EO missions

EO data spans a range of processing levels and thematic focuses, each suited to different analytical needs. Data providers typically distribute measurements across standardized processing levels (from raw instrument data to results from analyses of lower-level data) and most ML workflows operate on data with varying degrees of preprocessing. At its core, this ecosystem draws on a rich archive of satellite imagery from well-established missions (e.g., Sentinel-1, Sentinel-2, Sentinel-3, Landsat, MODIS, VIIRS, HLS, ALOS PALSAR), complemented by higher-resolution commercial data (e.g., Planet, Airbus, and Vantor, formerly known as Maxar) and specialized missions with a targeted scope (e.g., GRACE, ICESat-2, GEDI, BIOMASS). In Table 2.1, we summarize the key attributes of widely used missions that structure this diverse data landscape.

Table 2.1: Overview of widely used Earth observation missions. Revisit times assume full constellation availability. Spatial resolution ranges reflect different acquisition modes or spectral bands. Coverage symbols: 🌐 = global (land & ocean), 🌿 = land surfaces and coastal areas only, 🎯 = tasked/on-demand, * = the failure of Sentinel-1B induced significant data gaps, see Sentinel-1 Data Gaps.
Mission Sensor Spatial res. Revisit Cov. Active Typical use cases
Sentinel-1 SAR-C 5–20 m 6 days 🌐* 2014– Flood, oil spill mapping
Sentinel-2 Optical 10–60 m 5 days 🌿 2015– Land cover, agriculture
Sentinel-3 Optical 300–1200 m \(<2\) days 🌐 2016– Ocean color, SST
Landsat Optical 15–120 m 8-18 days 🌿 1972– Land cover change
MODIS Optical 250–1000 m 1–2 days 🌐 1999– Vegetation, fire, albedo
VIIRS Optical 375–750 m \(\sim 1\) day 🌐 2011– Nighttime lights, fire
ALOS PALSAR SAR-L 6–100 m 14–46 days 🌐 2006– Forest mapping
NISAR SAR-L/S 10–20 m 12 days 🌿 2025– Land/forests/ice changes
Planet (Dove) Optical 3–5 m Daily 🌿 2017– Change detection, agriculture
Planet (SkySat) Optical 0.5 m 🎯 🎯 2016– Sub-daily site-monitoring
Vantor (WorldView) Optical 0.3–2 m 🎯 🎯 2001– Change detection, defense
Airbus (PlΓ©iades/SPOT) Optical 0.3–6 m 🎯 🎯 2002– Large-area mapping, 3D
GEDI LiDAR 25 m footprint Sparse Β±51.6Β° 2019– Canopy height, biomass
ICESat-2 LiDAR 11 m footprint 91 days 🌐 2018– Ice sheet elevation
GRACE / GRACE-FO Gravimetry \(\sim 300\) km 30 days 🌐 2002– Groundwater, ice mass

Beyond satellite observations, derived products constitute a second category of EO data. These include land use and land cover maps (e.g., Copernicus Global Dynamic Land Cover [1]), terrain models (e.g., Copernicus GLO-30 Digital Elevation Model [2], NASA SRTM [3]), atmospheric variables (e.g., ERA5 [4]), climate (e.g., TerraClimate [5]) and vegetation (e.g., MOD13 [6]) indices, population density (e.g., WorldPop [7]) and numerous other thematic data layers. These products combine observations from one or more satellite missions with domain knowledge and modeling capabilities (that may or may not incorporate ML methods) to produce analysis-ready information. A more recent addition to this category is pre-computed geospatial embeddings: dense vector representations of locations on Earth produced by large models trained on satellite imagery and often complementary EO modalities. Unlike traditional thematic layers, which target a specific variable interpretable in its own units, embeddings are purportedly general-purpose and task-agnostic. Recent examples include AlphaEarth Foundations [8], TESSERA [9] and ESL [10]. Like other derived products, embeddings inherit the uncertainties of their upstream inputs, compounded by the biases of their pre-training data and learning objectives, caveats that should be accounted for in downstream analyses.

A more accessible entry point into EO data is offered by curated EO datasets for ML: preprocessed samples of the above-mentioned data sources, purportedly ready to feed into ML training pipelines. Such datasets are typically assembled by individual researchers [11], [12], [13], [14], [15] or through community efforts such as TorchGeo [16] or Major TOM [17], and hosted on platforms like Zenodo [18], HuggingFace [19], or Source Cooperative1. They lower the barrier to entry for ML practitioners who would otherwise face the non-trivial task of assembling training-ready inputs from raw or derived EO products. The trade-off, however, is a loss of control: adopting an existing curated dataset means inheriting its creators’ design choices (spatiotemporal coverage, selected data sources, sampling strategy, processing pipeline) which may not match the requirements of another downstream task.

