{"success":true,"data":{"author":"oxai4science (University of Oxford), from NASA SDO AIA and HMI data","bucketAccessType":"public","bucketName":"oxai4science/sdoml-lite","bucketPrefix":"","category":"Astronomy & Space","cloudProvider":"huggingface","createdAt":"2026-11-22T01:09:23.264Z","description":"Full-disk images of the Sun from NASA's Solar Dynamics Observatory (SDO), resampled for machine learning in solar physics and space-weather forecasting. Each sample is one observation time at 15-minute cadence from 13 May 2010 to 31 July 2024 (about 500,000 timestamps) with six 512 x 512 float32 channels: extreme-ultraviolet and ultraviolet images from the Atmospheric Imaging Assembly (AIA) at 131, 171, 193, 211 and 1600 Angstrom, which trace hot coronal plasma, flares and the chromosphere, and a line-of-sight magnetogram, the surface magnetic field, from the Helioseismic and Magnetic Imager (HMI, hmi.M_720s). Inputs are AIA Level 1 FITS files and HMI data; AIA channels are normalized from statistics computed in a first processing pass. Files are WebDataset TAR archives, one per day (5,194 files, about 3 TB), each holding up to 96 observation sets as NumPy .npy files keyed YYYY/MM/DD/HHMM. SDOML-lite is a separate product from the SDOML dataset on AWS and is not format-compatible with it; a PyTorch loader is in the oxai4science/sdoml-lite GitHub repository. Uses: solar flare and coronal mass ejection forecasting, generative and self-supervised models of the Sun, active-region tracking, and EUV irradiance proxies. Released by oxai4science under CC BY 4.0; acknowledge NASA SDO and the AIA and HMI teams.","displayTitle":"SDOML-lite: SDO AIA and HMI Full-Disk Solar Images at 15-Minute Cadence, 2010-2024, for Machine Learning (Oxford)","huggingFaceUrl":"https://huggingface.co/datasets/oxai4science/sdoml-lite","id":"873ef5d7-55b7-5664-b755-67a205d3692a","license":"CC-BY-4.0","licenseEvidenceUrl":"https://huggingface.co/datasets/oxai4science/sdoml-lite","objectCountEstimate":5199,"probeStatus":"ok","referenceUrl":"https://github.com/oxai4science/sdoml-lite","region":"us","schemaCard":{"accessHints":[{"language":"python-datasets","snippet":"from datasets import load_dataset\nds = load_dataset(\"oxai4science/sdoml-lite\")"},{"language":"python-hub","snippet":"from huggingface_hub import snapshot_download\nsnapshot_download(repo_id=\"oxai4science/sdoml-lite\", repo_type=\"dataset\")"}],"caveats":["HuggingFace metadata: author=oxai4science, license:cc-by-4.0, size_categories:100K<n<1M, downloads=4021"],"format":"mixed","generatedAt":"2026-11-22T01:10:25.449Z","layout":{"example":"assets/ar1520.png","pattern":"{repoId}/{filename}"},"mountPath":"/mnt/archive","sampleHeaders":[{"bytes":"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","encoding":"hex","key":"assets/ar1520.png","truncatedAt":512}],"sampleObjects":["assets/.gitattributes","assets/ar1520.png","assets/usage_example_1.png","data/sdoml-lite-0001.tar","data/sdoml-lite-0002.tar"],"sidecars":[{"bytes":"---\nlicense: cc-by-4.0\ntags:\n- image\n- video\n- timeseries\n- forecasting\n- astrophysics\n- heliophysics\n- physics\n- scientific-data\n- climate\n- weather\n- space-weather\npretty_name: SDOML-lite\nsize_categories:\n- 100K<n<1M\n---\n<div align=\"left\">\n  <img width=\"100%\" src=\"assets/ar1520.png\">\n</div>\n\n# SDOML-lite\n\nSDOML-lite is a lightweight alternative to the [SDOML](https://registry.opendata.aws/sdoml-fdl/) dataset specifically designed for machine learning applications in solar physics, providing continuous full-disk images of the Sun with magnetic field and extreme ultraviolet data in several wavelengths. The data source is the [Solar Dynamics Observatory (SDO)](https://sdo.gsfc.nasa.gov/) space telescope, a NASA mission that has been in operation since 2010.