Gaia 20eae: A newly discovered episodically accreting young star

The Gaia Alert System issued an alert on 2020 August 28, on Gaia 20eae when its light curve showed a $\sim$4.25 magnitude outburst. We present multi-wavelength photometric and spectroscopic follow-up observations of this source since 2020 August and identify it as the newest member of the FUor/EXor family of sources. We find that the present brightening of Gaia 20eae is not due to the dust clearing event but due to an intrinsic change in the spectral energy distribution. The light curve of Gaia 20eae shows a transition stage during which most of its brightness ($\sim$3.4 mag) has occurred at a short timescale of 34 days with a rise-rate of 3 mag/month. Gaia 20eae has now started to decay at a rate of 0.3 mag/month. We have detected a strong P Cygni profile in H$\alpha$ which indicates the presence of winds originating from regions close to the accretion. We find signatures of very strong and turbulent outflow and accretion in Gaia 20eae during this outburst phase. We have also detected a red-shifted absorption component in all the Ca II IR triplet lines consistent with signature of hot in-falling gas in the magnetospheric accretion funnel. This enables us to constrain the viewing angle with respect to the accretion funnel. Our investigation of Gaia 20eae points towards magnetospheric accretion being the phenomenon for the current outburst.


INTRODUCTION
Episodic accretion onto low-mass pre-main sequence (PMS) stars is no longer considered an oddity. It is now considered as one of the important stages in the grand scheme of evolution of the low-mass PMS stars, even though it is a poorly understood phenomenon. The arXiv:2112.01717v1 [astro-ph.SR] 3 Dec 2021 short outburst timescales compared to the millions of years spent in the formation stage of these PMS stars makes these events extremely rare, although statistically each PMS star is expected to experience ∼50 such short duration outbursts during its formation stages (Scholz et al. 2013). The outburst durations, although short in timescales, are capable of delivering a substantial fraction of circumstellar mass onto the central PMS star (Vorobyov & Basu 2006). These events have been observed to span the entire age range of young stars starting from the embedded Class 0 sources to the Class ii sources (Safron et al. 2015). Based on the outburst timescales and spectroscopic features, these classes of sources have been classically divided into two categories: FUors which experience a luminosity outburst of 4-5 mag that last for several decades and containing only absorption lines in their spectra and EXors experiencing a luminosity outburst of 2-3 mag which last for a timescale of few months to few years and contains emission lines in their spectra (Herbig 1977;Hartmann & Kenyon 1996;. The physical origin of the sudden enhancement of accretion rate is not yet clear. However, a variety of models ranging from thermal instability, magneto-rotational instability, combination of magneto-rotational instability and gravitational instability, disc fragmentation to external perturbations have been proposed (Audard et al. 2014). To arrive at a general consensus about the physics behind such sudden enhancement of accretion rates, a large sample of FUor/EXor sources are required to test the above instability models. However, only about 25 FUor/EXor sources have been discovered so far (Audard et al. 2014). Therefore, any newly discovered source provides an important test-bed to probe the various physical aspects of episodic accretion and their comparison with the previous sources.
