We present a comprehensive 3D lithospheric-scale model of the South China Sea region (SCS), which reveals the structural configuration of the area. This model delineates seven distinct geological units: (1) seawater, (2) sedimentary cover, (3) continental crystalline crust, (4) oceanic crust, (5) upper lithospheric mantle, (6) lower lithospheric mantle, and (7) sub-lithospheric mantle. The model covers an area of 960 km × 1260 km and reach down to a depth of 250 km. It is provided as uniformly spaced grids with 10 km intervals for each unit. The geometries and density distributions within the crust have been compiled and interpolated from a variety of datasets, predominantly seismic data (see section 6). To eliminate boundary effects, the model boundaries have been extended by more than 500 km in all horizontal directions, incorporating additional constraining data from the extended region. Additionally, we provide gridded gravity field data, a density voxel cube for the sub-lithospheric mantle, and relevant tomography data. Notably, the density of the lower lithospheric mantle was derived from 3D gravity inversion modeling.

Topography and bathymetry data were obtained from the ETOPO_2022 dataset (NOAA National Centers for Environmental Information, 2022). Then, we integrated reflection and refraction seismic profiles (Table 1 and 2) to constrain the sediment base and the Moho interface, and where seismic profiles were lacking, we used a global crustal model-ECM1 (Mooney et al., 2023) to fill gaps.

To derive sediment thickness from Multi-Channel Seismic (MCS) reflection data (as listed in Table 1, section 6), the two-way travel time (TWT) was converted to depth below the seafloor using specific time-depth conversion formulas (Table 3). For the Moho interface, the TWT within the crystalline crust layer is converted to depth below sediment basement by using the time-depth relationship established by Huang et al. (2023) (Table 3), which is based on velocity-depth profiles from seismic refraction data in the SCS. Additionally, the Ocean Bottom Seismometer (OBS) refraction data (Table 2), presented in depth terms, were extracted directly through digitization. For the upper mantle, we converted the Vs tomography data of Tang & Zheng (2013) into mantle temperature using the method of Priestley & McKenzie (2006) and defined the depth of the 1300°C isotherm as the Lithosphere-Asthenosphere Boundary (LAB).

To enhance the gravity response of our 3D density models, we referenced the free-air gravity disturbance at an altitude of 6 km above sea level, as derived from the EIGEN-6C4 model (Förste et al., 2014; Ince et al., 2019), a global gravity model that combines satellite and terrestrial data sources. We selected a height of 6 km that is above the highest topographical point of the model in order to ensure that all gravity observations are outside the subsurface space of relevant mass variations.