Body Wave Retrieval From Noise Crosscorrelation Functions Between Dense Arrays And Its Application To ChinArray

The seismic wave signals propagating between two stations can be extracted using ambient noise cross-correlation for further studies such as tomography and monitoring.Theoretically,both body waves and surface waves can be obtained in the noise cross-correlation function.Since the ambient noise consists mainly of surface waves,the surface waves are easier to extract and are the signals utilized in most studies.The extraction of the body waves in the noise cross-correlation function is affected by the stationary region of the body waves,the source distribution of the body wave noise source and reflection attenuation,and its signal is weaker.Using a dense station array to enhance the signal-to-noise ratio of the body wave signal in the noise cross-correlation function by spatial stacking is an effective method to extract and analyze the body wave signal.In this paper,the vertical component of the Japanese F-net(2013-2018,73 stations)and the Chinese seismic scientific detection array Phase II(2013-2016,667 stations)and Phase III(2016-2018,361 stations)were used to calculate the noise cross-correlation function between the dense array and to extract and analyze the body waves with spatial stacking.The main work and the conclusions obtained are as follows.(1)The direct P-wave and core reflection Pc P and Sc S body wave signals at station spacing of 15-40 degrees were obtained by the bin stacking using the noise cross-correlation between the vertical components of F-net and Chin Array II stations.Among them,the body wave signals with periods of 3-10 s are dominated by the direct P waves,while the body wave oscillations with periods of 30-60 s are dominated by Pc P and Sc S.The analysis of the signal-to-noise ratio of the body wave with the bin stacking width shows that the optimal stack width of the direct P-wave to obtain the highest signal-to-noise ratio is about 20 km when the stack width varies from 5 to 100 km,while the signal-to-noise ratio of Pc P and Sc S increases with the increase of the stack width.The optimal widths are related to the period and slowness of the signals.The direct P waves show strong unidirectionality and are mainly transmitted from the F-net to Chin Array.their signal-to-noise ratios show some seasonal variations,indicating that their energy sources are mainly from the body wave noise source in the North Pacific Ocean,which is related to the ocean activity.The signal-to-noise ratios of long-period Pc P and Sc S show no significant seasonal variations and are positively correlated with the occurrence of significant global earthquakes of magnitude 6.5 or higher in the corresponding time periods.(2)The extraction and orientation dependence of the direct P-wave signal with period 3-10 s are investigated using the noise intercorrelation function between the vertical components of the F-net and Chin Array III and the bin stacking method.The results show that the direct P-wave signals also exhibit strong unidirectionality within the epicenter distance of 10-30 degrees,and their energy mainly comes from the North Pacific Ocean.Meanwhile,the optimal bin stacking width is about 20-25 km and is controlled by the slowness of the P-wave signal at different epicentral distances as the epicentral distance increases and increases.Meanwhile,the signal-to-noise ratio of P-wave in the stacking results of similar epicentral distances and different orientations is significantly different.The analysis shows that the signal-to-noise ratio of the corresponding P-wave is related to the distribution of the body-wave noise sources in the North Pacific region,and the signal-tonoise ratio of the extracted P-wave is higher when the path orientation is close to the bodywave noise source,which shows the difference between the shape of the stationary zone of the body wave and the stationary zone of the surface wave,and its influence on the signal extraction.(3)The reflected P-wave signals of 410 km and 610 km discontinuity in North China are extracted and analyzed by using the vertical component ambient noise cross-correlation function of Chin Array III and the spatial stacking method of co-center points stacking.Through sub-regional co-reflection point stacking and phase weighted stacking,clear reflected P-wave signals of the discontinuities can be obtained and the lateral structure of the inter-section between the study areas can be analyzed.The superimposed profiles after time-depth conversion shows that the 410 km discontinuities below the study area is low in the west and high in the east,and the area of weak amplitude at the 660 km discontinuities corresponds to the north-south gravity gradient zone.The correlation results have a good correspondence with the results of the receiver function.Using the noise correlation function to extract the body wave signals can expand the types of signal extraction in noise correlation and be subsequently used to study the structure of the deep Earth.The results of existing studies show that the weaker body wave signals can be enhanced by using spatial stacking to obtain reliable body wave signals.At the same time,the extraction of the body wave signal has a strong correlation with the distribution and characteristics of the body wave noise sources.In the subsequent study,the extraction accuracy of the relevant body wave signals can be further improved by adopting a more refined spatial stacking method and considering the influence of the stationary zone of the body wave signals for the subsequent study.