The Gaia Photometric Alert System (Wyrzykowski et al. 2012;Hodgkin et al. 2013) is dedicated towards issuing transient alerts. Previously, it has issued three proven alerts for the eruptive young stars: Gaia 17bpi (Hillenbrand et al. 2018), Gaia 19ajj (Hillenbrand et al. 2019) and Gaia 18dvy (Szegedi-Elek et al. 2020). Among these, Gaia 17bpi and Gaia 18dvy are classified as FUors while Gaia 19ajj has been classified as an EXor, with its spectral features similar to that of V2492 Cyg (Hillenbrand et al. 2019). The Gaia alert system issued an notification on 2020 August 28 about Gaia 20eae with a transient identification number of AT2020nrs stating that it has undergone a 4.6 magnitude outburst. The rise timescales and the amplitude of the outburst suggest that this should be an FUor/EXor phenomenon. We have carried out optical and near-infrared (NIR) pho-tometric and spectroscopic observations and combined them with the archival optical and infrared (IR) data to identify the outburst features of Gaia 20eae. In this paper, we present the initial findings of this source. Section 2 provides details on the location and distance of Gaia 20eae. Section 3 describes about the observations and the data reduction procedures in detail. In Section 4 we describe the results that have been obtained while in Section 5 we conclude by our understanding of the present outburst in the context of the FUor/EXor phenomena. The W51 star-forming region is known to be one of the most massive and active star-forming sites of our Galaxy located at about 5 kpc distance from us (Ginsburg 2017). Retes-Romero et al. (2017) have listed two molecular cloud MC1 (size∼ 15 × 15 ) and MC2 (size∼ 30 × 36 ) in this direction at two different distances of 1.3 ± 0.2 kpc and 3.4 ± 0.4 kpc, respectively. The distance of these molecular clouds are derived kinematically using the CS (2 → 1) line velocities (Faúndez et al. 2004). The MC1 and MC2 are also associated with IRAS sources IRAS 19230+1506 and IRAS 19236+1456, respectively (Retes-Romero et al. 2017. These molecular clouds do not have any optical nebula associated with them, indicating that the star-formation activity has started recently. Also, the IRAS sources have ultra-compact H ii region (UCHii) colors indicating that these molecular clouds are high mass star-forming regions. In Figure  1, we show the location of Gaia 20eae along with the IRAS sources in the IR color-composite image generated from WISE 22 µm (red), WISE 12 µm (green), and 2MASS 2.2 µm (blue) images. Heated dust grains (22 µm emission) can be seen at several places including at the IRAS locations of sources. The warm dust towards the south of Gaia 20eae is surrounded by 12 µm emission which covers the prominent Polycyclic Aromatic Hydrocarbons (PAH) features at 11.3 µm, indicative of photon dominant region (PDR) under the influence of feedback from massive stars (see e.g. Peeters et al. 2004). This indicates that the Gaia 20eae is located at a site showing signatures of recent star-formation activities. Until the release of data from the Gaia mission 1 , there was no direct measurement of distance of the Gaia 20eae. Recently, adding corrections to the Gaia data release Figure 1. Color-composite image obtained by using the WISE 22 µm (red), WISE 12 µm (green) and 2MASS 2.2 µm (blue) images of the ∼ 30 × 30 Field of View (FOV) around Gaia 20eae. The locations of Gaia 20eae, IRAS 19230+1506 and IRAS 19236+1456 (Retes-Romero et al. 2017) are shown by cyan circles. Locations of standard stars from the ZTF sky survey is also shown with green circles. Sub-panels 1(a) and 1(b) show the pre-outburst and post-outburst phases of Gaia 20eae in optical color composite image taken from SDSS and HCT, respectively. 3 (DR3) parallax by using the Bayesian inference approach to account for the non-linearity of the transformation and the asymmetry of the resulting probability distribution, Bailer-Jones et al. (2021) estimated the distances of stars in our Galaxy. Therefore, for Gaia 20eae, we have adopted a distance of 3.2 ± 1 kpc as estimated by Bailer-Jones et al. (2021). Since, molecular cloud MC2 is located in the same direction at a distance of 3.4 ± 0.4 kpc (Retes-Romero et al. 2017), Gaia 20eae's seems to be associated with MC2. This makes it the farthest discovered FUor/EXor type source till date. Almost all the previously discovered FUors/EXors are located at a distance of ∼1 kpc or less. We have monitored Gaia 20eae photometrically in optical bands at 16 different epochs with the Himalayan Faint Optical Spectrograph Camera 2 (HFOSC, 4 epochs) on 2m Himalayan Chandra Telescope (HCT), Hanle, India, ANDOR 2K CCD (9 epochs) on 1.3m Devasthal Fast Optical Telescope (DFOT), Nainital India, and FlareCam 1K CCD 3 on 0.5m ARC Small Aperture Telescope (ARCSAT, 3 epochs), New Mexico 4 from 2020 August to December. We have also obtained the near infrared (NIR) photometric data of Gaia 20eae during its outburst state using the TIFR-ARIES Near Infrared Spectrometer (TANSPEC; Ojha et al. 2018) mounted on the 3.6m Devasthal Optical Telescope (DOT), Nainital, India on the night of 2020 October 24. Table 1 provides the complete log of photometric observations presented in this work.
We have used standard data reduction procedures for the image cleaning, photometry, and astrometry (for details, see Sharma et al. 2020). We have derived following color transformation equations using the available magnitudes in different filters (i.e., APASS DR10 archive 5 or Two Micron All Sky Survey (2MASS) archive 6 ) of all the stars in the frame at epoch JD = 2459141 (for optical) and JD = 2459144 (for NIR).
