AT 2018cow: Prototype FBOT
- AT 2018cow is a fast blue optical transient exhibiting extreme luminosity, rapid evolution, and diverse wavelength emissions.
- It demonstrates engine-driven explosion mechanisms through aspherical circumstellar interaction and evidence of a compact central object.
- Multiwavelength observations including X-ray QPOs and radio shock measurements critically constrain models ranging from magnetars to tidal disruption events.
AT 2018cow (“The Cow”) is the nearest and best-studied example of the Fast Blue Optical Transients (FBOTs), a class of extragalactic explosions distinguished by extreme luminosity, rapid photometric evolution, blue continua, and unusual multi-wavelength signatures. Its 2018 eruption in CGCG 137-068 (z=0.0141, d≈60 Mpc) triggered a global campaign spanning X-ray, UV, optical, mm, cm, and radio bands, establishing it as a prototype for both the FBOT phenomenon and a new regime of aspherical, engine-driven transients.
1. Photometric and Spectroscopic Evolution
AT2018cow displayed a rest-frame bolometric luminosity erg s⁻¹ (), with an unprecedentedly rapid rise to peak (t_rise ≲ 2.9 d) and post-peak decline rates of 0.2–0.4 mag d⁻¹ in the first ten days (Xiang et al., 2021, Margutti et al., 2018). The UV–optical SED maintained a high temperature (– K) and exhibited a receding blackbody photosphere, with dropping from cm at 2–3 d to < 10¹⁴ cm by 1 month (Perley et al., 2018, Chen et al., 2023). Early spectroscopy revealed a nearly featureless, hot continuum. Broad (FWHM ≳ 10,000 km s⁻¹) absorption features appeared after ∼3–8 d (v ∼ 0.3 c), then vanished; broad and intermediate-width (3000–14,000 km s⁻¹) He I, H I, and high-excitation lines emerged after ∼10 d, often redshifted by several thousand km s⁻¹ (Fox et al., 2019, Xiang et al., 2021). Weak, narrow He I features (v ∼ 700–1000 km s⁻¹) developed after 20 d, closely resembling those of SNe Ibn/IIn.
The late-time Hubble Space Telescope SEDs (50–60 d) remained smooth blackbodies, K, ; bolometric light declines followed (up to day 13), steepening to 0 thereafter (Chen et al., 2023).
2. Multiwavelength Emission and Physical Modeling
X-ray and γ-ray
AT2018cow was X-ray luminous, with 1(0.3–10 keV) peaking at 2 erg s⁻¹ in 3–10 d and persistent hard X-ray emission up to 3 keV (Margutti et al., 2018). The X-ray decay transitioned from 4 to 5 (post-20 d), with erratic 6days variability and spectral softening (Sandoval et al., 2018). At 7 keV, a spectral hump appeared during the first 815 d, vanishing as the Thomson optical depth dropped to unity, consistent with Compton-downscattering in expanding ejecta (Margutti et al., 2018, Govreen-Segal et al., 26 Jan 2026).
A high-amplitude quasi-periodic oscillation (QPO) at 224.4 Hz (9 cycles, Q ≳ 14, RMS ∼ 30%) was detected over the initial 60 d, providing direct evidence for a compact object—either a neutron star (0 ms, 1 G) or a low-mass black hole (2) (Pasham et al., 2021).
Radio to mm
cm – mm light curves exhibited spectral peaks (SSA turnovers) shifting from 3 GHz at 1–2 weeks to 41 GHz at hundreds of days, associated with a shock of 5–6 expanding into a dense medium (7–8 cm⁻³) (Ho et al., 2018, Nayana et al., 2021). The radio energy tied to the forward shock was 9 erg, and VLBI imaging constrained source expansion to 0 at 98 d, excluding any long-lived relativistic jet (Bietenholz et al., 2019, Mohan et al., 2019).
Polarization and Geometry
High-cadence polarimetry (RINGO3) recorded a brief (≲1 d) 7% optical polarization spike at 5.7 d, declining rapidly (Maund et al., 2023). These values exceed the spheroidal scattering limit and require an aspherical (disk-like) CSM with 1 viewed near-edge-on, suggesting an equatorially concentrated dense shell or disk intersected by the shock (Maund et al., 2023).
3. Progenitor, Circumstellar Medium, and Explosion Environment
Integral-field spectroscopy and resolved HI mapping of host CGCG 137-068 show AT2018cow occurred in a region of young stars (~10 Myr), moderately sub-solar metallicity (12+log(O/H)≃8.6), and slightly elevated SFR density, within—though not exactly coincident with—a 2 kpc HI ring (Lyman et al., 2020, Roychowdhury et al., 2019). The immediate environment lacks an unusual atomic gas concentration or a distinct star cluster, but displays localized features compatible with both bar/accretion-induced ring formation and past galaxy interaction (Michałowski et al., 2019, Roychowdhury et al., 2019). HI fraction and kinematics are not exceptional, placing CGCG 137-068 at the lower edge of the dwarf main sequence. The absence of evidence for a compact star cluster or kinematic substructure disfavors a local IMBH host.
