Biomazon: A Multimodal Dataset for 3D Forest Structure and Biomass Modeling in the Amazon Basin

Version 1.0

Mandal, Sayan; Sedona, Rocco; Besnard, Simon; Urbazaev, Mikhail; Zandi, Ehsan; Cavallaro, Gabriele, 2026, "Biomazon: A Multimodal Dataset for 3D Forest Structure and Biomass Modeling in the Amazon Basin", https://doi.org/10.26165/JUELICH-DATA/2YP2PZ, Jülich DATA, V1

Learn about Data Citation Standards.

Dataset Metrics

0 Downloads

Description	Biomazon is a large-scale, multi-modal remote sensing benchmark dataset covering the Amazon basin, designed for wall-to-wall estimation of forest vertical structure and aboveground biomass from dense multi-sensor Earth Observation data. The dataset fuses six complementary input modalities -- Sentinel-2 L2A optical surface reflectance, ALOS-2 PALSAR-2 L-band SAR, Sentinel-1 C-band SAR (each in ascending and descending orbits), Dynamic World V1 land-use/land-cover, Copernicus GLO-30 DEM elevation, and AlphaEarth Embeddings (AEX) foundation-model features -- totalling 88 input bands. The prediction targets are GEDI spaceborne LiDAR Relative Height profiles (RH0--RH100 in meter, 101 percentile bands capturing the full canopy height distribution) and GEDI Aboveground Biomass Density (AGBD, in Mg/ha), together comprising 102 label bands at 20 m resolution. All data span the period April 2019 to March 2023. Biomazon is intended for developing and benchmarking models that predict these 3D forest structure and biomass targets from the combined optical, SAR, elevation, land-cover, and foundation-model embedding inputs. The dataset is tiled on the Harmonized Landsat-Sentinel (HLS) MGRS tiling grid and delivered as 20m 256x256-pixel patches (5.12 km x 5.12 km) with 50% overlap, stored in HDF5 files (one per split: train / val / test). (2026-04-27)
Subject	Earth and Environmental Sciences
Keyword	benchmark dataset, GEDI, full RH profile, relative height, canopy height, aboveground biomass, multimodal, Amazon basin, Sentinel-2, Sentinel-1, ALOS-2 PALSAR-2, Copernicus GLO30 DEM, deep learning, joint modeling
Related Publication	Biomazon: A Multimodal Dataset for 3D Forest Structure and Biomass Modeling in the Amazon Basin To be submitted to IEEE JSTARS Special Issue "AI-Driven Multimodal Remote Sensing for Forestry Monitoring and Management"
Institute	JSC
Storage Location URL	https://datapub.fz-juelich.de/biomazon/

	1 File
	DATASET_SHA256SUMS.txt Plain Text - 2.3 KB - May 7, 2026 - 0 Downloads SHA-256: 81376fb74da47f7c31cf45bac89d957e838d3f2c8e406e4849474eaf56419823 SHA-256 checksum manifest for the Biomazon data release. Each line contains the SHA-256 hash and relative file path of a released file, enabling users to verify download completeness and file integrity after transfer or extraction.	Download