In order to calibrate the photometry of Gaia 20eae at other epochs, we have used these equations with intercept estimated from a set of 7 non-variable standard stars (see Table 3). These non-variable standard stars were identified from the Zwicky Transient Facility (ZTF) sky survey (Bellm et al. 2018) based on their zr band light curves (LCs 7 ). Table 2 lists the magnitudes of Gaia 20eae in different filters in different epochs of our observations.

Archival data
We have also obtained the photometric data from the time domain Gaia sky survey (Gaia Collaboration et al. 2016. Gaia sky survey maps the sky in G band to look out for the transients and regularly updates on their Gaia Alert Index website 8 . We have obtained the G band photometric data provided by the Gaia survey at its data archive 9 . We have also acquired the pre-outburst g, r, i, z band archival data from the Panoramic Survey Telescope and Rapid Response System (Pan-STARRS; PS1). The details about the PS1 surveys and latest data products are given in Chambers et al. (2016). We have downloaded the point source catalog from the data release 2 of the PS1 10 . Gaia 20eae was observed by the Zwicky ZTF sky survey (Bellm et al. 2018). We obtained the archival zr band photometric data of ZTF available from the 10 http://catalogs.mast.stsci.edu/ NASA/IPAC Infrared Science Archive 11 . We also obtained the recent photometric data in zg and zr bands of the observation made by the ZTF survey from Lasair 2.0 12 , a community broker service to access, visualize and extract science data. We obtained the pre-outburst mid-infrared (MIR) magnitudes of Gaia 20eae from the Spitzer archive 13 in 3.6 µm, 4.5 µm, 5.8 µm, 8.0 µm and 24.0 µm wave bands. Pre-outburst magnitudes of Gaia 20eae are also obtained from the W ISE archive 14 in 3.4 µm, 4.6 µm, 12 µm and 22 µm wave bands. Outburst magnitudes of Gaia 20eae are obtained from the W ISE/N EOW ISE survey 15 in 3.4 µm and 4.6 µm wave bands.

Spectroscopic data 3.2.1. Medium resolution Optical/NIR Spectroscopy
A photometric alert was issued by the Gaia alert system named AT2020nrs on 2020 August 28, 1:28 p.m. UTC. We immediately followed it using a medium resolution (R∼2000) spectrograph 'HFOSC' mounted on the 2m HCT starting from 2020 August 29 itself. Using the Gr7 and Gr8 grisms of HFOSC, our spectroscopic observations spanned the optical wavelength range from ∼ 4000Å to 9000Å. We have also observed the flux calibrator 'Feige 110' on each night after Gaia 20eae to flux calibrate our HFOSC spectra. On 2020 September 12, we also obtained a medium resolution optical spectrum (R∼1140, 1760 and 1920 for the Orange Arm, Red Arm and Far Red Arm respectively) using the LRS-2 Red Integral Field Unit spectrograph on 10-m Hobby-Eberly Telescope (HET) (Ramsey et al. 1998;Shetrone et al. 2007), USA. LRS2-R spectrum was reduced using standard LRS2 pipeline, Panacea 16 . Finally, we scaled our flux calibrated spectra to match the flux and slope obtained from the photometric flux values of the same date. In case, photometric magnitudes are not available on the same date we have scaled our spectra with the photometric flux values of the nearest date. This is done to correct for any residual systematics in the flux calibration for the HFOSC due to its sensitivity to seeing variations and centering errors on the slit. This is also important for the LRS2 IFU observation, since the night was hazy due to smoke from wildfires, resulting in a highly variable non-grey atmospheric extinction.
We obtained NIR spectra of Gaia 20eae using the TANSPEC with its 0 .5 slit providing a R∼2700 on the nights of 2020 October 24 and 2020 November 6. Standard NIR dithering technique i.e, obtaining the spectra at two different slit positions, was followed. The final spectrum of the object is obtained by subtracting the spectra obtained at the dithered positions to cancel the sky contribution. A telluric standard star was also observed immediately after the Gaia 20eae observations to remove the telluric features.
We have used the standard tasks of IRAF 17 to reduce medium resolution spectroscopic data. The task apall was used to extract the one dimensional spectrum. The extracted spectrum was then calibrated using identify task with the help of the calibration lamps taken immediately after the source spectrum. Finally, continuum task of IRAF is used to continuum normalize the spectra in order to measure the equivalent widths (EWs) of different lines. Standard IRAF tasks standard, sensfunc and calibrate were used to flux calibrate our spectra.

