Sentinel-1 ice surface velocities of Svalbard

Cite as:

Friedl, Peter; Seehaus, Thorsten; Braun, Matthias (2021): Sentinel-1 ice surface velocities of Svalbard. V. 1.0. GFZ Data Services. https://doi.org/10.5880/fidgeo.2021.016

Status

I N R E V I E W : Friedl, Peter; Seehaus, Thorsten; Braun, Matthias (2021): Sentinel-1 ice surface velocities of Svalbard. V. 1.0. GFZ Data Services. https://doi.org/10.5880/fidgeo.2021.016

Abstract

The data set comprises Sentinel-1 scene pair-velocity fields, as well as monthly and annually averaged velocity mosaics over Svalbard for the period January 2015 - November 2020. The data are provided as GeoTIFF rasters in UTM (scene-pair velocity fields) and polar stereographic north (mosaics) coordinate reference systems at a spatial resolution of 200 m and were derived by applying a well-established intensity offset tracking algorithm (Strozzi et al., 2002; Wegmüller et al., 2016; Friedl et al., 2018; Wendleder et al., 2018; Seehaus et al., 2018). For tracking, we used consecutive pairs of single or dual polarized Sentinel-1 SLC (Single Look Complex) TOPS (Terrain Observation with Progressive Scans in azimuth) SAR (Synthetic Aperture Radar) images recorded in IW (Interferometric Wide swath) mode at a pixel spacing of ~14 m in azimuth (az) and ~3 m in range (r), and a spatial coverage of ~250 x 250 km. For the time from 2015 to 2016, Sentinel-1 imagery is available at a minimum repeat cycle of 12 days and from 2016 onward at a minimum repeat cycle of 6 days.

The Sentinel-1 data were obtained from the ASF (Alaska Satellite Facility) DAAC (Distributed Active Archive Center), https://search.asf.alaska.edu. In case of dual polarized acquisitions (HH+HV or VV+VH), we only used the HH or VV channels for the processing.

Methods

Scene pair-velocity fields were generated by applying intensity offset tracking (feature tracking and speckle tracking) on two subsequent Sentinel-1 images (master and slave scene), using a window size of 250 x 50 pixels and a step size of 50 x 10 pixels. The results were (1) UTM-geocoded and orthorectified with the help of an external digital elevation model (3 arc second TanDEM-X Global DEM, Wessel et al., 2018), (2) filtered with an effective three-step filter approach (Lüttig et al., 2017) that removes > 99% of erroneous measurements and (3) corrected for remaining coregistration errors based on the median of the filtered range- and the azimuth-velocities measured over ice-free ground. The effective time stamp of each velocity field is derived as the mean date of the acquisition dates of the master and the slave scene.

Annual and monthly mosaics were derived from all filtered and corrected scene pair-velocity products that have a time stamp between 1 January–31 December of a year and between the first and the last day of a month, respectively. Before mosaicking, all scene-pair velocity fields were reprojected to polar stereographic north. We calculated the weighted means of the x- and y-velocity components, the velocity magnitude, the acquisition date, the time separation between the images, the displacement angle relative to true north, as well as the number of measurements per pixel. Additionally, we calculated the weighted standard deviation and the weighted standard error for the x and y velocity components, as well as the velocity magnitude. For averaging we took the SNR (Signal to Noise Ratio) of each velocity measurement pixel as weight.

Scene pair-velocity products and mosaics are provided as GeoTIFF rasters. The coverage files, containing the footprints of the Sentinel-1 scenes, are provided as shapefiles (.shp). The corresponding metadatafiles are text files (.txt). Quicklooks of the scene pair-velocities and the mosaics magnitudes (m d-1) are in .png format.

Authors

Friedl, Peter;University of Erlangen-Nuremberg, Erlangen, Germany
Seehaus, Thorsten;University of Erlangen-Nuremberg, Erlangen, Germany
Braun, Matthias;University of Erlangen-Nuremberg, Erlangen, Germany

Contact

Friedl, Peter (Researcher) ; University of Erlangen-Nuremberg;
Seehaus, Thorsten (Researcher) ; University of Erlangen-Nuremberg;

Keywords

ice velocity, remote sensing, feature tracking, speckle tracking, Sentinel-1, SAR

GCMD Science Keywords

EARTH SCIENCE > CRYOSPHERE > GLACIERS/ICE SHEETS > GLACIER MOTION/ICE SHEET MOTION
EARTH SCIENCE > CRYOSPHERE > GLACIERS/ICE SHEETS > GLACIERS
Earth Observation Satellites > Sentinel GMES > SENTINEL-1A
Earth Observation Satellites > Sentinel GMES > SENTINEL-1B
Earth Remote Sensing Instruments > Active Remote Sensing > Imaging Radars > SENTINEL-1 C-SAR