Citation Metadata

Dataset Persistent ID	doi:10.26165/JUELICH-DATA/2YP2PZ
Publication Date	2026-05-07
Title	Biomazon: A Multimodal Dataset for 3D Forest Structure and Biomass Modeling in the Amazon Basin
Author	Mandal, Sayan (Jülich Supercomputing Centre) - ORCID: https://orcid.org/0000-0003-3637-6029 Sedona, Rocco (Jülich Supercomputing Centre) - ORCID: https://orcid.org/0000-0003-4089-972X Besnard, Simon (GFZ Helmholtz Centre for Geosciences) - ORCID: https://orcid.org/0000-0002-1137-103X Urbazaev, Mikhail (GFZ Helmholtz Centre for Geosciences) - ORCID: https://orcid.org/0000-0002-0327-6278 Zandi, Ehsan (Jülich Supercomputing Centre) - ORCID: https://orcid.org/0000-0003-0135-257X Cavallaro, Gabriele (Jülich Supercomputing Centre) - ORCID: https://orcid.org/0000-0002-3239-9904
Contact	Use email button above to contact. Mandal, Sayan (Jülich Supercomputing Centre) Sedona, Rocco (Jülich Supercomputing Centre)
Description	Biomazon is a large-scale, multi-modal remote sensing benchmark dataset covering the Amazon basin, designed for wall-to-wall estimation of forest vertical structure and aboveground biomass from dense multi-sensor Earth Observation data. The dataset fuses six complementary input modalities -- Sentinel-2 L2A optical surface reflectance, ALOS-2 PALSAR-2 L-band SAR, Sentinel-1 C-band SAR (each in ascending and descending orbits), Dynamic World V1 land-use/land-cover, Copernicus GLO-30 DEM elevation, and AlphaEarth Embeddings (AEX) foundation-model features -- totalling 88 input bands. The prediction targets are GEDI spaceborne LiDAR Relative Height profiles (RH0--RH100 in meter, 101 percentile bands capturing the full canopy height distribution) and GEDI Aboveground Biomass Density (AGBD, in Mg/ha), together comprising 102 label bands at 20 m resolution. All data span the period April 2019 to March 2023. Biomazon is intended for developing and benchmarking models that predict these 3D forest structure and biomass targets from the combined optical, SAR, elevation, land-cover, and foundation-model embedding inputs. The dataset is tiled on the Harmonized Landsat-Sentinel (HLS) MGRS tiling grid and delivered as 20m 256x256-pixel patches (5.12 km x 5.12 km) with 50% overlap, stored in HDF5 files (one per split: train / val / test). (2026-04-27)
Subject	Earth and Environmental Sciences
Keyword	benchmark dataset GEDI full RH profile relative height canopy height aboveground biomass multimodal Amazon basin Sentinel-2 Sentinel-1 ALOS-2 PALSAR-2 Copernicus GLO30 DEM deep learning joint modeling
Related Publication	Biomazon: A Multimodal Dataset for 3D Forest Structure and Biomass Modeling in the Amazon Basin To be submitted to IEEE JSTARS Special Issue "AI-Driven Multimodal Remote Sensing for Forestry Monitoring and Management"
Grant Information	3D-ABC, Helmholtz Foundation Model Initiative GCS/NIC: 3d-abc: 61954
Depositor	Mandal, Sayan
Deposit Date	2026-04-27
Time Period Covered	Start: 2019-04-01 ; End: 2023-03-30
Software	Python, Version: 3.11.9 earthengine-api, Version: 1.6.3 gediDB, Version: 2025.4.25
Data Sources	(A) Sentinel-2 L2A SR - Contains modified Copernicus Sentinel data 2019-2023. - European Space Agency, “Copernicus Sentinel-2 (processed by ESA), 2021, MSI Level-2A BOA Reflectance Product. Collection 1,” 2021. [Online]. Available: https://doi.org/10.5270/S2_-znk9xsj - Google Earth Engine Data Catalog, “Harmonized Sentinel-2 MSI: Mul- tiSpectral Instrument, Level-2A (SR),” https://developers.google.com/earth-engine/datasets/catalog/COPERNICUS_S2_SR_HARMONIZED, [Accessed 20-10-2025].; (B) Cloud Score+ - V. J. Pasquarella, C. F. Brown, W. Czerwinski, and W. J. Rucklidge, “Comprehensive quality assessment of optical satellite imagery using weakly supervised video learning,” in 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), 2023, pp. 2125–2135. - Google Earth Engine Data Catalog, “Cloud Score+ S2 HARMONIZED V1,” https://developers.google.com/earth-engine/datasets/catalog/GOOGLE_CLOUD_SCORE_PLUS_V1_S2_HARMONIZED, [Accessed 20-10-2025].; (C) Sentinel-1 - Contains modified Copernicus Sentinel data 2019-2023. - Google Earth Engine Data Catalog, “Sentinel-1 SAR GRD: C- band Synthetic Aperture Radar Ground Range Detected, log scaling,” https://developers.google.com/earth-engine/datasets/catalog/COPERNICUS_S1_GRD, [Accessed 07-10-2025]; (D) ALOS-2 PALSAR-2 ScanSAR Level 2.2 - This data have been provided by Earth Observation Research Center (EORC) of Japan Aerospace Exploration Agency (JAXA) via Google Earth Engine - Google Earth Engine Data Catalog, “PALSAR-2 ScanSAR Level 2.2,” https://developers.google.com/earth-engine/datasets/catalog/JAXA_ALOS_PALSAR-2_Level2_2_ScanSAR, [Accessed 06-10-2025].; (E) Copernicus GLO-30 DEM - © DLR e.V. 2010-2014 and © Airbus Defence and Space GmbH 2014-2018 provided under COPERNICUS by the European Union and ESA; all rights reserved. - Google Earth Engine Data Catalog, “Copernicus DEM GLO-30: Global 30m Digital Elevation Model ,” https://developers.google.com/earth-engine/datasets/catalog/COPERNICUS_DEM_GLO30, [Accessed 06-10-2025]; (F) Dynamic World V1 LULC - C. F. Brown, S. P. Brumby, B. Guzder-Williams, T. Birch, S. B. Hyde, J. Mazzariello, W. Czerwinski, V. J. Pasquarella, R. Haertel, S. Ilyushchenko et al., “Dynamic world, near real-time global 10 m land use land cover mapping,” Scientific data, vol. 9, no. 1, p. 251, 2022. - Google Earth Engine Data Catalog, “Dynamic World V1,” https://developers.google.com/earth-engine/datasets/catalog/GOOGLE_DYNAMICWORLD_V1, [Accessed 04-10-2025].; (G) AlphaEarth Embedding - The AlphaEarth Foundations Satellite Embedding dataset is produced by Google and Google DeepMind. - C. F. Brown, M. R. Kazmierski, V. J. Pasquarella, W. J. Rucklidge, M. Samsikova, C. Zhang, E. Shelhamer, E. Lahera, O. Wiles, S. Ilyushchenko, N. Gorelick, L. L. Zhang, S. Alj, E. Schechter, S. Askay, O. Guinan, R. Moore, A. Boukouvalas, and P. Kohli, “Alphaearth foundations: An embedding field model for accurate and efficient global mapping from sparse label data,” 2025. [Online]. Available: https://arxiv.org/abs/2507.22291 - Google Earth Engine Data Catalog, “Satellite Embedding V1,” https://developers.google.com/earth-engine/datasets/catalog/GOOGLE_SATELLITE_EMBEDDING_V1_ANNUAL, [Accessed 17-09-2025].; (H) GEDI L2A Relative Height and L4A Above Ground Biomass Density - Besnard et al., (2025). gediDB: A toolbox for processing and providing Global Ecosystem Dynamics Investigation (GEDI) L2A-B and L4A-C data. Journal of Open Source Software, 10(113), 8593, https://doi.org/10.21105/joss.08593

