Adaptive weighting of InSAR and GNSS data for coseismic deformation and slip inversion of the M _W 6.4 Meinong earthquake, Taiwan

Empirical weighting is a major source of uncertainty in joint inversion of InSAR and GNSS data. An objective weighting strategy based on Helmert variance component estimation (H-VCE) is applied to the 2016 Mw 6.4 Meinong earthquake in Taiwan. The weighted data are used to reconstruct the three-dimensional (3-D) coseismic deformation field and constrain fault geometry. Slip distribution is inverted using Green’s functions computed for a CRUST1.0 layered elastic structure. The preferred model indicates a buried low-angle thrust with a strike of ∼317° and a dip of ∼33°. Slip is heterogeneous and concentrated at depths of 8–15 km, with a maximum of ∼0.56 m. The corresponding scalar seismic moment is M0 = 4.12 × 1018 N·m (Mw 6.41). The inferred slip pattern is broadly complementary to the aftershock distribution and is supported by independent LOS validation. H-VCE reduces residual misfit relative to empirical weighting and improves inversion stability. The resulting method is useful for geodetic source inversion of blind-thrust earthquakes in structurally complex fold-and-thrust belts.