Scribd
Upload
18 views5 pages

Command Line inSAR Processing

InSAR

Uploaded by

prabhu
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
18 views5 pages

Command Line inSAR Processing

InSAR

Uploaded by

prabhu
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 5

SENTINEL-1

Command-Line InSAR Processing


Issued February 2020

Alex McVittie

Copyright © 2020 SkyWatch Space Applications Inc. https://skywatch.com


http://step.esa.int
Introduction
This tutorial will guide you through the setup of creating an InSAR processing framework for Sentinel-
1 SLC radar imagery to create interferograms.

System requirements
Interferogram generation is resource-intensive. Having at minimum a quad-core CPU, 16gb of RAM
and 20gb of free disk space is recommended. You also need to have an EarthData (Free to register)
account to download the Sentinel-1 imagery. As the scenario in this tutorial is from 2016, it is in offline
mode on the Copernicus data hub, increasing wait time to access. NASA’s ASF Vertex has the data
readily available.
You will also need to have SNAP installed on a desktop with a monitor to set up the processing graph,
prior to running this on a server with command-line only.

Scenario
For this InSAR tutorial, we will be generating an interferogram that visualizes the surface deformation
caused by the April 16th 2016 Kumamoto earthquake in Japan. Two Sentinel-1 products are needed, one
for before the event, and one for after. They can be downloaded here:
https://datapool.asf.alaska.edu/SLC/SA/
S1A_IW_SLC__1SSV_20160408T091355_20160408T091430_010728_01001F_83EB.zip
https://datapool.asf.alaska.edu/SLC/SA/
S1A_IW_SLC__1SSV_20160420T091355_20160420T091423_010903_010569_F9CE.zip

Building the processing graph


We will be building upon an existing graph that is provided with SNAP – the TOPSAR Coreg
Interferogram graph, available in the graph builder through Graphs > InSAR Graphs.
While this generates an initial interferogram and debursts it, there are a few enhancement steps required
after the TOPSAR-Deburst. Topographic phase removal must be run, along with multi-looking and
Goldstein phase filtering. Topographic phase removal (Radar > Interferometric > Products >
TopoPhaseRemoval) improves the accuracy of the interferogram by reducing the influence created by
terrain using a DEM. SNAP will auto-select a DEM to download, but you may wish to modify the
parameters to choose a different DEM in the processor options.
At this point, the interferogram has a large amount of noise, with patterns in the image being difficult to
detect. Multi-looking (Radar > SAR Utilities > Multilook) will improve the phase fidelity of the the
interferogram, which performs averaging of neighbouring pixels.
Within the processing parameters tab, the number of range and azimuth looks defines the window size
that averages the product. Setting the Range Looks to 6 and the Azimuth Looks to 2 results in a 6x2
window to average pixels with, creating an approximately 25x25m resolution raster.
Finally, we must apply the Goldstein phase filter (Radar > Interferometric > Filtering >
GoldsteinPhaseFiltering) to improve the interferometric phase. Our graph should now look like the
following:

Save this graph XML file and open the XML file in your preferred text editor. To build a repeatable
workflow, we want to substitute a few values in the graph with variables that can be brought in from a
text file – this allows you to run this graph in batch on many sentinel-1 image pairs, if you needed to
produce many sets of interferograms. SNAP’s gpt can substitute ${variableName} with a variable
value provided in a configuration file.
Within the first Read, add the following directly after <parameters/> :
<parameters class="com.bc.ceres.binding.dom.XppDomElement">
<file>${product1}</file></parameters>

If this already exists due to having products open in SNAP and the reads being auto-populated, replace
the path in the <file> tags with the ${product1} variable.
Do the same thing with the second Read, but have ${product2} in the <file> tags instead.
Replace <subswath/> in both TOPSAR-Split operators with <subswath>${subswath}</subswath>,
along with replacing <selectedPolarisations/> with
<selectedPolarisations>${polarization}</selectedPolarisations>
Additionally, within the TOPSAR-Split operators, remove the firstBurstIndex and lastBurstIndex tags
and values.
Within Multilook, replace <nRgLooks>6</nRgLooks> with <nRgLooks>${rangeLook}</nRgLooks>
and <nAzLooks>2</nAzLooks> with <nAzLooks>${azLook}</nAzLooks>
Within Write, replace <file>/tmp/target.dim</file> with <file>${outputFile}</file>
With our graph XML file prepared, let’s create a parameter file that we can use to give SNAP the
information it needs to transform our raw products into a filtered interferogram.
Create a new text document titled input_parameters.properties . Within it, populate it as such, replacing
the paths for the two products with the path to where you have saved your products:
product1=/media/skywatch/USB/2016/
S1A_IW_SLC__1SSV_20160408T091355_20160408T091430_010728_01001F_83EB.zip

product2=/media/skywatch/USB/2016/
S1A_IW_SLC__1SSV_20160420T091355_20160420T091423_010903_010569_F9CE.zip

subswath=IW1

selectedPolarisations=VV

rangeLook=6

azLook=2

outputFile=/data/japan_earthquake.dim

To execute this, in a shell simply run


gpt processing_graph.xml -e -p input_parameters.conf

and it will write the filtered interferogram out to (in this case) /data/japan_earthquake.dim
The GPT processing framework has many additional options that you can set. Run gpt -h to see a full
list of options, such as setting available RAM and multithreading options.
Output interferogram – the deformation fringe pattern very clearly shows the areas affected by the
earthquake.

You might also like