5. Conclusion
The main conclusions are:
- Current strain rate models based only on point GNSS measurements are
reliable at longer wavelengths (> 30-40 km) but have
large differences at shorter wavelengths so they cannot achieve the
100 nanostrain/yr accuracy threshold for assessing seismic hazard. In
addition, there are significant transient deformation processes (e.g.,
hydrologic, magmatic, and tectonic) having strain rate signals that
exceed the 100 nanostrain/yr threshold needed for hazard assessment.
- We developed an approach to integrate Sentinel-1 InSAR and GNSS
time-series (4.5 years) over the entire San Andreas fault system from
the ascending and descending look directions. This analysis was
enabled by the frequent, high-quality, observations from the
Sentinel-1 satellites.
- A video tour of the high-resolution LOS velocity maps reveals a wide
array of deformation processes including: active faults and stepovers;
extraction and recharge of groundwater, petroleum and geothermal
fluids; and continuous expansion of the surface of dry lake beds
(Movie S1, and available at
https://www.youtube.com/watch?v=SxNLQKmHWpY ).
- The two components of average LOS velocity are used to refine and
update estimates of creep rate along the major strands of the San
Andreas fault system.
- The two LOS components are decomposed into 3-components of velocity by
assuming the direction of deformation matches those predicted by a
GNSS-only velocity model.
- The higher spatial resolution vector velocity maps are used to
estimate the three components of horizontal crustal strain rate. The
results show significant off-fault strain, yet challenges remain in
separating the contribution from tectonic and hydrologic sources and
whether hydrologic strain will increase seismic hazards.
- Given the 20-year plus observation plan of the twin Sentinel-1
satellites, as well as continued GNSS operations, these high spatial
resolution, time-dependent products will continue to improve.