Uncertainty Quantification and Field Source Inversion for the Continental-Scale Time-Varying Gravity Dataset: A Case Study in SE Tibet, China

Southeastern (SE) Tibet on the Chinese mainland is geologically active and plays an important role in subsurface deformation and mass transfers. Hybrid gravimetry using both absolute and relative gravimeters is an efficient tool for monitoring surface and underground mass transfers. But for hybrid gravity network surveys, uncertainties are influenced by measurement errors, while sparseness of the network and environmental artifacts must be identified and minimized prior to studying gravity change. In this study, Bayesian gravity adjustment (BGA) was applied for the first time to the hybrid gravity network in SE Tibet during 2014–2016, which effectively reduced measurement uncertainties via its estimated scale factors and drift rates, thereby demonstrating its suitability for the large and complex gravity network in SE Tibet. To estimate the field source resolution, reduce environmental artifacts, and invert mass redistributions on the deep crust, an equivalent source inversion (ESI) based on the spatiotemporal smoothness regularization constraint model and the Akaike Bayesian information criterion parameter estimation method was applied to datasets processed by BGA. With respect to processing synthetic gravity data with spatiotemporal noises, the ESI was an effective algorithm, with the optimal field source resolution in SE Tibet being 0.75° × 0.75°. The apparent density change at a depth of 20 km was then inverted, with an average rate of −0.6 to 0.6 kg/m 3 /year, which was approximately 0.22‰ of the average crustal density. In addition, its spatial distribution showed close consistency with active tectonic block boundaries and low-velocity/high-conductivity zones. Comprehensively considering the hydrological effects, GPS observation studies, and geophysical and petrological evidence in the region, this study suggests that the crustal mass redistributions in SE Tibet are possibly controlled by active tectonic block boundaries and fluids distributed in the deep crust.