README.md

Technical Support Engineer Challenge: Monitoring Amazonas Deforestation

Green-Plant Management GmbH is a vegetation management specialist currently monitoring deforestation in the Amazonas region for a government project. They require a repeatable, automated pipeline to analyze Sentinel-2 time-series data using the UP42 API.

This Jupyter Notebook provides a step-by-step Python workflow to perform NDVI time-series analysis for a small AOI in the Amazonas, leveraging the UP42 ecosystem to discover, order, and download suitable Sentinel-2 imagery.

In this example, a comparison of two Sentinel-2 images with a one-year difference was chosen to monitor deforestation: for August 2024 and August 2025. August was chosen because it is one of the driest months in Brazil and it is used by the Program to Calculate Deforestation in the Amazon (PRODES) as a reference month (source).

(Created by Viktor Kochkin, April 2026)

Contents

Search and order Sentinel-2 images

• 1. Install dependencies

• 2. Get an access token to call API endpoints

• 3. Call the "Get a geospatial collection" endpoint

• 4. Load Area of Interest Polygon

• 5. Search the catalog to get a list of available Sentinel-2 scenes

• 6. Preview found scenes (thumbnails)

• 7. Order selected Sentinel-2 scenes

• 8. Get information about catalog orders statuses

• 9. Data management with SpatioTemporal Asset Catalog (STAC)

NDVI Analysis

• 1. Download assets for NDVI calculation

• 2. Calculate NDVI

• 3. Plot NDVI images

• 4. Plot NDVI Difference Map