FZJ Metadata

Institute	JSC
PoF IV topic	Domain-Specific Simulation & Data Life Cycle Labs (SDLs) and Research Groups (POF4-5111)
Storage Location URL	https://datapub.fz-juelich.de/biomazon/

Waiver

Our Community Norms as well as good scientific practices expect that proper credit is given via citation. Please use the data citation above, generated by the Dataverse.

CC0 - "Public Domain Dedication" Creative Commons CC0 1.0 Public Domain Dedication icon

Guestbook

No guestbook is assigned to this dataset, you will not be prompted to provide any information on file download.

Preview Guestbook

Upon downloading files the guestbook asks for the following information.

Guestbook Name

Collected Data

Account Information

Dataset	Summary	Contributors	Published
No records found.

Share Dataset

Share this dataset on your favorite social media networks.

Dataset Citations

Citations for this dataset are retrieved from Crossref via DataCite using Make Data Count standards. For more information about dataset metrics, please refer to the User Guide.

Sorry, no citations were found.

Restricted Files Selected

The restricted file(s) selected may not be downloaded because you have not been granted access.

Restricted Files Selected

The restricted file(s) selected may not be downloaded because you have not been granted access.

Click Continue to download the files you have access to download.

Delete Dataset

Are you sure you want to delete this dataset and all of its files? You cannot undelete this dataset.

Delete Draft Version

Are you sure you want to delete this draft version? Files will be reverted to the most recently published version. You cannot undelete this draft.

Unpublished Dataset Private URL

Use a Private URL to allow those without Dataverse accounts to access your unpublished dataset. For more information about the Private URL feature, please refer to the User Guide.

Private URL has not been created.

Unpublished Dataset Private URL

Are you sure you want to disable the Private URL? If you have shared the Private URL with others they will no longer be able to use it to access your unpublished dataset.

Delete Files

The file(s) will be deleted after you click on the Delete button.

Files will not be removed from previously published versions of the dataset.

Compute

This dataset contains restricted files you may not compute on because you have not been granted access.

Deaccession Dataset

Are you sure you want to deaccession? The selected version(s) will no longer be viewable by the public.

Deaccession Dataset

Are you sure you want to deaccession this dataset? It will no longer be viewable by the public.

Version Differences Details

Please select two versions to view the differences.

Version Differences Details

Version:
Last Updated:

Select File(s)

Please select a file or files to be downloaded.

Select File(s)

Please select a file or files for access request.

Select File(s)

Please select a file or files to be deleted.

Select File(s)

Please select unrestricted file(s) to be restricted.

Select File(s)

Please select restricted file(s) to be unrestricted.

Select File(s)

Please select a file or files to be edited.

Select File(s)

Please select a file or files to be edited.

Edit Tags

Select existing file tags or create new tags to describe your files. Each file can have more than one tag.

Request Access

You need to Log In to request access to this file.

Dataset Terms

Please confirm and/or complete the information needed below in order to continue.

Package File Download

Use the Download URL in a Wget command or a download manager to download this package file. Download via web browser is not recommended. User Guide - Downloading a Dataverse Package via URL

- -

Download URL

https://data.fz-juelich.de/api/access/datafile/

Request Access

Please confirm and/or complete the information needed below in order to request access to files in this dataset.

Compute Batch

Clear Batch

Dataset	Dataset Persistent ID

Compute Batch

File Restrictions

Terms of Access

Request Access

Enable access request

Submit for Review

You will not be able to make changes to this dataset while it is in review.

Publish Dataset

Are you sure you want to publish this dataset? Once you do so it must remain published.

Publish Dataset

This dataset cannot be published until Earth System Science is published. Would you like to publish both right now?

Once you publish this dataset it must remain published.

Publish Dataset

Are you sure you want to republish this dataset?

Select if this is a minor or major version update.

Minor Release (1.1)

Major Release (2.0)

Publish Dataset

This dataset cannot be published until Earth System Science is published by its administrator.

Publish Dataset

This dataset cannot be published until Earth System Science and Jülich Supercomputing Centre (JSC) are published.

Return to Author

Return this dataset to contributor for modification.