GRASSINEL

University Project (FSU Jena) by Patrick Fischer, Marlin M. Mueller and Jonas Ziemer

This tool combines pyroSAR Sentinel-1 preprocessing capabilites with GRASS GIS (Version: 7.9dev) functionality for Sentinel satellite imagery.

It allows for automatic downloading of user-defined Sentinel-1 scenes, pre-processing, subsetting and multitemporal operations. Data is handled in GRASS internal formats and all available GRASS operations can be applied

Functionality Overview:

• Automatic Setup of GRASS environment
• Automatic download of Sentinel-1 data using the GRASS addon i.sentinel.download
• Pre-processing of downloaded Sentinel-1 using pyroSAR functionality (ESA SNAP is required for the processing, see here)
• Automatic subsetting of satellite data to specified region of interest
• GRASS functionality:
  • Creation of space time raster datasets (STRDS) in GRASS
  • Different multitemporal operations in GRASS
• Future functionality:
  • pyWPS implementation

Developed in Python 3.6, currently only supported on linux-based OS

Installation

In case you have git installed you can install the package as follows:

pip install git+https://github.com/marlinmm/GRASSINEL.git

Workflow of the GRASSINEL module:

GRASSINEL workflow

Examples

Example of created timeseries processed using pyroSAR and created using GRASS GIS time series functionality Sentinel-1 time series from 2020-06-06 to 2020-07-29 using only relative orbit number 168 in desending direction with 20m resolution with terrain flattening and thermal noise removal active (roughly 5 mins on 14 cores per scene)

Sentinel-1 time series from 2020-06-01 to 2020-06-07 using all available datasets for the chosen time frame in ascending direction with 50m resolution without terrain flattening and thermal noise removal (roughly 2 mins on 8 cores per scene)

Point plot of Sen-1 acquisition dates for VV and VH polariazation printed by the "visualize_stc" function

Screenshot of the GRASS addon for pyroSAR processing inside GRASS GISS