BioMassters: A Benchmark Dataset for Forest Biomass Estimation using Multi-modal Satellite Time-series https://nascetti-a.github.io/BioMasster/

The objective of this repository is to provide a deep learning ready dataset to predict yearly Above Ground Biomass (AGB) for Finnish forests using multi-temporal satellite imagery from the European Space Agency and European Commission's joint Sentinel-1 and Sentinel-2 satellite missions, designed to collect a rich array of Earth observation data

Reference data:

• pixel-wise above-ground biomass maps
• Measurements were collected using LiDAR calibrated with in-situ measurements.
• Total 13000 patches, each patch covering 2,560 X 2,560 meter area.

Feature data:

• Sentinel-1 SAR (band order: ASC VV, ASC VH, DSC VV, DSC VH)
• Sentinel-2 MSI (band order: B2, B3, B4, B5, B6, B7, B8, B8A, B11, B12, CLP)
• 12 months of data (1 image per month)
• Total 310,000 patches

Data Specifications:

Data Size:

  dataset         | # files  | size
  --------------------------------------
  train_features  |  189078  | 215.9GB
  test_features   |   63348  |  73.0GB
  train_agbm      |    8689  |   2.1GB

Citation:

@inproceedings{nascetti2023biomassters,
  title={BioMassters: A Benchmark Dataset for Forest Biomass Estimation using Multi-modal Satellite Time-series},
  author={Nascetti, Andrea and Yadav, Ritu and Brodt, Kirill and Qu, Qixun and Fan, Hongwei and Shendryk, Yuri and Shah, Isha and Chung, Christine},
  booktitle={Thirty-seventh Conference on Neural Information Processing Systems Datasets and Benchmarks Track},
  year={2023}
}