zibnix / gost_sar

Collection of tools developed by GOST team for extracting information from SAR data

SAR applications

This repository contains resources and examples to help you leverage SAR data (and optical data) for various mapping applications, in Google Earth Engine and Google COLAB. SAR systems, contrary to optical systems, do not depend on natural illumination and it is able to penetrate clouds, being able to provide weather-independent data. More information about SAR image formation and SAR satellite data and applications can be found here

The illustration below shows an example of SAR (gray) and optical (true color RGB) satellite images acquired over the Rome metropolitan area (Contains Copernicus Sentinel-1 and Sentinel-2 data [2021]).

Resources

1. Flood analysis and mapping

• guidance for choosing the appropriate methodology to map flood extent
• GEE scripts for flood mapping and damage assessment using S1 radar data and S2 optical data
• COLAB script for cloud masking and flood mapping using S2 optical data

2. Coherence change detection

• COLAB scripts using S1 data

3. Oil spill detection

• GEE script

4. Volcano eruption

• GEE script

5. Damage assessment using SAR Inteferometric Coherence This use case has been tested for Mariupol Damage assessment using 2 approaches:

• Coherence Change Detection: In this case it has bee assessed the damage to infrastructure of Mariupol city during the war until 5th April 2022. We have analyzed possible changes and damaged buildings using Coherence Change Detection exploiting the Sentinel-1 amplitude and phase information leveraging some methods developed above. Additionally, using open layers, they were able to quantify affected buildings, roads and bridges. In this damage assessment, they are also providing information on the industrial and commercial damaged buildings. The results had been validated with damaged locations reported by the media (i.e. BBC news) and by crowdsourcing platforms.
• Anomaly detection of coherence values: A continuation of the previous work using Anomaly detection techniques to detect changes (or damage) using large time series of SAR interferometric coherence values.

6. Nation-wide Radar Vegetation Index (RVI). In order to overcome the limitations of satellite optical sensors we have generated nationwide RVI maps over Ukraine that are further employed for analyzing the start/stop of the crop season from 2020 until 2022.

Get started

• Google Earth Engine

• Google COLAB

Let us know if you have tried any of the solutions presented, we'd love to hear about your use cases!

Disclaimer

The code available in this repository may produce results containing geographic information with limitations due to the scale, resolution, date and interpretation of the original source materials. No liability concerning the content or the use thereof is assumed by the producer.

The Global Operational Support Team, DECAT, The World Bank.