GEDTM30

Overview

GEDTM30 is a Global 1-arc-second (~30m) Digital Terrain Model (DTM) built using a machine-learning-based data fusion approach. This dataset was generated using a global-to-local random forest model trained on ICEsat-2 and GEDI data, leveraging almost 30 billion of the highest-quality elevation points.

GEDTM30 is also used to generate 15 land surface parameters at six scales (30, 60, 120, 240, 480 and 960m), covering aspects of topographic position, light and shadow, landform characteristics, and hydrology. The repository demonstrates the modeling and land surface parameters / terrain derivation.

Data Components

1. GEDTM30: Terrain Height Prediction

Represents the predicted terrain height.

2. Uncertainty Map of Terrain Prediction

Provides an uncertainty map of the terrain prediction, derived from the standard deviation of individual tree predictions in the Random Forest model.

3. Global-to-local pixel selection mask

A mask that indicates which pixel comes from the global model or local model.

Pixel Value	Source / Category	Description
0	Global-to-Local Model	Represents areas where the global-to-local model (or downscaled data) is applied
1	Global Model	Represents areas where the global model data is used directly without local refinement.
2	Open water	Represent the open water pixels
255	Non-data (No Data)	No Data

4. 15 land surface parameters

Produced by DTM parametrization, representing different terrain features. Metadata of each parameter is currently stored at scale.csv. The optimized Equi7 tiling system for parameterization is currently stored at equi7_tiles.

• Landform: Slope in Degree, Geomorphons
• Light and Shadow: Positive Openness, Negative Openness, Hillshade
• Curvature: Minimal Curvature, Maximal Curvature, Profile Curvature, Tangential Curvature, Ring Curvature, Shape Index
• Local Topographic Position: Difference from Mean Elevation, Spherical Standard Deviation of the Normals
• Hydrology: Specific Catchment Area, LS Factor, Topographic Wetness Index, D8 flow direction

5. Version status

Version	Digital terrain model	Update record	Note
GEDTM30 v1.2	Recommended	Mar 6, 2026	1-arc-sec alignment, update LSPs.
GEDTM30 v1.1	Usable	Jun 11, 2025	Fixed surface water, tile border value jumping.
GEDTM30 v1.01	Deprecated	Jun 11, 2025	Duplicated version of v1.1 with files error.
GEDTM30 v1	Not recommended	May 22, 2025	1st peer-reviewed version
GEDTM30 v0	Deprecated	Feb 24, 2025
EDTM v1.1	Not recommended	Feb 25, 2023	MERITDEM is exclusively presented at some pixels
EDTM v1	Deprecated	Feb 13, 2023	Removed. License issue of FABDEM

Metadata and Scaling

Layer	Scale	Data Type	No Data
Ensemble Digital Terrain Model	10	Int32	-2,147,483,647
Standard Deviation EDTM	100	UInt16	65,535
Difference from Mean Elevation	100	Int16	32,767
Geomorphons	1	Byte	255
Hillshade	1	UInt16	65,535
LS Factor	1,000	UInt16	65,535
Maximal Curvature	1,000	Int16	32,767
Minimal Curvature	1,000	Int16	32,767
Negative Openness	100	UInt16	65,535
Positive Openness	100	UInt16	65,535
Profile Curvature	1,000	Int16	32,767
Ring Curvature	10,000	Int16	32,767
Shape Index	1,000	Int16	32,767
Slope in Degree	100	UInt16	65,535
Specific Catchment Area	1,000	UInt16	65,535
Spherical Standard Deviation of the Normals	100	Int16	32,767
Tangential Curvature	1,000	Int16	32,767
Topographic Wetness Index	100	Int16	32,767
D8 flow direction	1	Byte	255

Usage

This dataset is designed for researchers, developers, and professionals working in earth sciences, GIS, and remote sensing. It can be integrated into various geospatial analysis workflows to enhance terrain representation and modeling accuracy. This dataset covers the entire world and is well-suited for applications in:

• Topography

• Hydrology

• Geomorphometry

• Others

Getting Started

• Access and test the model and parametrization, please clone this repository:

git clone https://codeberg.org/openlandmap/GEDTM30.git

• COG urls of GEDTM30 and Land surface parameters can be found in this csv here: metadata/cog_list.csv

• Download the data set from Zenodo (10.5281/zenodo.14900181)

• Access and visualize COGs in QGIS.

Please follow the step in the GIF below.

Instruction: right click --> copy link --> paste to QGIS Layer >> Add Layer >> Add Raster Layer >> select Protocol:HTTP(S), cloud, etc. >> paste the url and you can visualize in QGIS.

Citation

If you use this dataset in your research or application, please cite as:

@dataset{ho_2025_14900181,
  author       = {Ho, Yufeng and Hengl, Tomislav},
  title        = {Global Ensemble Digital Terrain Model 30m (GEDTM30)},
  month        = feb,
  year         = 2025,
  publisher    = {Zenodo},
  version      = {v20250130},
  doi          = {10.5281/zenodo.14900181},
  url          = {https://doi.org/10.5281/zenodo.14900181},
}

Technical documentation can be cited as:

@Article{yufengho2025GEDTM30,
AUTHOR = {Ho, Yu-Feng and Grohmann, Carlos H and Lindsay, John and Reuter, Hannes I and Parente, Leandro and Witjes, Martijn and Hengl, Tomislav},
TITLE = {{Global Ensemble Digital Terrain modeling and parametrization at 30 m resolution (GEDTM30): a data fusion approach based on ICESat-2, GEDI and multisource data}},
JOURNAL = {PeerJ},
VOLUME = {13},
YEAR = {2025?},
PAGES = {e19673},
DOI = {10.7717/peerj.19673}
}

License

This dataset is released under fully open license CC-BY 4.0.

Acknowledgements & Funding

This work is supported by OpenGeoHub Foundation and has received funding from the European Commission (EC) through the projects:

• Open-Earth-Monitor Cyberinfrastructure: Environmental information to support EU’s Green Deal (1 Jun. 2022 – 31 May 2026 - 101059548)

Contact

For any questions or contributions, feel free to open an issue or reach out via [yu-feng.ho@opengeohub.org].