Wavefield decomposition for passive ocean bottom seismological data
Wavefield decomposition for passive ocean bottom seismological data
复制标题
被动海底地震数据的波场分解
DOI：
10.1111/j.1365-246x.2005.02761.x
复制
发表时间：
2005
影响因子：
2.8
通讯作者：
T. Dahm
中科院分区：
文献类型：
--
作者：
M. Thorwart;T. Dahm
摘要
来自海底站的地震数据显示，水听器和地震仪通道上存在强烈的水层多次波，具体取决于水深和海底阻抗对比。这些多次波可以通过波场分解（WDC）分解为上行和下行波场来衰减。我们将弹性分解方案应用于第勒尼安海的地震记录，以展示不同被动地震学研究的潜力。我们开发了一种三步分解方法，为我们的地震数据提供了最佳结果。分解过程中可以限制水听器和地震仪之间的相对放大系数。在我们的例子中，发现了 0.25 到 0.4 之间的因子。我们发现，局部深部地震记录中的水层多次波显着减少，因此可以研究 P 相尾波中的构造相。附近两个台站的海底 P 波阻抗对比估计为 1.3 和 1.5 量级，表明最上层非常泥泞且松散。我们用合成数据证明，水层多次波会严重扭曲远震接收器功能。 WDC 显着提高了接收器函数中结构相位的信噪比。单站对实际数据的应用证实了理论预测。我们预测了走时残差的偏差效应，这是通过不同台站的远震记录的互相关来估计的，在主周期为 50 秒的波中，最多可达 8 秒。波形分解消除了这种明显的走时残留并实现无偏估计。理论预测通过两次远震地震的实际数据得到验证。根据我们的结果，我们强烈建议将弹性 WDC 应用于宽带台站的被动地震记录。
SUMMARY
Seismic data from ocean bottom stations show strong water-layer multiples on the hydrophone and seismometer channels depending on the water depth and the seafloor impedance contrast. These multiples can be attenuated by means of a wavefield decomposition (WDC) into up and downgoing wavefields. We apply an elastic decomposition scheme to earthquake recordings in the Tyrrhenian Sea to demonstrate the potentials for different passive seismological studies. A three-step decomposition approach has been developed giving optimal results for our earthquake data. Relative amplification factors between hydrophone and seismometer can be constrained during decomposition. In our case, factors in between 0.25 and 0.4 have been found. We show that water-layer multiples on recordings of local deep earthquakes are significantly reduced, so that structural phases in the coda of the P phase can be studied. The seafloor P-wave impedance contrast at two nearby stations is estimated to the orders of 1.3 and 1.5, indicating that the uppermost layer was very muddy and unconsolidated. We demonstrate with synthetic data that teleseismic receiver functions are strongly distorted by means of water-layer multiples. The WDC significantly improves the signal-to-noise ratio of structural phases in receiver functions. A single station application to real data confirms the theoretical predictions. We predict the biasing effect of traveltime residuals as estimated by cross correlation of teleseismic recordings from different stations in the order of up to 8 s at waves with a dominant period of 50 s. Waveform decomposition removes this apparent traveltime residual and enables unbiased estimates. The theoretical predictions are verified with real data from two teleseismic earthquakes. Based on our results we strongly recommend applying an elastic WDC to passive seismological recordings on broad-band stations.