{"success":true,"data":{"author":"NASA-IBM AI4Science (Surya / SuryaBench team); Daniel da Silva (NASA GSFC)","bucketAccessType":"public","bucketName":"nasa-ibm-ai4science/surya-bench-coronal-extrapolation","bucketPrefix":"","category":"Astronomy & Space","cloudProvider":"huggingface","createdAt":"2026-11-16T02:38:29.645Z","description":"Benchmark targets for emulating global coronal magnetic field models from photospheric magnetograms, part of SuryaBench, the downstream task suite of the NASA-IBM Surya heliophysics foundation model. Each daily instance is the output of the physics-based ADAPT-WSA Potential Field Source Surface (PFSS) model driven by SDO/HMI magnetograms: spherical harmonic coefficients of the coronal magnetic potential between the photosphere and a 2.51 solar radius source surface, stored in FITS HDU 3 as a (2, 91, 91) triangular, Schmidt-normalised array up to order 90, alongside photospheric polarity maps and open/closed field-line masks from WSA. All 12 ADAPT ensemble realizations (R000 to R011) are included for 5,347 days from 13 May 2010 to 31 December 2024. Splits are given as train.csv (15 February to 31 December, 2010 to 2019), validation.csv and leaky_validation.csv (January and early February windows) and test.csv (2020 to 2024), each listing file paths into data.tar.gz (about 36 GB compressed). Tasks: PFSS surrogate modelling, coronal field reconstruction and open-flux and solar wind connectivity for space weather. Read with astropy. CC-BY-4.0 as stated in the card; contact Daniel da Silva (NASA GSFC).","displayTitle":"SuryaBench Coronal Field Extrapolation: ADAPT-WSA PFSS Spherical Harmonic Coefficients from SDO/HMI (2010-2024)","id":"ac83a02e-7568-52c4-acdb-b8fa46a25b74","license":"CC-BY-4.0","licenseEvidenceUrl":"https://huggingface.co/datasets/nasa-ibm-ai4science/surya-bench-coronal-extrapolation","objectCountEstimate":9,"probeStatus":"ok","referenceUrl":"https://github.com/NASA-IMPACT/Surya","region":"us","schemaCard":{"accessHints":[{"language":"python-datasets","snippet":"from datasets import load_dataset\nds = load_dataset(\"nasa-ibm-ai4science/surya-bench-coronal-extrapolation\")"},{"language":"python-hub","snippet":"from huggingface_hub import snapshot_download\nsnapshot_download(repo_id=\"nasa-ibm-ai4science/surya-bench-coronal-extrapolation\", repo_type=\"dataset\")"}],"caveats":["HuggingFace metadata: author=nasa-ibm-ai4science, license:cc-by-4.0, size_categories:1K<n<10K, format:csv, downloads=105"],"format":"mixed","generatedAt":"2026-11-16T02:38:35.807Z","layout":{"example":"leaky_validation.csv","pattern":"{repoId}/{filename}"},"mountPath":"/mnt/archive","sampleHeaders":[{"bytes":"timestamp,R000,R001,R002,R003,R004,R005,R006,R007,R008,R009,R010,R011\n2011-01-01 08:00:00,data/wsa_201101010800R000_ahmi.fits,data/wsa_201101010800R001_ahmi.fits,data/wsa_201101010800R002_ahmi.fits,data/wsa_201101010800R003_ahmi.fits,data/wsa_201101010800R004_ahmi.fits,data/wsa_201101010800R005_ahmi.fits,data/wsa_201101010800R006_ahmi.fits,data/wsa_201101010800R007_ahmi.fits,data/wsa_201101010800R008_ahmi.fits,data/wsa_201101010800R009_ahmi.fits,data/wsa_201101010800R010_ahmi.fits,data/wsa_201101010800R011_","encoding":"utf8","key":"leaky_validation.csv","truncatedAt":512}],"sampleObjects":["CoronalFieldExtrapolation_examples.tar","data.tar.gz","leaky_validation.csv","test.csv","train.csv"],"sidecars":[{"bytes":"---\nlicense: cc-by-4.0\nsize_categories:\n- 100B<n<1T\ntags:\n- Helio\n- coronal-extrapolation\nconfigs:\n- config_name: default\n  default: true\n  data_files:\n  - split: train\n    path: train.csv\n  - split: validation\n    path: validation.csv\n  - split: test\n    path: test.csv\n  - split: leaky_validation\n    path: leaky_validation.csv\n---\n# Coronal Field Extrapolation Dataset\n\n## Dataset Summary\n\nThis dataset contains spherical harmonic coefficients of the coronal magnetic potential generated by emulating the physics-based **ADAPT-WSA PFSS** (Potential Field Source Surface) code, driven by **SDO/HMI** solar magnetogram observations. The target spherical harmonic coefficients represent the magnetic potential between the photosphere and the source surface (set to 2.51 Rs).  \nEach file also contains additional variables from the WSA model output, such as:  \n- Photospheric polarity maps  \n- Open/closed field line masks  \n\nBecause ADAPT-WSA is an ensemble model, data is included for all **12 realizations** (`R000` – `R011`) in the filenames. Spherical harmonic data are stored in FITS HDU index **3**, formatted as a `(2, 91, 91)` array. Each `(91, 91)` array is triangular, with unused elements left as zero. The dataset covers May 2010 to December 2024 and includes daily simulations, with splits designed for consistent training, validation, and testing across years.\n\n\n## Supported Tasks and Applications\n\n- **Surrogate modeling**: Training ML models to emulate ADAPT-WSA PFSS outputs  \n- **Coronal field reconstruction**: Predicting global magnetic field structure from photospheric inputs  \n- **Space weather forecasting**: Estimating open/closed field lines and solar wind connectivity  \n- **Physics emulation**: Foundation models replacing computationally expensive PFSS runs  \n\n\n## Data Structure\n\n### Data Files\n\n- `train.csv`: Instances from Feb 15 to Dec 31 in each year between 2010–2019  \n- `validation.csv`: Instances from Jan 15–28 of each year between 2010–2019  \n- `test.csv`: All instances from each year between 2020–2024  \n- `leaky_validation.csv`: Instances from Jan 1–14 and Jan 29–Feb 11 of each year between 2010–2019  \n\nCompressed simulation model outputs containing spherical harmonics are available in `data.tar.gz`. Individual files are stored in FITS format. There are 12 ensemble realizations per one timestamp. Each file includes:  \n  - Spherical harmonic coefficients (HDU index 3)  \n  - Photospheric polarity information  \n  - Open/closed field masks  \n- The paths to realizations (`R000`–`R011`) are provided in above mentioned csv files. \n\n### Features\n\n- **Spherical Harmonics**: `(2, 91, 91)` triangular arrays  \n  - Order: up to 90  \n  - Normalization: Schmidt  \n\n\n## Dataset Details \n\n| Field | Description |\n|------------------------|---------------------------------------------|\n| **Temporal Coverage** | May 13, 2010 – Dec 31, 2024 |\n| **Data Format** | FITS |\n| **Data Shape** | (2, 91, 91) per instance |\n| **Data Size** | 5,347 instances * 12 realizations  |\n| **Cadence** | 1 day |\n| **Total File Size** | ~36.2 GB (compressed) |\n| **Size per file** | ~763 KB |\n\n\n## Example Code (Python)\n\n```python\nfrom astropy.io import fits\nimport numpy as np\nimport torch\n\nfits_file = fits.open(\"wsa_201005010800R000_ahmi.fits\")\nsph_data = fits_file[3].data.copy()\nfits_file.close()\n\ncoefficients = torch.from_numpy(\n    np.array(\n        [\n            sph_data[0, :, :][np.triu_indices(sph_data.shape[1])],\n            sph_data[1, :, :][np.triu_indices(sph_data.shape[1])],\n        ]\n    )\n)\n```\n\n## License\nThis dataset is licensed under the Creative Commons Attribution 4.0 International (CC BY 4.0) License.\n\n\n## Contact\n\nDaniel da Silva, [daniel.e.dasilva@nasa.gov](daniel.e.dasilva@nasa.gov)\n\n![WSA Fields](wsa_fields.png \"WSA Fields\")","encoding":"utf8","key":"README.md","name":"README.md","size":3801,"truncated":false}],"sourceObjectKey":"leaky_validation.csv","starterCell":"import pandas as pd\ndf = pd.read_csv(\"hf://datasets/nasa-ibm-ai4science/surya-bench-coronal-extrapolation/leaky_validation.csv\")\ndf.head()","starterSamplePath":"leaky_validation.csv","totalObjects":9},"summary":"Benchmark targets for emulating global coronal magnetic field models from photospheric magnetograms, part of SuryaBench, the downstream task suite of the NASA-IBM Surya…","title":"SuryaBench Coronal Field Extrapolation: ADAPT-WSA PFSS Spherical Harmonic Coefficients from SDO/HMI (2010-2024)","updatedAt":"2026-11-16T16:52:31.701Z","visibility":"public","websiteUrl":"https://huggingface.co/datasets/nasa-ibm-ai4science/surya-bench-coronal-extrapolation"}}