right border 
• As can be seen in both figures, the results heavily 
depend on the coherence of processed area. 
Regarding this parameter, Kost’any area shows better 
results (it is spatially continuous and probably stable). 
In case of ErvSnice corridor, a problem occurred that 
the reference point was not stable, so that the results 
cannot be interpreted straight away. 
• However, the idea of referencing the phase with 
regard to one point is correct. What is not correct is 
the phase of the reference point, and although its 
precision is improved using the average of the 
neighbouring points, there still remains a large error 
(in addition, the average of the wrapped phase is 
ambiguous). 
6. CONCLUSIONS AND FUTURE WORK 
Using InSAR method for land subsidence and monitoring of 
other types of deformation seems to be promising to give 
reasonable results in the future. However, it is very much 
influenced by the coherence of processed area. That is closely 
connected with the problem of spatial continuousness, which 
affects the algorithm based on neighbouring points. Even 
finding a stable reference point is very difficult, for in many 
cases no stable points are located near the processed area or are 
separated from the examined area by an area of low coherence.