Forest Loss → MIDI (Saxony)

This repo contains a small pipeline of Jupyter notebooks that:

1. Download and preprocess Global Forest Change data for Saxony (Germany),
2. Extract and classify spatial clusters of forest loss, and
3. Turn those clusters into MIDI tracks – a sonification of deforestation patterns over time.

The workflow focuses on the federal state of Saxony (Sachsen) and uses the Hansen Global Forest Change v1.12 (2000–2024) dataset. :contentReference[oaicite:0]{index=0}

---

Data Source & Attribution

Forest change data:

Source: Hansen/UMD/Google/USGS/NASA :contentReference[oaicite:1]{index=1}

The underlying dataset is:

Hansen, M. C., P. V. Potapov, R. Moore, M. Hancher, S. A. Turubanova, A. Tyukavina, D. Thau, S. V. Stehman, S. J. Goetz, T. R. Loveland, A. Kommareddy, A. Egorov, L. Chini, C. O. Justice, and J. R. G. Townshend. 2013. High-Resolution Global Maps of 21st-Century Forest Cover Change. Science 342 (15 November): 850–853. https://doi.org/10.1126/science.1244693. Data available via the GLAD Global Forest Change app. :contentReference[oaicite:2]{index=2}

Tiles are downloaded from the official Hansen/UMD Google Cloud bucket. :contentReference[oaicite:3]{index=3}

Administrative boundaries for Germany/Saxony are pulled from the GADM 4.1 shapefile (level 1 = states).

---

Repository Structure

Key pieces:

• DownloadAndPrepareData.ipynb
  Clip Hansen rasters to Saxony, compute yearly statistics, and make overview plots.

• BuildAndClassifiyClusters.ipynb
  Label connected patches (clusters) of forest loss, compute shape metrics, classify them as points / lines / planes, and export per-year cluster datasets.

• MIDIcreation.ipynb
  Map cluster properties to musical parameters (pitch, velocity, duration, onset) and export yearly MIDI files.

Typical folders/files created by the notebooks:

• data/ – Hansen GeoTIFF tiles and other raw data
• gadm_germany/ – GADM shapefiles for German states
• hansen_lossyear_saxony.tif, hansen_treecover2000_saxony.tif – clipped rasters
• saxony_forest_loss_extended.csv – yearly forest loss metrics for Saxony
• loss_clusters_all_years.gpkg / loss_clusters_all_years.csv – cluster-level features across all years
• loss_clusters_summary_by_year.csv – per-year summary stats
• *.png – plots (time series, maps, cluster charts)
• midi_output_clock/ – exported MIDI files and parameter .txt files

---

Dependencies

All notebooks are standard Python/Jupyter and use:

• Core: python, jupyter, numpy, pandas
• Geo stack: geopandas, shapely, rasterio, rasterio.mask, rasterio.warp
• Image/metrics: matplotlib, scikit-image
• ML / stats: scikit-learn
• I/O / HTTP: requests, zipfile, io, os
• MIDI: pretty_midi

Install (minimal example, adjust to your environment):

pip install numpy pandas geopandas shapely rasterio scikit-image scikit-learn matplotlib pretty_midi requests