Common-mode error and multipath mitigation for subdaily crustal deformation monitoring with high-rate GPS observations

Abstract

High-rate precise point positioning with a global positioning system (GPS) is recognized as a powerful tool for large earthquake monitoring. However, its ability to monitor subdaily subtle deformation is still limited by unmodeled errors, which primarily contain the common-mode error (CME) and multipath. In this study, a combination of modified principal component analysis and the sidereal filtering method is employed to separate and extract the CME and multipath from high-rate GPS displacements. The effectiveness of the proposed method is validated by processing 1 Hz GPS data at 131 static stations over 30 days, during which the 4 and 6 July 2019 Ridgecrest earthquakes occurred. The results show no obvious early afterslip within the first 5 h following the 6 July Mw 7.1 earthquake. The CME mainly consists of the first principal component and can be attributed to satellite orbit and clock errors. The proportion of the CME is higher than that of the multipath in the high-rate displacements, namely 36.7% versus 14.4%, 23.8% versus 18.1%, and 36.1% versus 12.5% for the east, north, and up components, respectively. By applying this method, the high-rate displacements can achieve precisions of 0.47, 0.50, and 1.60 cm in the three components. The results of the rapid fault-slip inversions indicate that the PPP fixed solution augmented with spatiotemporal filtering can improve fault-slip inversion and moment magnitude estimation and may lead to new findings for geophysics.