Satellite-to-DEM Translation Dataset: Landsat, Sentinel, and 30 m DEMs of the Iberian Peninsula
收藏Zenodo2026-04-22 更新2026-05-26 收录
下载链接:
https://zenodo.org/doi/10.5281/zenodo.14647631
下载链接
链接失效反馈官方服务:
资源简介:
S. Leyva, M. Ortega, J. de Moura, " Deep Learning Approaches to DEM Generation: Attention Mechanisms and Gradient-Based Loss Function ", PFG – Journal of Photogrammetry, Remote Sensing and Geoinformation Science, 2026. DOI : 10.1007/s41064-026-00393-y
This dataset is provided solely for statistical and research purposes. No private data or fees are required for its use.
Dataset Description
This dataset was specifically designed for satellite-to-DEM translation tasks, covering the Iberian Peninsula, a region characterized by diverse landscapes, ecosystems, terrain types, and land uses. It combines multi-source satellite imagery and elevation data to provide a comprehensive representation of the region's geographic and topographic diversity.
The dataset includes:
Harmonized Sentinel-2 MSI imagery (Level-2A):High-resolution multi-spectral images filtered for the period June 1, 2023, to August 31, 2024, with RGB bands resampled to 30 meters.
USGS Landsat 9 imagery (Level 2, Collection 2, Tier 1):Atmospherically corrected surface reflectance images filtered for June 1, 2022, to September 30, 2024, with RGB bands at 30 meters resolution.
Copernicus DEM GLO-30 elevation model:A global digital elevation model derived from radar data with a resolution of 30 meters.
All data were processed in the EPSG:3035 projected coordinate system for minimal distortion and clipped to the Iberian Peninsula. The dataset was partitioned into 256 × 256 pixel tiles (~7.68 × 7.68 km) for computational efficiency, resulting in a total of 10,237 non-overlapping tiles for each dataset. Images were normalized and preprocessed to remove cloud artifacts and ensure high-quality inputs for machine learning tasks.
This dataset serves as a robust resource for geospatial and remote sensing research, especially for machine learning applications in terrain modeling, environmental analysis, and Earth observation studies.
Acknowledgments
This dataset was created using publicly available data from multiple sources. The authors acknowledge the contributions of the following:
Harmonized Sentinel-2 MSI imagery:Contains modified Copernicus Sentinel data [2026]. The data is provided under the Copernicus Sentinel Data Terms and Conditions.© European Union, Copernicus Sentinel data [2026].
USGS Landsat 9 imagery:Landsat 9 data courtesy of the U.S. Geological Survey. These datasets are in the public domain and may be freely used, transferred, or reproduced without copyright restrictions.
Copernicus DEM GLO-30:Contains data produced using Copernicus WorldDEM-30 © 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.
提供机构:
Zenodo
创建时间:
2025-01-14



