142 
Figure 2-a. Differential interferogram 
after modified Goldstein filtering(No.l) 
Figure 2-b. Differential interferogram 
after modified Goldstein filtering(No.5) 
123700 123800 123900 124000 124100 124200 124300 
Figure 3. The average annual subsidence rate map from 1993 to 2000-2000 
41900 
41800 
41700 
41600 
Figure 4. Main mines distribution map of Fushun city (Zhao, 2003) 
3.3 Result 
Assuming that the deformation is linear, we then compute the 
average deformation rate within 7 years from the former 5 
vertical change maps. The annual elevation change of Funshun 
city is overlaid on the shaded SAR image acquired on July 14, 
1998, seen in Figure 3. The vertical and horizontal coordinate 
represent latitude and longitude amplified 1000 times, 
respectively. 
Figure 4 shows several main mines of Fushun city. There are 
three main large mine fileds, such as WOSM (Western Open-pet 
Coal Mine), Shengli, Laohutai, and Longfeng mine fields and 
EOSM (Eastern Open-pet Coal Mine) etc. Compared Figure2 and 
Figure 4, we can see the two “holes” due to severe decoherence 
shows the boundary of WOSM and Shengli, Laohutai and 
Longfeng mine fields. In the “holes”, we can only compute the 
range where dramatic deformation occurred, and can not provide 
accurate amplitude of deformation. Except this, the amplitude and 
range of deformation are both displayed in Figure 4. From Figure 
4 we can find that the relative subsidence value of east part 
around the WOSM and Shengli mine field are generally more 
than that of west part around Laohutai and Longfeng mine field, 
which coincide with the open report( Xu ,2005). 50% area of 
eastern Fushun has the 6-8 cm/y velocity, and most of the 
western Fushun has the velocity of more than 10 cm/y. 
4 DISCUSSION 
The above case studies have supplied much ground surface 
displacement information, and verified the D-InSAR technique 
can detect cm level motion effectively and rapidly with low cost. 
Meanwhile, we can find that the interferogram suffers from 
spatial and temporal decorrelation, which leads to the occurrence 
of some resultless area in the change maps. Besides this, the 
terrain of Fushun urban is rigorous, and high-precision 
topographic information is lacked. All these disadvantages make 
the two pass D-InSAR little difficult. Provided that more SAR 
SLC images are obtained, the results will be improved a lot. 
In addition, the D-InSAR technique provide relative motion of 
land surface, in collaboration with precise leveling surveying, 
GPS ,and GIS and professional subsidence model, we can 
evaluate the accuracy of D-InSAR technique, which is what we 
will do in the future. 
REFERENCES 
Ding, X. L., Li Z. W., 2006.Selection of Filtering Parameter for 
the Goldstein Radar Interferogram Filter. IEEE Transactions on 
Geoscience and Remote Sensing(submitted) 
Zhao, G X., Li, L., Chang Y. G, 2003.Design scheme of 
geological hazard monitoring system in Fushun urban, Liaoning 
Province, The Chinese Journal of Geological Hazard and Control, 
14(3), pp. 100-105 
Ge, L., Rizos C., Han S.W., and Zebker, H., 2001. Mining 
subsidence monitoring using the combined InSAR and GPS 
approach. The 10th FIG international symposium on deformation 
measurements, California: Orange 
Ge, L. L., Chang, H. C., Rizos, C., and Omura, M., 2004, Mine