Depth Of Moho In North China From Ambient Noise Autocorrelations

Body waves as well as surface waves can be extracted from seismic ambient noise contains.Compared with surface waves,retrieval of body waves from noise is much more difficult.As body waves contain much information on the Earth’s structure,it is new frontiers of the ambient noise seismology to extract body wave signals and apply them to exploring the deep interior of the Earth.Body waves contained in ambient noise are rather weak.In order to enhance PmP waves which are reflected normally from the Moho interface,we deal with ambient noise in the way:phase autocorrelations of noise are first calculated;autocorrelograms of many periods for the same station are stacked with a two-step approach.In the two-step approach,all the autocorrelograms are divided into groups,and the ones within the same group are stacked linearly into a trace;the stacked traces among the groups are then stacked with phase weights.As for PmP waves on the finally resulting stacked autocorrelograms,we pick them with an internationally common method:based on receiver function results,we calculate time windows expected to contain the PmP waves,taking the uncertainty of 5%of the average crustal velocities;in such a time window,the maxima in the second derivatives of the envelopes of the finally stacked traces are selected as the PmP traveltimes;the PmP traveltimes are converted to depths using the average crustal velocities.In this study,we have processed ambient noise for three years recoded by 120 stations in North China and obtained Moho depths of 106 stations.The results show that in the Ordos area,Moho first becomes shallower from south to north and then turns to be deep in the Yinshan orogenic belt;along the Zhangbo seismotectonic belt,the Moho depth increases as a whole from southeast to northwest,with small fluctuations between the two ends.Compared with the receiver function results,the Moho depths derived from noise autocorrelation functions have an average deviation of 0.97 km,and the corresponding two-way travel time deviation is 0.31s.Waveform characteristics of the autocorrelation functions indicate that the crust-mantle transition zone in North China has different velocity gradient patterns.It is potential as an exploration method to extract P-wave reflections from ambient noise autocorrelation and apply them on determining the interfaces(such as Moho)in the Earth.