| dc.description.abstract | In order to investigate stress adjustments in the source region of the 1999 Chi-Chi earthquake, the regional seismic activities, focal mechanisms, and crustal stresses are together analyzed in this study. In addition, we have identified the true fault planes by applying several methods of pattern recognition on aftershock distributions. The results are further used to examine their relevance to the known fault structures.
In the study of seismic activities, we first define the aftershocks of the Chi-Chi earthquake by the earthquake clustering method. The locations of aftershocks are then adjusted by the hypocenter collapsing method. The results show six distinct clusters of aftershocks. An analysis of the b value shows a lack of larger earthquakes before the Chi-Chi earthquake which was subsequently compensated by intensive its aftershocks. Using a moving window in time, we have observed that the b value gradually increases from 0.8 to 1.0 beginning 1994 until 1999. Whether this anomalously linear increase of the b value is a precursor of the Chi-Chi earthquake deserves further investigations.
In the study of focal mechanisms, we found the larger and deeper earthquakes often exhibit thrust faulting whereas smaller and shallower earthquakes exhibit strike-slip type. The fact that larger earthquakes are dominantly thrust faulting is a reflection of the regional crustal stress regimes in the Chi-Chi source. With respect to the Chelungpu fault, thrust faulting dominates the hanging wall areas to the east, strike-slip faulting near its southern and northern ends, and southeastern side, whereas normal faulting in its central part and to the eastern side of the Central Mountain Range. Furthermore, the relative ratios of the numbers of the three types of focal mechanism are almost constant throughout the three-year period after the Chi-Chi.
Regarding fault plane identification, a traditional way to effectively identify a fault plane is based on aftershock distributions. We have successfully applied several methods of pattern recognition, including earthquake clustering, hypocenter collapsing, principal component analysis, and equal area projection, to determine the true fault plane from the two nodal planes of a focal mechanism. The results of 115 identified fault planes are informative for subsequent studies on possible association of seismic activities with fault structures, including pre-existing faults, new fractures, or underground blind faults.
In the study of crustal stresses, we analyzed the stress patterns and changes of the source region due to the Chi-Chi earthquake by the method of moment tensor summation, where the moment tensor of an individual earthquake is constructed from its focal mechanism solution and a scalar moment derived from the moment-magnitude relationship. The results reveal that although the distributions of seismicity and focal-mechanism patterns are different before and after the 1999 Chi-Chi earthquake, the overall patterns of crustal stresses remain similar. A grid analysis of stress estimation has established a database of empirical focal mechanisms, moment tensors, and crustal stresses of the Chi-Chi source region.
In summary, the results of composite analyses of seismic activities, focal mechanisms, and crustal stresses can be summed up as follows. In a shear stress environment, the strike-slip deformation is usually concentrated around known faults, often induces earthquakes with small magnitudes. In a compressional stress environment, the thrust deformation often induces earthquakes which will distribute in broad areas and usually triggers complex types of focal mechanisms on pre-existing faults with significant magnitudes. | en_US |