The pre-explosion CSM features a composite structure: an inner, dense equatorial shell (2 cm, 3–4), likely produced by eruptive mass loss up to ~2 y before explosion, and a more diffuse, extended wind (5; 19–45 y pre-explosion) (Nayana et al., 2021, Fox et al., 2019). These are consistent with either a massive star (possibly WR) experiencing rapid pre-SN ejection or a binary/CE event.
4. Competing Theoretical Interpretations
AT2018cow’s phenomenology has motivated several distinct models:
| Model Type | Key Elements | Main Constraints/Challenges |
|---|---|---|
| Shock in aspherical CSM | Shock propagates through dense (disk-like) CSM | Reproduces coordinated optical/X-ray, X-ray hump and instabilities; requires 6–7 erg, 8–0.05 M9, 0 M1 (Govreen-Segal et al., 26 Jan 2026) |
| Magnetar central engine | Millisecond-P NS (P₀~3.7ms, 2 G) | Simultaneous fit to UV–X-ray with 3 M4, 5; struggles with late UV plateau (Li et al., 2024) |
| Tidal-disruption event (TDE) | Disruption of low-mass star by IMBH (6–7 M8) | Late-time UV “plateau,” tiny (9 cm) blackbody, and slow decay at 2–5 yr closely match disk emission; super-Eddington requirement is not explained theoretically (Inkenhaag et al., 9 Oct 2025) |
| Luminous, interacting SN Ibn/IIn | Compact, stripped (possible WR) progenitor with CSM | Early blue continuum and emission-line structure are reproduced, but rapid decline and X-ray properties not naturally modeled (Xiang et al., 2021) |
Central radioactive decay models are excluded by the minimal 0Ni mass derived from light curves and lack of UV line blanketing in SEDs (Chen et al., 2023, Margutti et al., 2018). Classical, long-lived relativistic jets are excluded by strict VLBI size and expansion limits (Bietenholz et al., 2019, Mohan et al., 2019).
5. Late-time Evolution (1–5 Years) and Central Source
HST imaging at 714–2043 d post-explosion revealed a persistent, luminous, blue (1) source with minimal (2 mag) fading in both optical and UV bands (Sun et al., 2022, Inkenhaag et al., 9 Oct 2025). The inferred blackbody temperature exceeds 3K and 4, with 5, orders of magnitude smaller than the UV/optical photospheric radii of CCSNe with CSM interaction at comparable epochs (Inkenhaag et al., 2024). No normal stellar, echo, or standard CSM-interaction scenario matches the suite of late-time observations; either a massive, ultra-young star cluster or prolonged central-engine (magnetar/TDE accretion) emission is implied (Sun et al., 2022).
Direct comparison to 51 nearby core-collapse SNe with HST UV at 2–5 yr shows AT2018cow to be notably more UV-luminous and to fade much slower than any detected SN, with the compact photospheric radius difficult to reconcile with interaction models (Inkenhaag et al., 2024). UV flux at 65 yr matches disk TDE model predictions (smooth 7 or plateau) but decays much slower than CSM-interacting supernovae (Inkenhaag et al., 9 Oct 2025).
6. Synthesis: Progenitor, Explosion, and Broader Context
The environment, quasi-stripped ejecta, dense compact CSM, and multiwavelength energetics are compatible with advanced core-collapse of a moderately massive star (M8–25 M9) experiencing sudden, asymmetric mass loss (possibly via binary interaction or violent pulsational ejection) (Lyman et al., 2020). Magnetar or black-hole accretion central engines fit the early rapid optical/X-ray decay, high velocities, and QPO, but require non-standard late-time energy input to explain the UV/optical plateau. Conversely, disk TDE scenarios by IMBHs replicate the late-time UV and 0 behaviors, but need to invoke highly super-Eddington emission and account for the lack of local IMBH host evidence.
The consensus is that AT2018cow and similar LFBOTs are powered by a central compact object—magnetar, low-mass black hole, or IMBH—embedded within a unique, likely aspherical CSM environment, with observed diversity set by differences in the angular structure of both the progenitor mass loss and the explosion itself (Margutti et al., 2018, Govreen-Segal et al., 26 Jan 2026).
7. Open Questions and Implications
AT2018cow remains a nexus for the study of central-engine astrophysics, non-spherical CSM interaction, and the end states of intermediate-mass stars. Its late-time UV–optical emission is a stringent discriminator for engine vs. interaction models, with ongoing HST monitoring expected to conclusively rule in or out TDE scenarios by ∼8 yr post-explosion (Inkenhaag et al., 9 Oct 2025). The event’s asphericity, rapid coupling between X-ray and optical decay, and radio/millimeter signals define the prototypical observational hallmarks of FBOTs and establish AT2018cow as the reference point for next-generation time-domain surveys and multimessenger engine-driven studies.