| Summary: | 碩士 === 國立成功大學 === 地球科學系 === 89 === The destructive Chichi earthquake (ML=7.3) occurred on September 21, 1999, in Central Taiwan. It produced a rupture trace more than 100 km long following the Chelungpu Fault. Many geodetic measurements have been taken to this area for detecting the co-seismic deformation following by Chi-Chi earthquake. Both the seismological records and the GPS measurements indicated significant co-seismic slip varying from a few metres to nearly 10 metres. However, the disadvantages of those methodologies are not only labor-intense but also no cost-effective. Furthermore, the spatial coverage of the geodetic observations is too sparse to clarify overall regional deformation pattern on the footwall and hangingwall of the Chelungpu Fault. Therefore, development of a new technique that can provide intense spatial distribution and finance loading is the main target of this study. As Interferometric Synthetic Aperture Radar (InSAR) allowed successful reconstruction of earthquake deformation, volcano inflation and deflation, landslide and subsidence, thus the significant surface displacement produced by Chi-Chi earthquake offered the excellent opportunity to apply Interferometric SAR reconstruct the coseismic deformation patterns. Meanwhile we try to overcome the problems of noisy interferograms caused by full-forested and rural areas in Taiwan.
In order to eliminate the topographic contribution, we used 3-pass differential InSAR (D-InSAR) approach to generate digital elevation model (DEM), which is suitable using for monitoring surface deformation over the experiment area. Although coherence is not perfect in most of full-forested hangingwall areas of the Chelungpu Fault and rural areas but we find that fringes in footwall area such as Taishung city show the significant co-seismic displacements. We also find a lobed pattern that suggests up to 15 cm of subsidence within a small area and follows the same patterns as compared with synthetic fringes from dislocation model based on GPS measurement.
A crucial problem deals with the accuracy and error sources revealed by our InSAR measurements and by coherence coefficient at different land-cover. The difference between D-InSAR and GPS measurements is less than 8 cm in city area but up to 50cm on the forested hanging wall. To this aspect, it is inadequate to apply InSAR technique in high mountains and dense vegetation area of Taiwan. In this study we obtain the reliable fringes in Taishung city, however in other areas such as rural areas and forest areas, the results are quite poor. It suggests that InSAR is indeed a convenient and useful tool in investigating the elastic deformation in high-population with intense building urban area. It is an important concern because the dense populations live in the big city in Taiwan. From the seismic hazard of view, the monitor of deformation built up during the interseismic period is a major concern for seismic hazard assessments in Taiwan. To better investigate this essential concern, further works including more InSAR analyses and geodetic measurements should be carried out in the near future.
|
|---|
