Spectroscopic observations are widely used for the identification and characterization of young stellar objects. Emission and/or absorption from several elements are attributed to the stellar photosphere and the accretion disk in these systems.
It is accepted that in low- and intermediate-mass YSOs the accretion onto the star occurs via magnetospheric accretion (e.g. Muzerolle et al. 2004). In this model, the disc is truncated at a radius, RT, where the pressure from magnetic field lines from the star balances the ram pressure of the accretion disc. The material at RT is then guided onto the star through magnetic channels (or also called accretion columns). The gas travels along the accretion columns at approximately free-fall velocities (200-300 km s-1 Muzerolle et al. 2001), and terminates at the stellar surface. This produces an accretion shock leading to the UV/blue continuum excess observed in classical T Tauri stars (see e.g. Hartmann et al. 1994). Before and after reaching the stellar surface, the gas is heated and ionized, and emission lines result from recombining and accreting hydrogen gas in these magnetic channels (Beck et al. 2010). Figure 1, shows a schematic view of the magnetospheric accretion model.

Figure 1. Schematic view of the magnetospheric accretion in young stellar objects. Figure from Hartmann et al. (2016)
In the near-infrared, the n = 7 - 4 transition of the Brackett series of hydrogen (Br γ) has been found to be strongly correlated to the mass accretion rate (Muzerolle et al. 1998) and it serves as a signpost of circumstellar accretion (Najita et al. 1996). This emission line can be used to study reddened class II objects or class I stars that are usually too deeply embedded to be observed at optical wavelengths.
The 12CO first-overtone ro-vibrational Δν = 2 bandhead absorption, spanning the wavelength range 2.293 − 2.415 μm is a common signature of the cool atmospheres of stars with spectral type later than G0 (Wallace & Hinkle 1997). The strength of the CO absorption at these wavelengths is greater in giant and supergiant stars. The presence of this signature was considered to be associated with field stars rather than YSOs in the study of molecular clouds (Elias 1978). However, YSOs were later discovered to have such absorption features and were recognized to arise from the stellar photosphere of CTTS (Casali & Matthews 1992). The lack of such features in the 2 μm spectra of younger Class I YSOs can be explained by the features being heavily veiled by an emission from the hot circumstellar dust (Connelley & Greene 2010). Figure 2 shows some examples of the spectra of Class I and Class II YSOs.

Figure 2. Typical near-IR spectra of Class I (left) and Class II (right) YSOs in NGC1333. Figure from Fiorellino et al. (2023)
Continuous monitoring of variable YSOs with the use of spectroscopy allows us to understand the physical mechanism driving the variability in these systems.
A class of variable YSOs commonly known as FUor/EX Lupi-type outbursts display unique characteristics in their spectra. These are used to confirm changes in the accretion rate as the driving mechanism.
During an FUor outburst, the spectrum is dominated by strong 12CO and H2O absorption, with a lack of emission features. At high accretion rates, the accretion disk dominates emission in the system, where absorption lines arise due to a cooler disk surface compared with the viscously-heated midplane (Liu et al. 2022). The lack of Br γ emission in the spectra of FUors has been explained as a break of magnetospheric accretion (or “crushing of the magnetosphere”). For sufficiently large accretion rates, as expected for FUors, the truncation radius can become the stellar radius, thus suppressing the accretion flows and the mechanism from which H I lines are expected to arise (see Fischer et al. 2012).

Figure 3. (left) Spectrum of EX Lupi during quiescence (top) and outburst (bottom). Spectra of V900 Mon and FU Ori showing the strong 12CO and H2O absorption characteristic of FUors in outburst (right).
The spectra of EX Lup-type objects is different from those of FUors, as they are dominated by emission lines during maximum light. In the near-IR, Na I (2.2 μm) and 12CO bandhead emission arises from the surface layers of a hot inner disc. These lines go into absorption during the quiescent state (Sipos et al. 2009). The accepted interpretation is that, at low accretion rates, the central star dominates the emission from the system, leading to the photospheric absorption profile. On the other hand, an increase in the value of the accretion rate leads to an increase of UV radiation at the accretion shock, which in turn increases the inner disc temperature and thus favors CO emission (Lorenzetti et al. 2009). In general, EX Lup type objects observed at minimum light were found to be no different from typical T Tauri stars (Herbig 2008; Lorenzetti et al. 2012)
Changes in the extinction along the line of sight can also drive high-amplitude variability in YSOs. In this case, however, we do not expect to see a change in the emission/absorption properties of the source. For example, AA Tau shows a long-term fading event (ΔV∼2 mag) that started in 2011 (Bouvier et al. 2013). Spectroscopic monitoring during different stages of the light curve (Figure 4), showed that AA Tau displayed typical emission absorption/emission features at both the bright and faint states in the light curve, with some changes in the strength of these features. The shape of the continuum, however, changes between the two epochs. This change can be attributed to an increase in extinction towards the source as well as an increase in the veiling flux (Covey et al. 2021).

Figure 4. (left) Optical/Infrared light curve of AA Tau. Spectra of AA Tau taken during the bright and faint states of the light curve. Figure from Covey et al. (2021).
Here at the Star Formation group in the Seoul National University, we have been conducting a spectroscopic follow-up programme of more than 100 YSOs that show variability at mid-IR wavelengths (discovered through NEOWISE photometric monitoring). The YSOs arise from the catalogues of Park et al. (2021) and Kuhn et al. (2021) and comprise ∼125000 objects. The observations have been carried out in partnership with research groups at the Kavli Institute for Astronomy and Astrophysics (China), the Indian Institute of Science Education and Research (IISER) Tirupati (India), the Department of Physics and Astronomy at the University of Victoria (Canada), and the Korea Astronomy and Space Science Institute (KASI, Republic of Korea).
The near-IR observations were obtained using low- and high-resolution spectrographs at the Gemini, Palomar, and IRTF observatories. The proposals have requested observations with FLAMINGOS2 and IGRINS through Gemini partnerships with the Republic of Korea and Canada. We have also accessed time on Palomar/Triplespec through the Telescope Access Program in the People’s Republic of China, and NASA IRTF/SpeX through open time offered by the USA.
In Contreras Pena et al. (2023) we published the first work resulting from the spectroscopic monitoring of YSOs. In here we report the classification of nine nearby (d<1.2 kpc) objects that show the characteristics of known classes of eruptive variable YSOs. We find one FU Orionis (FUor)-like source, one EX Lupi type object, and six YSOs with mixed characteristics or V1647 Ori like objects. The varied characteristics observed in our sample are consistent with recent discoveries of eruptive YSOs. We discuss how a wide range in YSO outburst parameters (central mass, maximum accretion rate during outburst, evolutionary stage, and/or instability leading to the outburst) may play a significant role in the observed spectrophotometric properties of YSO outbursts. Figure 5 shows the light curves and spectra of two YSOs that are discussed in Contreras Pena et al. (2023).

Figure 5. Light curves (top), W1 vs W1-W2 colour-magnitude diagram (middle) and near-IR spectra of YSOs LDN 1455 IRS3 (left) and HOPS 267 (right). Figure from Contreras Pena et al. (2023)
Several other publications, using the spectroscopic data gathered by our group, are currently under preparation, and planning to be submitted within this year (2023).