\n\nNASA’s SDO mission has generated over 20 petabytes of high-resolution solar imagery since its launch in 2010, representing one of the most comprehensive records of science data ever collected. This dataset offers an extraordinary opportunity for scientific discovery and machine learning, enabling advances in space weather forecasting, climate modeling, solar flare and coronal mass ejection (CME) prediction, image-to-image translation, and unsupervised representation learning of solar dynamics. Yet, the sheer scale and complexity of the raw archive have made it largely inaccessible to the broader ML community. SDOML-lite bridges this gap by delivering a curated, normalized, and temporally consistent subset of SDO data, specifically designed for seamless integration into large-scale machine learning pipelines.\n\n_IMPORTANT: SDOML and SDOML-lite datasets are different in structure and data distributions. SDOML-lite is inspired by SDOML, but it is not based on SDOML data and there is no compatibility between the two formats._\n\n* **Temporal Coverage:** 13 May 2010 - 31 July 2024\n* **Data Source:** NASA SDO/AIA (Level 1 FITS files) and SDO/HMI (`hmi.M_720s` series data)\n* **Format:** WebDataset (one TAR file per day, `sdoml-lite-xxxx.tar`)\n* **Total samples:** ~ 500,000 timestamps\n* **Total Size:** ~ 3 TB\n* **Number of Chunks (Days):** 5194\n* **Image Resolution:** 512x512 pixels for all channels\n* **Data Type:** `float32` (NumPy `.npy` files)\n- **Channels per sample**: 6 (separate `.npy` files per channel). AIA: 131 Å, 171 Å, 193 Å, 211 Å, 1600 Å (`aia_0131`, `aia_0171`, `aia_0193`, `aia_211`, `aia_1600`). HMI: Line-of-sight magnetogram (`hmi_m`)\n* **Cadence:** Images are provided every 15 minutes within each daily TAR file. Each TAR file contains up to 96 observation sets.\n\n## Intended Uses\n\nThis dataset is designed for machine learning research at the intersection of solar physics, space weather, and climate science. Potential applications include:\n\n* Generative and self-supervised modeling of solar dynamics.\n* Spatiotemporal forecasting of solar flares and coronal mass ejections (CMEs).\n* Real-time space weather nowcasting from full-disk EUV and magnetic imagery.\n* Tracking and modeling the evolution of active regions and magnetic flux.\n* Learning data-driven proxies for solar irradiance for use in climate models.\n* Cross-channel image translation and modality alignment.\n\n## Dataset Structure\n\nThe dataset is organized into 5194 `.tar` files, with one `.tar` file per day within the temporal coverage period. Each daily `.tar` file contains multiple observation sets, one for every 15-minute interval of that day (up to 96 observation sets per day). An observation set consists of 6 NumPy (`.npy`) files, one for each specified channel.\n\nEach `.tar` file corresponds to one day of data and contains one `.npy` file per channel, per 15-minute timestamp. For each timestamp, the internal file path has the following structure.\n\n```\nYYYY/MM/DD/HHMM.aia_131.npy\nYYYY/MM/DD/HHMM.aia_171.npy\nYYYY/MM/DD/HHMM.aia_193.npy\nYYYY/MM/DD/HHMM.aia_211.npy\nYYYY/MM/DD/HHMM.aia_1600.npy\nYYYY/MM/DD/HHMM.hmi_m.npy\n```\n\nIn WebDataset convention, the `YYYY/MM/DD/HHMM` substring of the path becomes the key for each data sample, and the `aia_0131.npy`, `aia_0171.npy`, `aia_0193.npy`, `aia_0211.npy`, ","encoding":"utf8","key":"README.md","name":"README.md","size":9959,"truncated":true}],"sourceObjectKey":"assets/ar1520.png","starterCell":"from huggingface_hub import hf_hub_download\npath = hf_hub_download(repo_id=\"oxai4science/sdoml-lite\", filename=\"assets/ar1520.png\", repo_type=\"dataset\")\nprint(path)","starterSamplePath":"assets/ar1520.png","totalObjects":5199},"summary":"Full-disk images of the Sun from NASA's Solar Dynamics Observatory (SDO), resampled for machine learning in solar physics and space-weather forecasting.","title":"SDOML-lite: SDO AIA and HMI Full-Disk Solar Images at 15-Minute Cadence, 2010-2024, for Machine Learning (Oxford)","updatedAt":"2026-11-22T04:03:06.730Z","visibility":"public","websiteUrl":"https://huggingface.co/datasets/oxai4science/sdoml-lite"}}