High Resolution Near Infrared Spectroscopy
We observed a high resolution NIR spectrum of Gaia 20eae on 2020 September 12 using the Habitable Zone Planet Finder (HPF) (Mahadevan et al. 2012;Mahadevan et al. 2014) on the 10m HET. HPF covers the wavelength range of 8100 -12800Å, at a spectral resolution of R∼55,000. The H2RG up-the-ramp raw data cube was reduced to 1D spectra by the procedures described in Ninan et al. (2018); Kaplan et al. (2018); Stefansson et al. (2020). The wavelength calibration was done using a laser frequency comb calibrator as described in Metcalf et al. (2019). Barycentric correction was applied to all spectra with the values calculated using barycorrpy (Kanodia & Wright 2018).
In summary, we have monitored Gaia 20eae spectroscopically at 10 different epochs with the HFOSC (6), TANSPEC (2), LRS2-R (1) and HPF (1).  Gaia 20eae is named as 'SSTGLMC G050.2584-00.5077' and classified as a candidate young stellar source (YSO) due to its red color in MIR bands using the Spitzer photometry by Robitaille et al. (2008). Later on, this source was classified as Class ii YSO based on its IR spectral index (MC1-M15 in the Table 4 of Retes-Romero et al. 2017). Retes-Romero et al. (2017) also derived its mass as 1.5 M assuming an age of 2 Myr for a typical Class ii source and a distance of 1.3 kpc. As Bailer-Jones et al. (2021) have estimated a distance of 3.2 kpc for Gaia 20eae from Gaia DR3, this would result in a different mass estimation. However, in the absence of direct measurement of A V around this source, it would be very difficult to derive its accurate physical parameters (e.g., age/mass).

Light Curve
The upper panel of Figure 2 shows the Light Curve (LC) of Gaia 20eae in Gaia G, ZTF zg and zr, Johnson − Cousins B, V , R and I and SDSS g, r, i and z bands. It is worthwhile to mention here that, although the ZTF (Bellm et al. 2018) and SDSS filters cover similar wavelengths, they have differences in their cutoff wavelength and transmission curve. The Gaia G band data cover the longest time span of the LC starting from 2014 October 27 (JD=2456957) upto 2020 December 9 (JD=2459192). The ZTF zr band has data from 2018 April 9 (JD=2458218) to 2020 November 27 (JD=2459180), whereas zg band data are available from 2020 May 16 (JD=2458985) to 2020 November 25 (JD=2459178). The ZTF photometric data have better temporal sampling (2-3 days) as compared to the Gaia G band data (10-15 days). The LC clearly demonstrates a long quiescent period with minor fluctuations until 2019 October 28 (JD=2458785), after that, it began to transit to the present outburst stage. The pre-outburst magnitudes of Gaia 20eae were G ∼ 19.49 mag (2019 November 30; JD=2458818) and zr ∼ 19.40 mag (on 2019 October 28; JD=2458785). The mean G and zr magnitudes of Gaia 20eae during the quiescent phase were 19.14 ± 0.26 mag (from 2014 October 27 to 2019 November 30) and 19.46 ± 0.17 mag (from 2018 April 9 to 2019 October 28), respectively.
As the quiescent phase LC of Gaia 20eae shows small scale fluctuations, we searched for periodic variability in it using the ZTF zr band data. Periodic variability has been reported in the PMS stars which is due to the rotation of the star having hot and cool spots on their photosphere. We have used the Period 18 software, which works upon the principle of Lomb-Scargle (LS) periodogram (Lomb 1976;Scargle 1982), to determine the period of Gaia 20eae and to phase fold the LC. The advantage of the LS method is that it is effective even in case of the data set being non-uniformly sampled. We have also used the NASA Exoplanet Archive Periodogram 19 service for cross verification. The periods obtained in both the cases matched well. The period of Gaia 20eae thus comes out to be 2.1±0.004 days. The period detected in the quiescent LC of the Gaia 20eae might correspond to the rotational period of the star. This type of period is commonly observed in Class ii/iii type of YSOs as shown by Sinha et al. (2020). Figure  3 shows the phase folded LC of the Gaia 20eae during its quiescent phase. The amplitude of variation is of the order of 0.2 mag which is also typical of Class ii/iii type of YSOs (Sinha et al. 2020 In the lower panel of Figure 2, we show the LC of Gaia 20eae in B, V , R, I, G, zg, zr, g, r, i, z bands in the outburst phase. The LC starts from 2020 March 3 (JD = 2458800) upto the latest data point on 2020 December 9 (JD = 2459192). The LC of Gaia 20eae is peculiar in the sense that the rise to peak brightness consists of two parts: an initial slow rise from the quiescent phase starting from JD = 2458800 to JD = 2458995 and then a rapid rise to the peak brightness from JD = 2459014 to JD = 2459047, reaching maximum brightness on JD = 2459047, and then a slowly decaying phase (JD = 2459047 to JD = 2459145). It also shows small scale  Quiescent phase folded LC of Gaia 20eae as obtained from the ZTF zr band data. The period is determined by using the Period software and also cross-matched with the NASA Exoplanet Archive Periodogram service.
