A New Global Database of Secular Horizontal GNSS Velocities

Steady-state secular motions of the Earth’s surface (i.e., motions not influenced by transient processes such as earthquakes or volcanic eruptions) reflect plate boundary interseismic strain accumulation, plate motions, post-glacial rebound, sea-level rise, or dynamic topography. Over the past three decades, the expansion of Global Navigation Satellite System (GNSS) networks has densified the number and spatial coverage of station position and velocity observations with improved measurement accuracy. In this study, we focus on horizontal motions and aim to create the most up-to-date, spatially dense velocity field. We gather secular velocities at ~35000 unique GNSS stations distributed globally, covering both tectonically active and stable regions. Roughly 18000 velocities are determined at the Nevada Geodetic Laboratory (NGL) from time-series of mostly continuous GNSS observations. However, the spatial coverage of the NGL velocity solution suffers from the absence of available RINEX (Receiver Independent Exchange Format) data in places such as most of the India-Eurasia collision zone. We therefore compile about 17000 additional continuous and campaign GNSS velocities from ~400 published studies and transform these auxiliary velocities onto the NGL velocity solution through a least-squares inversion. For several large earthquakes with sufficient GNSS observations (e.g., 2004 M9.1 Sumatra, 2011 M9.1 Tohoku, 2010 M8.8 Maule, and others), we correct GNSS time-series and auxiliary velocities for postseismic viscoelastic deformation using forward modeling based on a gravitational spherical Earth with a 1D rheological structure. For other earthquakes, we correct the GNSS time-series by removing the postseismic time-series fitted by an empirical logarithmic function. Additionally, we develop and apply a velocity outlier detection and removal algorithm to generate our final global velocity database. Our velocity field is an update to the compilation from the 2014 Global Strain Rate Model (GSRM) and greatly extends the scope of existing global velocity solutions. Our new database will be used to produce the next GSRM and to provide a starting velocity field for future integration with InSAR analysis.