Horizontal Velocities in a Global Reference Frame Derived from
Sentinel-1 Along-track Interferometry
Abstract
InSAR measurements of ground displacement are relative, due to unknown
integer ambiguities introduced during propagation of the signal through
the atmosphere. However, these ambiguities mostly cancel when using
spectral diversity to estimate along-track (azimuth) velocities allowing
measurements to be made with respect to a terrestrial reference frame.
Here, we calculate along-track velocities for a partial global dataset
of Sentinel-1 acquisitions as processed by the COMET LiCSAR system, and
find good agreement with model values from ITRF2014 plate motion model.
We include corrections for solid-Earth tides and gradients of
ionospheric total electron content based on a moderate resolution model
IRI2016. Application of tidal corrections improves the average velocity
precision from 23 to 11 mm/yr. Ionospheric corrections, however, have
significant effect only in near-equatorial regions. The median
difference between along-track velocities and values predicted by
ITRF2014 is 3 mm/yr. A preliminary study using reprocessed precise orbit
determination products in a limited dataset shows significant
improvement in both precision and accuracy. By combining data from
ascending and descending orbits we are able to estimate north-south
(N-S) and east-west (E-W) velocities with an average precision of 3 and
16 mm/yr, respectively. Although we have calculated these estimates over
large 250 x 250 km areas, such measurements can also be made at much
higher resolution, albeit with lower precision. These “absolute”
measurements can be particularly useful for global velocity and strain
rate estimation, where GNSS measurements are sparse. We will also
investigate large-scale averages of across-track (range) pixel offsets,
which are most sensitive to E-W and vertical displacements, and perform
a comparison to a GNSS network in selected areas.