fluctuations with amplitude of ∼0.2 mag on a time scale of few days. We were not able to derive the periodicity of these fluctuations using the LS periodogram. We call the rapid rise part and slow decaying part of the LC as transition phase and active plateau phase, respectively, and the same have been labeled and shadowed with different colors in the lower panel of Figure 2. We have calculated the rise-rate and decay-rate of Gaia 20eae in different wavelengths from the available Gaia and ZTF data by fitting a least square straight line in the different phases of the LC. The fits for the data points in the transition and active plateau phases are shown in the lower panel of Figure 2. The overall best-fit rise-rate from the quiescent phase to the maximum brightness is calculated to be 0.6 mag/month in the G band, over a duration of ∼247 days. This rise rate is prone to higher uncertainties as there are lots of data gaps in the LC initially. The rise-rate in transition phase was found to be similar i.e. ∼0.1 mag day −1 or ∼3 mag month −1 in G, zg and zr bands. The decay rate in the active plateau phase is calculated as 0.01 mag day −1 (or 0.3 mag month −1 ) for a duration of ∼98 days, which is an order less than the rise-rate. It is to be noted here that Gaia 20eae has not returned to its quiescent state. Hence, the decay rate that we calculated is by considering the data range that we have presented in this study.
The maximum brightness in the current outburst phase in the zg and zr bands is recorded on 2020 July 16 (JD=2459047) as 17.02 mag and 15.09 mag, respectively. In the G band, the source reached the maximum brightness of 14.89 magnitude on 2020 August 26 (JD = 2459088). Thus, the current outburst magnitude amplitudes are: ∆G = 4.25 mag and ∆zr = 4.37 mag.  Days The LCs in B, V, R and I also follow the trend of the ZTF and Gaia LCs. The quiescent phase J, H, and K s magnitudes of Gaia 20eae are 14.67 mag, 13.36 mag and 12.16 mag, respectively as obtained from the UKIDSS DR10plus. During the present outburst stage, the J, H, and K s magnitudes of Gaia 20eae as obtained from TANSPEC are 12.41 mag, 11.27 mag and 10.40 mag, respectively. This implies that the present outburst is similar to the FUor family of objects and is almost wavelength independent. Figure 4 left panel represents the Spectral Energy Distributions (SEDs) of Gaia 20eae during its quiescent and active states represented in red and cyan curves, respectively. We constructed the quiescent phase SED using the multiwavelength data (optical to MIR wavelengths, i.e. 0.44 (B), 0.55 (V ), 0.65 (R), 0.80(I), 1.2 (J), 1.6 (H), 2.2 (K s ), 3.4(W 1), 3.6(I1), 4.6(W 2), 4.5(I2), 5.8(I3), 8.0(I4), 12(W 3), 22(W 4) and 24(I4) µm) taken from the data archives (PS1, 2MASS, Spitzer and W ISE). The PS1 magnitudes were transformed to Johnson Cousins system by using the equations given by Tonry et al. (2012). For the outburst state SED, we have used current optical and NIR band observations as well as NEOWISE data. Apart from having a shift in the brightness, there is clearly a change in the shape of the SED at the longer wavelengths as compared to the shorter wavelengths. This change in the SED can be quantified by comparing the differences in observed magnitudes of Gaia 20eae with that of pure dust clearing events. Right panel of Figure 4 shows the observed magnitude differences of Gaia 20eae between its quiescent and active state by a blue curve. We compare the magnitude variations against ∆A z = −2 mag and ∆A z = −4 mag for R V = 3.1 and R V = 5.5 dust laws (see also, McGehee et al. 2004). From the deviation of the observed magnitude difference with the dust clearing events, we can conclude that the brightening of Gaia 20eae cannot be explained by the diminishing of the line-of-sight extinction, rather the enhancement of accretion rate is the likely cause.

