Kinematic Reference Frame for the Crustal Deformation Cycle: Transients and Model Discrepancies

We describe a four-dimensional kinematic reference frame that relies on surface displacements, rather than station velocities to detect transient deformation (e.g., postseismic, fault creep, slow slip events, preseismic) atop time-independent secular motions, which is critical to our understanding of physical processes and may require a reassessment of seismic risk. Our approach also accounts for motions related to natural (e.g., hydrology, climate) and anthropogenic (e.g. groundwater effects). The daily GNSS displacements are estimated by the NASA MEaSUREs Extended Solid Earth Science ESDR System (ESESES) project that uses a unified metadata database and a combination of GNSS solutions generated by two independent analysis centers at JPL and SIO. The Earth Science Data Records (ESDRs) now span twenty-seven years. Where feasible, integrating Sentinel-1 radar interferometry imagery with a 12-day repeat cycle and 200 m resolution with daily displacements of a GNSS network with a typical spatial resolution of 10-40 km provides a four-dimensional kinematic reference frame with high spatio-temporal resolution and mm displacement and sub-mm/yr velocity, and <50 nanostrain/yr accuracies. We compare the observed displacements to the motions predicted by a secular fault slip model that may have been derived from a variety of geodetic, seismic and geologic sources. The slip model provides the information on high velocity gradients near faults but can also be provided by the InSAR observations, where available. Multiplying the predicted surface velocities by time since a reference epoch, we can difference them from the observed daily displacements to identify transients as well as secular deviations from the model. The residuals can then be interpolated into displacement and strain rate grids in plate boundary zones. Because of the irregular nature of vertical deformation, we interpolate GNSS vertical displacements without the use of an underlying model.