Receiver Function Imaging And Joint Inversion Method: Theory And Application

Seismic imaging is one of the most important way to map the structure of the Earth. Detailed knowledge of the crustal structure constrained by seismic imaging with datasets from dense array is fundamental to our understanding of the formation mechanisms and evolution processes of the crust as well as dynamics of the Earth's deep interior. High-resolution images of the crust are also essential to characterize oil and gas reservoirs or mineral deposit fields, and to understand earthquake and other geological hazards. In this study, we used receiver function (RF) imaging to map the Moho depth variations in North China and also developed a stepwise joint inversion with RF and surface wave data to constrain detailed crustal structure. Through real-data study and new method development, we resolve detailed knowledge of the crustal structure, therefore provide important constraints at deep depths on our understanding of crustal and mantle dynamic processes as well as the lithosphere evolution in typical tectonic regions.First, we employed a wave equation-based migration method to RF data with both Ps conversions and surface-reflected multiples from -200 broadband stations,to construct a high-resolution Moho depth model for northeastern North China Craton (NCC). The Moho depths are distinctly different on the opposite sides of the North-South Gravity Lineament (NSGL), as deep as >40 km in the west but shallow to <32 km in the east, with a rapid Moho uplift of ?10 km in between. This broadly consists with previous passive source and active source studies in the region. The Moho depths further exhibit significant N-S heterogeneities, beneath both sides of the NSGL. While a deeper Moho (-42 km) appears in the interior of the central NCC,spatially coincident with the presence of high-velocity anomaly from the lower crust to uppermost mantle and the refractory compositions in lithospheric mantle, a relatively shallow depth of ?3 8 km exists in the northern margin of pre-existing zones in the central to western NCC, perhaps corresponded to the center depth of the thick crust-to-mantle transition zone inferred petrologically and seismically. The former probably relates to the relics of the Archean crust, and the latter may reflect intense crust-mantle interactions associated with the underplating and upwelling processes. This difference implies uneven modifications of the lithosphere to the west of the NSGL. To the east of the NSGL, the crust beneath the Yan Mountains(?32-40 km) in the north is relatively thicker than the dramatically thinned crust beneath the Bohai Bay Basin (BBB) (?26-32 km) in the south. Such an observation may reflect the imprints of earlier N-S compressional deformation along the northern margin of the NCC associated with the amalgamation of the NCC with the Siberian craton, and different responses of the boundary area from the craton interior to the subsequent lithospheric extension since the Late Mesozoic. The Moho within the BBB further shallows from NE (?32 km) to SW (?26 km), but corresponded to southward LAB drop and gravity decrease. This correlation probably suggests a more buoyant lithospheric mantle in the interior of the BBB. Through comparisons of the Moho depth with surface altitude, gravity anomaly, and lithospheric thickness,we attribute the dominant W-E concordant structural difference to the Paleo-Pacific plate subduction under eastern Asia since Mesozoic. However, the small-scale N-S structural changes may have resulted from the structural heterogeneity of the cratonic lithosphere inherited since the formation of the NCC in the Paleoproterozoic or spatially uneven effects on the cratonic lithosphere of subsequent thermos-tectonic events during the long-term evolution of the craton.We also propose a stepwise joint linearized inversion method using RF, surface wave dispersion and Rayleigh wave ZH ratio to better resolve 1D crustal shear wave velocity (Vs) as well as the Vp/Vs ratio structure. A three-stage inversion strategy,which can take advantages of each dataset due to their complementary sensitivities to the crust, is used to gradually resolve the velocity structure. We firstly invert surface wave dispersion and ZH ratio data to obtain a 1D smooth and absolute Vs model, and then incorporate RF data in the joint inversion to obtain a finer Vs model with better constraints on interface structures. Both compressional wave velocity (Vp)and density model are computed from Vs model by empirical relationships in the first two stages. At last, three datasets are recombined to further constrain the Vp structure based on the obtained Vs model, meanwhile, layered Vp/Vs ratio structure is also retrieved. Through synthetic tests and Monte Carlo error analyses, we demonstrate that the proposed joint inversion method can not only resolve crustal Vs model very well but also resolve Vp model for relatively thicker crustal layers. Thus,we have been able to resolve reliable depth-dependent Vp/Vs ratios for the crust,which is important for understanding lithology and physical state of the crust at depths in combination with other geophysical datasets.