Multi-station PPP network solutions for robust velocity field inversion of crustal deformation

To enhance the accuracy and robustness of GNSS crustal deformation monitoring, this study develops a high-precision regional solution that overcomes the ambiguity-fixing difficulty and rigid constraints of conventional undifferenced PPP. The method forms a multi-station PPP network that reconstructs fixable double-difference integer ambiguities, applies adaptive prior constraints to each station, determines the optimal stochastic model via power-spectral analysis, and inverts the regional velocity field with maximum-likelihood estimation. Validation using 2021—2023 observations from 10 sites in north-western Australia, single-day positioning accuracies better than 1 cm horizontally and 2 cm vertically. Introducing a white+flicker noise model into the coordinate time series provides realistic velocity uncertainties. The derived field shows a uniform northeastward motion of approximately 70 mm/a and vertically negligible yet strongly seasonal displacements. By jointly exploiting the precision of undifferenced observations and the reliability of fixed double-differenced ambiguities, the proposed approach enables robust large-scale velocity field inversion and offers a dependable tool for plate kinematics and crustal deformation research.