Collection of geospatial data science projects

Change detection in the amazon forest

This project utilizes Sentinel-2 and Sentinel-1 satellite imagery to detect deforestation in the Amazon rainforest by analyzing changes in vegetation between 2017 and 2023.

Description

This project includes four main Python notebooks:

1. sentinel2_download.ipynb: Downloads Sentinel-2 satellite images of a specified region within the Amazon rainforest and saves them to a tif file using the STACK API.

2. sentinel2_change_detection.ipynb: Analyzes the changes in NDVI over time to detect areas where deforestation might have occurred. It outputs visualizations of these changes.

3. sentinel1_download.ipynb: Downloads Sentinel-1 satellite images of a specified region within the Amazon rainforest and saves them to a tif file using the Sentinel Hub API.

4. sentinel1_change_detection.ipynb: Computes change detection using a stack of radar images. It outputs visualization of the change map.

Installation

To set up the environment for this project, you need to install the required Python libraries. You can install them using the following command:

pip install -r requirements.txt

Outputs

The Sentinel-2 satellite images over time:

The change maps over time:

The Sentinel-1 satellite images over time:

The change map:

License

This project is licensed under the MIT License - see the LICENSE file for details.

Acknowledgements

This project uses data from the Sentinel-1 & Sentinel-2 satellite mission provided by the European Space Agency (ESA).

*The code for Sentinel-1 change detection is from the following paper:

• Colored visualization of multitemporal data for change detection: issues and methods. Elise Colin Koeniguer et al. EUSAR 2018