The choice among these dataset categories carries important implications for research design. Furthermore, selecting any data source requires weighing its spatiotemporal coverage, the reliability of its long-term open accessibility, and the practical costs of download and storage.

[1]
Copernicus Global Land Service, β€œCopernicus Global Land Cover 100m - Algorithm Theoretical Basis Document (ATBD), User Manual.” 2021. Available: https://land.copernicus.eu/en/technical-library/global-dynamic-land-cover-algorithm-theoretical-basis-document-v3.0
[2]
ESA, β€œCopernicus Digital Elevation Model,” Copernicus DEM. European Space Agency, 2022. doi: 10.5270/esa-c5d3d65.
[3]
NASA JPL, β€œNASA Shuttle Radar Topography Mission Global 1 arc second number.” NASA Land Processes Distributed Active Archive Center, 2013. doi: 10.5067/MEASURES/SRTM/SRTMGL1N.003.
[4]
H. Hersbach et al., β€œThe ERA5 global reanalysis,” Quarterly Journal of the Royal Meteorological Society, vol. 146, no. 730, pp. 1999–2049, 2020, doi: 10.1002/qj.3803.
[5]
J. T. Abatzoglou, S. Z. Dobrowski, S. A. Parks, and K. C. Hegewisch, β€œTerraClimate, a high-resolution global dataset of monthly climate and climatic water balance from 1958–2015,” Scientific Data, vol. 5, no. 1, Jan. 2018, doi: 10.1038/sdata.2017.191.
[6]
K. Didan, β€œMODIS/Terra Vegetation Indices 16-Day L3 Global 250m SIN Grid V061.” NASA Land Processes Distributed Active Archive Center, 2021. doi: 10.5067/MODIS/MOD13Q1.061.
[7]
A. J. Tatem, β€œWorldPop, open data for spatial demography,” Scientific Data, vol. 4, no. 1, Jan. 2017, doi: 10.1038/sdata.2017.4.
[8]
C. F. Brown et al., β€œAlphaEarth Foundations: An embedding field model for accurate and efficient global mapping from sparse label data,” arXiv preprint arXiv:2507.22291, 2025, Available: https://arxiv.org/abs/2507.22291
[9]
Z. Feng et al., β€œTESSERA: Temporal Embeddings of Surface Spectra for Earth Representation and Analysis,” arXiv preprint arXiv:2506.20380, 2025, Available: https://arxiv.org/abs/2506.20380
[10]
S. Chen et al., β€œDemocratizing planetary-scale analysis: An ultra-lightweight earth embedding database for accurate and flexible global land monitoring,” arXiv preprint arXiv:2601.11183, 2026, Available: https://arxiv.org/abs/2601.11183
[11]
Y. Wang, N. A. A. Braham, Z. Xiong, C. Liu, C. M. Albrecht, and X. X. Zhu, β€œSSL4EO-S12: A large-scale multi-modal, multi-temporal dataset for self-supervised learning in earth observation,” arXiv preprint arXiv:2211.07044, 2022.
[12]
A. Toker et al., β€œDynamicEarthNet: Daily multi-spectral satellite dataset for semantic change segmentation,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2022.
[13]
V. Sainte Fare Garnot and L. Landrieu, β€œPanoptic segmentation of satellite image time series with convolutional temporal attention networks,” ICCV, 2021.
[14]
E. Plekhanova et al., β€œSSL4Eco: A global seasonal dataset for geospatial foundation models in ecology,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2025.
[15]
F. Fogel et al., β€œOpen-canopy: Towards very high resolution forest monitoring,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition (CVPR), Jun. 2025, pp. 1395–1406.
[16]
A. J. Stewart, C. Robinson, I. A. Corley, A. Ortiz, J. M. Lavista Ferres, and A. Banerjee, β€œTorchGeo: Deep Learning With Geospatial Data,” ACM Transactions on Spatial Algorithms and Systems, vol. 11, no. 4, pp. 1–28, Aug. 2025.
[17]
A. Francis and M. Czerkawski, β€œMajor TOM: Expandable Datasets for Earth Observation,” in IGARSS 2024 - 2024 IEEE international geoscience and remote sensing symposium, 2024, pp. 2935–2940. doi: 10.1109/IGARSS53475.2024.10640760.
[18]
European Organization For Nuclear Research and OpenAIRE, β€œZenodo.” CERN, 2013. doi: 10.25495/7GXK-RD71.
[19]
Q. Lhoest et al., β€œDatasets: A Community Library for Natural Language Processing,” in Proceedings of the 2021 conference on empirical methods in natural language processing: System demonstrations, Online; Punta Cana, Dominican Republic: Association for Computational Linguistics, Nov. 2021, pp. 175–184. Available: https://aclanthology.org/2021.emnlp-demo.21

  1. https://source.coop/productsβ†©οΈŽ