Exploring Moho sharpness in Northeastern North China Craton with frequency-dependence analysis of Ps receiver function

The North China Craton (NCC), one of the oldest cratons in the world, has attracted wide attention in Earth Science for decades because of the unusual Mesozoic destruction of its cratonic lithosphere. Understanding the deep processes and mechanism of this craton destruction demands detailed knowledge about the deep structure of this region. In this study, we calculate P-wave receiver functions (RFs) with two-year teleseismic records from the North China Seismic Array ( 200 stations) deployed in the northeastern NCC. We observe both diffused and concentered PpPs signals from the Moho in RF waveforms, which indicates heterogeneous Moho sharpness variations in the study region. Synthetic Ps phases generated from broad positive velocity gradients at the depth of the Moho (referred as Pms) show a clear frequency dependence nature, which in turn is required to constrain the sharpness of the velocity gradient. Practically, characterizing such a frequency dependence feature in real data is challenging, because of low signal-to-noise ratio, contaminations by multiples generated from shallow structure, distorted signal stacking especially in double-peak Pms signals, etc. We attempt to address these issues by, firstly, utilizing a high-resolution Moho depth model of this region to predict theoretical delay times of Pms that facilitate more accurate Pms identifications. The Moho depth model is derived by wave-equation based poststack depth migration on both Ps phase and surface-reflected multiples in RFs in our previous study (Zhang et al., submitted to JGR). Second, we select data from a major back azimuth range of 100° - 220° that includes 70% teleseismic events due to the uneven data coverage and to avoid azimuthal influence as well. Finally, we apply an adaptive cross-correlation stacking of Pms signals in RFs for each station within different frequency bands. High-quality Pms signals at different frequencies will be selected after careful visual inspection and adaptive cross-correlation stacking. At last, we will model the stacked Pms signals within different frequency bands to obtain the final sharpness of crust-mantle boundary, which may shed new lights on understanding the mechanism of cratonic reactivation and destruction in the NCC.