This repository contains an end-to-end workflow for extracting, transforming, and loading (ETL) Earth observation data to support wildfire monitoring and environmental analysis in Portugal.

The project integrates multiple geospatial datasets from Google Earth Engine (GEE) and Meteostat to generate per-tile environmental indicators (NDVI, soil moisture, brightness temperature, and local weather conditions).

Note:

• All heavy computations are performed on servers such Google Earth Engine or Meteostat.
• CSV exports are ignored in version control (.gitignore).
• You can reproduce the environment using.

conda env create -f environment.yml -n fire-etl
conda activate fire-etl

├── data/
│   ├── csv/          # Processed exports from Earth Engine & Meteostat
│   ├── geojson/      # Country grid and geometry inputs
│   └── ...
├── notebooks/
│   ├── extract_transform_load_data_1.ipynb
│   ├── extract_transform_load_data_2.ipynb
│   └── ...
├── environment.yml   # Conda environment definition
├── .gitignore        # Ignore large data & temp files
└── README.md

Execute the notebooks in order:

• extract_transform_load_data_1.ipynb: Create the grid and export NDVI, SMAP, FIRMS data per tile.
• extract_transform_load_data_2.ipynb: Retrieve weather conditions from Meteostat for each tile.