--- license: other license_name: iras-mission-terms license_link: https://irsa.ipac.caltech.edu/Missions/iras.html pretty_name: "IRAS Faint Source Catalog" language: - en description: "The IRAS Faint Source Catalog (FSC), version 2.0, contains 173,044 infrared point sources detected by the Infrared Astronomical Satellite (IRAS) at 12, 25, 60, and 100 micrometres. Published by Moshir" task_categories: - tabular-classification tags: - space - infrared - astronomy - iras - mid-infrared - sky-survey - open-data - tabular-data - parquet size_categories: - 100K All-sky infrared survey mosaic showing the Milky Way in infrared light

Credit: NASA/GSFC

*Part of a [dataset collection](https://huggingface.co/collections/juliensimon/astronomy-datasets-69c24caf2f17e36128946743) on Hugging Face.* ## Dataset description The IRAS Faint Source Catalog (FSC), version 2.0, contains 173,044 infrared point sources detected by the Infrared Astronomical Satellite (IRAS) at 12, 25, 60, and 100 micrometres. Published by Moshir et al. (1992), the FSC extends the IRAS Point Source Catalog to fainter flux levels by applying more stringent processing at the cost of sky coverage (~75% of sky, avoiding the galactic plane and regions of high infrared cirrus). IRAS was a joint NASA/Netherlands/UK mission that performed the first sensitive all-sky survey in the mid-infrared (August 1983 – November 1983). Its four photometric bands — 12, 25, 60, and 100 μm — probed thermal emission from warm dust, revealing a zoo of previously hidden astronomical objects: ultraluminous infrared galaxies (ULIRGs) driven by starbursts and AGN, proto-planetary debris disks around main-sequence stars, asymptotic giant branch (AGB) stars with dust shells, and the large-scale structure of the interstellar medium through infrared cirrus emission. The four IRAS bands trace progressively cooler dust: 12 μm traces hot dust and PAH emission from star-forming regions; 25 μm traces warm circumstellar dust around evolved stars; 60 μm traces cool dust in star-forming regions and starburst galaxies; 100 μm traces cold interstellar dust (T ~ 20–30 K). The colour ratios between bands (e.g., F60/F25, F100/F60) are powerful diagnostics for classifying the dominant heating source. This dataset fills the mid-infrared gap in the collection: no other dataset covers all-sky mid-IR point sources across all source types. It serves as the reference catalog for identifying infrared counterparts of radio, optical, and X-ray sources, and for building multiwavelength SEDs. This dataset is suitable for **tabular classification** tasks. ## Schema | Column | Type | Description | Sample | Null % | |--------|------|-------------|--------|--------| | `iras_name` | str | IRAS source designation in format FHHMM.m+DDMM (prefix F = faint catalog) | F02558+0026 | 0.0% | | `ra_1950_deg` | float64 | Right ascension (B1950 epoch, degrees) — note: 1950 epoch, not J2000 | 43.960833333333326 | 0.0% | | `dec_1950_deg` | float64 | Declination (B1950 epoch, degrees) — note: 1950 epoch, not J2000 | 0.4394444444444443 | 0.0% | | `major_axis_arcsec` | int64 | Semi-major axis of 95% confidence position ellipse (arcsec) | 28 | 0.0% | | `minor_axis_arcsec` | int64 | Semi-minor axis of 95% confidence position ellipse (arcsec) | 9 | 0.0% | | `pos_angle_deg` | int64 | Position angle of major axis east of north (degrees) | 75 | 0.0% | | `flux_12um_jy` | float64 | Flux density at 12 μm in Jansky (1 Jy = 10⁻²⁶ W/m²/Hz) | 0.1188 | 0.0% | | `flux_12um_err_pct` | int64 | Percentage uncertainty on flux_12um_jy | 21 | 0.0% | | `flux_12um_quality` | int64 | Flux quality at 12 μm: 1=high quality, 2=moderate, 3=uncertain, 0=upper limit | 1 | 0.0% | | `flux_25um_jy` | float64 | Flux density at 25 μm in Jansky | 0.1935 | 0.0% | | `flux_25um_err_pct` | int64 | Percentage uncertainty on flux_25um_jy | 22 | 0.0% | | `flux_25um_quality` | int64 | Flux quality at 25 μm: 1=high quality, 2=moderate, 3=uncertain, 0=upper limit | 1 | 0.0% | | `flux_60um_jy` | float64 | Flux density at 60 μm in Jansky — dominant band for cool dust emission | 0.2338 | 0.0% | | `flux_60um_err_pct` | int64 | Percentage uncertainty on flux_60um_jy | 22 | 0.0% | | `flux_60um_quality` | int64 | Flux quality at 60 μm: 1=high quality, 2=moderate, 3=uncertain, 0=upper limit | 3 | 0.0% | | `flux_100um_jy` | float64 | Flux density at 100 μm in Jansky — dominant band for cold ISM dust | 1.081 | 0.0% | | `flux_100um_err_pct` | int64 | Percentage uncertainty on flux_100um_jy | 21 | 0.0% | | `flux_100um_quality` | int64 | Flux quality at 100 μm: 1=high quality, 2=moderate, 3=uncertain, 0=upper limit | 1 | 0.0% | | `reliability_pct` | int64 | Source reliability percentage (0–99); ≥90 indicates high-confidence point source | 85 | 0.0% | | `snr_12um` | float64 | Signal-to-noise ratio at 12 μm | 14.0 | 33.4% | | `snr_25um` | float64 | Signal-to-noise ratio at 25 μm | 10.0 | 65.6% | | `snr_60um` | float64 | Signal-to-noise ratio at 60 μm | 6.0 | 59.1% | | `snr_100um` | float64 | Signal-to-noise ratio at 100 μm | 5.0 | 80.1% | | `num_associations` | int64 | Number of counterpart associations in the IRAS Association File | 0 | 0.0% | | `source_type` | float64 | Source type flag: 1=point source, 2=small extended, 0=no flux at any band | 1.0 | 33.4% | ## Quick stats - **173,044** mid-infrared sources across 12, 25, 60 and 100 μm bands - **57,845** high-quality 12 μm detections; **102,262** high-quality 60 μm detections - Median 60 μm flux: **0.215 Jy** — covers both galactic and extragalactic populations - Coordinates in B1950 epoch; ~75% sky coverage (galactic plane excluded) ## Usage ```python from datasets import load_dataset ds = load_dataset("juliensimon/iras-faint-source-catalog", split="train") df = ds.to_pandas() ``` ```python from datasets import load_dataset ds = load_dataset("juliensimon/iras-faint-source-catalog", split="train") df = ds.to_pandas() # Colour-colour diagram (F25/F12 vs F60/F25) to separate star-forming galaxies # from evolved stars and cirrus import matplotlib.pyplot as plt import numpy as np good = df[ (df["flux_12um_quality"] >= 1) & (df["flux_25um_quality"] >= 1) & (df["flux_60um_quality"] >= 1) & (df["flux_12um_jy"] > 0) & (df["flux_25um_jy"] > 0) & (df["flux_60um_jy"] > 0) ] log_f25_f12 = np.log10(good["flux_25um_jy"] / good["flux_12um_jy"]) log_f60_f25 = np.log10(good["flux_60um_jy"] / good["flux_25um_jy"]) plt.figure(figsize=(8, 7)) plt.hexbin(log_f25_f12, log_f60_f25, gridsize=60, cmap="YlOrRd", bins="log") plt.colorbar(label="log(count)") plt.xlabel("log(F25/F12)") plt.ylabel("log(F60/F25)") plt.title("IRAS FSC colour-colour diagram") plt.show() ``` ## Data source https://vizier.cds.unistra.fr/viz-bin/VizieR-3?-source=II/156A/main ## Related datasets - [juliensimon/neowise-asteroid-catalog](https://huggingface.co/datasets/juliensimon/neowise-asteroid-catalog) - [juliensimon/4xmm-dr14-xray-sources](https://huggingface.co/datasets/juliensimon/4xmm-dr14-xray-sources) - [juliensimon/rosat-bright-source-catalog](https://huggingface.co/datasets/juliensimon/rosat-bright-source-catalog) > If you find this dataset useful, please consider [giving it a like](https://huggingface.co/datasets/juliensimon/iras-faint-source-catalog) on Hugging Face. It helps others discover it. ## About the author Created by [Julien Simon](https://julien.org) — AI Operating Partner at Fortino Capital. Part of the [Space Datasets](https://julien.org/datasets) collection. ## Citation ```bibtex @dataset{iras_faint_source_catalog, title = {IRAS Faint Source Catalog}, author = {juliensimon}, year = {2026}, url = {https://huggingface.co/datasets/juliensimon/iras-faint-source-catalog}, publisher = {Hugging Face} } ``` ## License [NASA/IPAC IRAS Mission Data](https://irsa.ipac.caltech.edu/Missions/iras.html)