Characterisation of an EXor outburst SPICY 97589

Stellar outbursts from variable or periodic accretion are thought to be ubiquitous across young stellar populations. However, relatively few outbursting objects have been discovered to date. Here, we present the characterisation of a new EXor-type episodic accretor. We aim to characterise the nature of the 2023 outburst of SPICY 97589/Gaia23bab and characterise the stellar source for the first time, while exploring how an accretion outburst contributes to disk evolution. We employ multi-waveband medium-resolution spectroscopy with UVB-VIS-NIR coverage during the peak of the 2023 outburst and the post-outburst quiescent object. The broad wavelength coverage of the dataset allows for robust measurements of the accretion rate using known line tracers. The addition of quiescent spectra provides a good estimation of stellar parameters of the central star while also informing us on the evolution of the disk during outburst phases. We find the stellar source to be a 3410,K, M3.0 type star with a luminosity of 0.41 $L_\odot$ and an estimated stellar mass of 0.29 $M_\odot$. We measure the accretion rate of SPICY 97589 to be $\dot M = 2.38\pm0.58\times10^{-7},\mathrm{M_\odot yr^{-1}}$. This value is at two orders of magnitude greater than the quiescent accretion rate. Thus, we confirm that the 2023 outburst was driven by an influx of material from the surrounding environment to the central star, an accretion outburst. The spectral fingerprint of emission lines is also characteristic of an outbursting EXor-type source, including variable disk winds.

💡 Research Summary

This paper presents a comprehensive multi‑wavelength spectroscopic and photometric study of the young stellar object SPICY 97589 (also known as Gaia23bab), which underwent a pronounced EXor‑type outburst in early 2023. Using the ESO X‑Shooter instrument, the authors obtained medium‑resolution spectra covering the UVB (300–560 nm), VIS (560–1020 nm), and NIR (1020–2400 nm) ranges at two epochs: 2023‑04‑19, roughly two months after the outburst peak when the source was still ≈2 mag brighter than quiescence, and 2024‑07‑25 after the source had returned to its quiescent state. The observations employed narrow slits (0.8″ UVB, 0.7″ VIS, 0.6″ NIR) to achieve resolutions of R≈6700, 11400, and 8100 respectively, and were reduced with the standard X‑Shooter pipeline and telluric‑corrected with Molecfit, ensuring consistent absolute flux calibration across both epochs.

Long‑term photometric monitoring from Gaia (optical G‑band) and WISE (mid‑infrared W1/W2) reveals two distinct outbursts, one in 2017 and a stronger, longer event in 2023. The 2023 outburst peaked at +2.51 mag above the quiescent G‑band level, with a full‑width at half‑maximum of ≈237 days and a total duration of ≈671 days, exceeding the 2017 event (peak +2.24 mag, duration ≈358 days). In the mid‑infrared, the brightening is more modest (≈1.2 mag in W1 and ≈1.3 mag in W2), indicating that the bulk of the luminosity increase originates from hotter regions closer to the star rather than the cooler outer disk.

To determine the stellar parameters, the authors applied the FRAPPE fitting tool to the quiescent spectrum, combining an accretion‑slab model, a Class III photospheric template, and an extinction law (Cardelli et al. 1989, RV=3.1). The best‑fit solution yields a spectral type M3.0 (Teff = 3410 K), visual extinction AV = 4.4 mag, stellar luminosity L* = 0.41 ± 0.09 L⊙, and, using Baraffe et al. (2015) isochrones, a mass M* = 0.29 ± 0.05 M⊙ and an age of ≈0.75 Myr. These values differ markedly from earlier estimates based on spectral energy distribution fitting (Giannini et al. 2024), underscoring the importance of direct spectroscopic constraints.

Accretion rates were derived from 22 emission lines (hydrogen Balmer, Paschen, and Brackett series, as well as He I and Ca II). Equivalent widths were measured on de‑reddened, continuum‑subtracted spectra, and line fluxes were computed by multiplying EW by the local continuum flux. Using empirical relations log Lacc = A + B log Lline from Alcalá et al. (2017) and Fairlamb et al. (2017), the authors converted line luminosities to accretion luminosities. For the 2023 outburst spectrum, the mean accretion luminosity corresponds to a mass accretion rate Ṁ = 2.38 ± 0.58 × 10⁻⁷ M⊙ yr⁻¹. This is roughly two orders of magnitude higher than the quiescent rate (≈10⁻⁹ M⊙ yr⁻¹) inferred from the same line diagnostics, confirming that the outburst was driven by a sudden surge of material onto the star. The Balmer lines exhibit pronounced P‑Cygni profiles, indicative of strong, variable disk winds, while higher‑order Brackett lines appear only during outburst, reflecting the presence of dense, hot gas close to the stellar surface.

The paper situates its findings within the broader context of episodic accretion in young stellar objects. It confirms that EXor events, though shorter and less luminous than FUor eruptions, can still increase the accretion rate by factors of 100–1000, potentially influencing disk chemistry, dust processing, and the early stages of planet formation. By comparing their accretion rate estimates with those of previous works (Giannini et al. 2024; Nagy et al. 2025), the authors demonstrate consistency across independent datasets and methodologies.

In conclusion, the study provides a robust, multi‑epoch, multi‑wavelength characterization of an EXor outburst, delivering precise stellar parameters, a detailed accretion rate measurement, and evidence for wind‑driven mass loss. The results highlight the value of simultaneous UV‑optical‑NIR spectroscopy for disentangling the contributions of stellar photospheres, accretion shocks, and circumstellar disks during rapid accretion events. Future high‑resolution spectroscopic monitoring and interferometric imaging will be essential to map the geometry of the inflowing and outflowing material and to assess the long‑term impact of such episodic accretion on protoplanetary disk evolution.