The evolution of the photometric Spectral Energy Distribution
Using the multi-epoch data from the PS1 archive (quiescent phase), 2m HCT and 1.3m DFOT (outburst phase), we have also examined the evolution of the reddening invariant colors, 'Q xyz ' of Gaia 20eae as it transitioned from quiescent state to the eruptive state. The reddening invariant colors have a generic form of : , where x, y, and z are the observed magnitudes in each passband (McGehee et al. 2004). A color change having ∆ Q xyz = 0 indicates that the changes in SED is not due to pure dust clearing. The estimated reddening-invariant colors (for R V = 3.1) for Gaia 20eae as it transitioned from the quiescent state to the active state are listed in Table 4. The large change in most of the Q xyz values also points towards that the increase in the brightness of Gaia 20eae is not consistent with a dust-clearing event, rather an intrinsic change occurred in the SED. It is also to be noted that in the 'Active Plateau Region' of Gaia 20eae starting from JD=2549141, the value of Q V RI is close to 0. This might imply that there was no change in the SED intrinsically in this duration. Figure 5 shows our medium resolution spectra covering a very broad wavelength range (∼0.4 -2.4 µm). The optical spectra are flux-calibrated whereas the NIR spectra are only normalized spectra. Both the optical and NIR spectra of Gaia 20eae consist of a mixture of the lines typically observed in the FUor and EXor family of sources. Evolution of these spectral lines at different epochs of the outburst phase are shown in Figure 6. Similar to the FUor sources, Gaia 20eae exhibits blueshifted absorption features in Na I resonance line and Hβ line, indicative of the powerful winds from the source. It also shows strong P Cygni profile in Hα and Ca II IR triplets (IRT) in emission. Fe II line at λ5018Å which, is seen in emission in EX Lupi, is found to be in absorption but the Fe II line at λ6433Å is found to be in emission. K I λ7694Å and O I λ8446Å lines are found to be in absorption. The strength of Hβ and Na I D lines can be seen decreasing during the outburst phase of Gaia 20eae. Spectroscopically, in the optical regime the spectrum of Gaia 20eae resembles a FUor, whereas in the NIR regime it is more or less similar to an EXor.

Spectral features
Our medium resolution NIR spectra show several distinct spectral features, most of them are in absorption. The gaps in the spectra represent atmospheric absorption windows due to the broad H 2 O and OH bands. We could identify some prominent lines: Ca II IRT, He I at λ10830Å and the CO bandheads.
CO bandhead in K band: The CO (2-0) and CO(3-1) bandhead absorption are one of the defining characteristics of FUors . The CO bandheads in Gaia 20eae are in emission, implying a temperature inversion at the surface of the protoplanetary disc. This is similar to that observed in other EXor sources like V2492 Cyg (Aspin 2011). Thus based on the CO bandhead lines, Gaia 20eae resembles more of an EXor source. 20 Radial velocity of Gaia 20eae: Due to lack of symmetric photospheric lines in the high resolution spectrum of Gaia 20eae, it is hard to estimate the radial velocity of the star. The chromospheric Fe I emission lines were found to be the most symmetric lines in the spectrum, and the line center of the Fe λ8387.77Å at 20 km s −1 is taken as the stellar radial velocity with respect to the solar system barycentre in this study. The corresponding velocity in the local standard of rest reference frame comes out to be ∼35 km s −1 . It is to be mentioned that the peak velocity of the 13 CO for the molecular cloud 'MC2' is 42 km s −1 with respect to the local standard of rest (Retes-Romero et al. 2017 The panel (d) in Figure 6 represents the evolution of O I line at λ7773Å of Gaia 20eae during our spectroscopic monitoring period. The formation of O I line at λ7773Å in T Tauri stars, which is an indicator of turbulence, is attributed to the presence of warm gas in the envelope surrounding the disc or in the hot photosphere above the disc (Hamann & Persson 1992a). Table 5   have also tabulated the EW values of other lines during different epochs of the outburst phase of Gaia 20eae in Table 5. Outflow wind velocity of the Gaia 20eae is estimated from the blue-shifted absorption minima of Hα, Na I D and Hβ lines . The estimated values of wind velocity by Doppler shift at different epochs in the outburst phase of Gaia 20eae are listed in the Table 5. The values show the variation from −630 to −203 km s −1 . The mean velocity of the outflow wind velocity for Hα, Na I D and Hβ comes out to be −505 ± 62 km s −1 , −356 ± 49 km s −1 and −339 ± 136 km s −1 , respectively. The typical error in the outflow wind velocity estimation is ∼25 km s −1 , therefore, a large scatter in its values can be attributed to intrinsic variation of outflow winds during the outburst phase.
