/
Displacement Time Series using S1-GUNW-MERGED with GIAnT

Displacement Time Series using S1-GUNW-MERGED with GIAnT

Owned by Alexander Torres
Last updated: Jul 08, 2020 by Hook Hua
3 min read

Github Repo and branch

Related files:

  • giant_time_series/scripts/create_filtered_gunw_merged_stack.py

  • giant_time_series/scripts/create_filtered_gunw_merged_stack.sh

  • giant_time_series/scripts/create_displacement_time_series.py

  • giant_time_series/scripts/create_displacement_time_series.sh

Filtered GUNW Merged Stack

Description:

The filtered GUNW-MERGED stack dataset (filtered-gunw-merged-stack) is primarily the HDF5 RAW-STACK.h5 and PROC-STACK.h5 outputs of GIAnT's PrepIgramStack.py and ProcessStack.py, respectively. Prior to running these PGEs, a first-order filtering step is performed to filter out GUNW-MERGED whose track don't match those specified. Additionally, GUNW-MERGED are filtered whose reference bounding box contains no data that pass the coherence threshold or which do not cover the region of interest.

Usage:

  1. In tosca interface, draw bounding box on the region of interest.

  2. Facet on the S1-GUNW-MERGED dataset.

  3. Facet on the dataset version (currently: v2.0.2).

  4. Facet on the track number.

  5. Click on On-Demand.

  6. For Action, select `GIAnT - Create filtered GUNW-MERGED stack [<version>].

  7. In the parameters section below, ensure track matches the track you initially faceted on. This ensures that S1-GUNW-MERGED products for other tracks are filtered out in case the user failed to facet down to them.

  8. Populate ref_point.

  9. Adjust other parameters accordingly.

  10. Click Process Now.

Example of Facet on Tosca

Outputs:

  • RAW-STACK.h5.gz - gzip-compressed HDF5 file of the filtered stack of GUNW-MERGED

  • PROC-STACK.h5.gz - gzip-compressed HDF5 file of the filtered stack of GUNW-MERGED with atmospheric and orbit corrections applied

  • browse.png - visual browse of temporal connectivity

  • gaps.txt - record of any temporal gaps detected in the stack

  • create_filtered_gunw_merged_stack.log - verbose log which can be used to determine what GUNW-MERGED were filtered and why

  • filt_info.pkl - pickle file containing IFG and filter information

  • data.xml, sbas.xml - other inputs needed by downstream displacement time series PGEs

 

Displacement Time Series

Description:

The displacement time series dataset (displacement-time-series) is primarily the
HDF5 LS-PARAMS.h5 (for SBAS-inversion) and NSBAS-PARAMS.h5 (for NSBAS-inversion)
outputs of GIAnT's SBASInvertWrapper.py and NSBASInvertWrapper.py, respectively.

Usage:

  1. In tosca interface, draw bounding box on the region of interest.

  2. Facet on the filtered-gunw-merged-stack dataset.

  3. Click on On-Demand.

  4. For Action, select `GIAnT - Create Displacement Time Series [<version>].

  5. In the parameters section below, select the inversion method: sbas or nsbas.

  6. Click Process Now.

Example of Facet on Tosca

Outputs:

  • LS-PARAMS.h5.gz or NSBAS-PARAMS.h5.gz- gzip-compressed HDF5 file of the displacement time series produced via the SBAS or NSBAS inversion method

  • browse.png - visual browse of initial time series step

Visualization

  1. Open HDF5 in Panoply: File->Open.

  2. Double click on rawts (raw time series) or recons (reconstructed time series) variable.

  3. Select Create a georeferenced Longitude-Latitude plot. Click on Create.

  4. Zoom in to the region of interest. On MacOSX, hold down the command key while you click and drag a bounding box over the region of interest.

  5. Click on the Scale tab and set Scale Range Min. to -100 and Max. to 100. You can play around with the values.

  6. Click on the Arrays tab and cycle through the time slices by clicking the up arrow button.