论文部分内容阅读
1992年6月28日加州兰德斯地震(M_w=7.3)震前有一个7小时的前震序列,它至少由28个地震组成。在这次研究中,我们根据附近台站记录的地震波形互相关的高精度定位,对几何条件和时间进程进行了仔细调查。与其试图获得各次地震的走时,不如使用排列波形,这样可以大大提高计时的精度,并且使初至波的走时选取非常准确。我们用改进的走时进行综合再定位,减少了相对定位误差,使其在水平方向小于100m,垂直方向小于200m。改进后的定位,使前震序列的几何条件变得很清晰。兰德斯前震发生在主震断层面上的直立阶跃处,长500m。由前震序列所确定的起始带在直立阶跃的南面向东南发展,在北面向北发展。这个几何条件被前震序列的震源机制所证实。它是右旋的,并且沿着断层两个相连部分上前震定位所确定的走向。当前震在阶跃处发生时,它顺时针旋转。前震序列的范围在水平和垂直方向上均约1km。由所有先前的前震引起的库仑应力变化的模型表明,前震很有可能不相互激发。这个结果对于Mw=4.4的紧临前震尤为清楚。因为序列中的应力传递似乎没有在它的发展中起到重要作用,我们推断出一个下伏的无震起始过程,很有可能为无震蠕动。其他的研究表明,地震的起始现象可以由断层带的不均匀性所控制。在兰德斯地震中,确实出现了这种情况,虽然不均匀的尺度很小,以至用常规定位技术都无法确定。
There was a 7-hour foreshock sequence ahead of the Lander earthquake in California (M_w = 7.3) on June 28, 1992, which consisted of at least 28 earthquakes. In this study, we conducted a careful investigation of the geometry and timing based on the high-precision positioning of the seismic waveforms recorded by nearby stations. Instead of trying to get the travel time of various earthquakes, it is better to use the arranged waveforms, which can greatly improve the accuracy of timing and make the selection of the travel time of the first arrival wave very accurate. We use the improved travel time comprehensive re-positioning, reducing the relative positioning error, so that it is less than 100m in the horizontal direction, the vertical direction of less than 200m. Improved positioning, the geometric conditions of the pre-earthquake sequence becomes very clear. The Landers foreshock occurs at an upright step on the fault plane of the main shock of 500 m in length. The onset zone, as determined by the foreshock sequence, develops southeast to the south of the upright step and north to the north. This geometry is confirmed by the focal mechanism of the pre-seismic sequence. It is dextral and follows the direction determined by the foreshock positioning on the two contiguous parts of the fault. When the current shock occurs at a step, it rotates clockwise. The range of the foreshock sequence is about 1km in both the horizontal and vertical directions. Models of Coulomb stress changes caused by all previous foreshocks show that foreshocks are most likely not excited by each other. This result is particularly clear for the immediate foreshock of Mw = 4.4. Since stress transfer in the sequence does not seem to play an important role in its development, we infer an underlying anaerobic initial process, most likely aseismic creep. Other studies show that the onset of an earthquake can be controlled by the inhomogeneity of the fault zone. This did happen in the Landers earthquake, though the scale of the inhomogeneities was so small that it was impossible to determine with conventional positioning techniques.