Resonant scattering from meta-stable Helium atoms are excellent traces of the outflow winds from YSOs (Edwards et al. 2003). The EUV to X-ray radiation from magnetospheric accretion or chromosphere activity can cause significant formation of meta-stable triplet ground state of Helium atoms. During the ∼2.5 hours when they typically survive in this meta-stable state, they could resonant scatter the λ10830Å photons result-ing in a strong absorption signal at the local velocity of the gas. Figure 7 shows the very strong blue-shifted He λ10830Å absorption signature in the high resolution spectrum of Gaia 20eae. On the red side, the absorption profile extends to about +200 km s −1 and on the blue side, the absorption profile extends beyond −400 km s −1 . Unfortunately, the high resolution spectrum beyond −400 km s −1 could not be measured since it falls outside the detector in the HPF spectrograph. Our medium resolution TANSPEC spectra in the panel (f) in Figure 6 shows the blue-shifted absorption extending to −513 km s −1 on 2020 October 24 and reducing to −493 km s −1 by 2020 November 6. Such strong blueshifted He λ10830Å triplet signatures are common in YSOs with strong outflows. Gaia 20eae is not an exception. The reduction in the blue-shifted wing velocity of He λ10830Å over a span of two weeks could be either due to change in the structure of the outflow winds or due to drop in the EUV -X-Ray irradiation. Ca II IR triplet emission lines are believed to originate in the active chromosphere as well as in the magneto−spheric accretion funnel regions (Hamann & Persson 1992b;Muzerolle et al. 1998). Our observation of Gaia 20eae on 2020 September 12 detected a redshifted absorption component (with respect to the stellar rest frame) in all the three Ca II IR triplet lines (See Figure 8). The smooth curves show the best fit of a double Gaussian model, where the first component fits the broad emission line and the second component fits the red-shifted absorption component at +25 km s −1 . The red-shifted absorption cannot originate in stellar wind or outflows. One possible region of origin could be the hot in-falling gas in the magnetospheric accretion funnel (Edwards et al. 1994). The lower panel in Figure 8 shows A) is inconsistent with the optically thin line formation scenario (1:9:5). 21 Constrain on the viewing angle: The detection of the absorption profile implies the line of sight is along the increasing temperature gradient, and we are not seeing an infalling hot gas in the foreground of a cooler environment. i.e., the viewing angle is through the accretion funnel to the footprint on the stellar surface where it is hottest. For a star of mass M, radius R * , and the disc infall radius r m , the velocity of the infalling gas along the magnetic field line direction is given by the formula v p (r) = [ 2GM R * ( R * r − R * rm )] 1/2 (Hartmann et al. 1994). For Gaia 20eae, this would imply a velocity of ∼350 km s −1 at the base of the funnel, and in the order of ∼50 km s −1 (or ∼25 km s −1 along line of sight) at a radius very close to start of infall near the truncated accretion disc. This combined with the requirement of hotter background against which this low velocity infall gas is viewed, constrains the viewing angle as shown in Figure 9. Figure 10 shows the Hydrogen Paschen lines from Pa (14-3) to Pa γ (6-3). Only the lines which are not completely lost in telluric bands are plotted here. The higher energy level Paschen lines are detected as broad absorption lines extending from −250 km s −1 to +250 km s −1 . However, in the lowest energy levels lines, Paγ λ10938.086 and Paδ λ10049.369, on top of the broad absorption component, we also detect an emission component at the core of the lines. The strength of this emission component decreases as we go to lines of higher energy levels.

DISCUSSION AND CONCLUSIONS
Gaia 20eae is the farthest discovered FUor/EXor type Class ii YSO undergoing outburst of ∼4.25 mag in G band. We have found that the present brightening of Gaia 20eae is not due to the dust clearing from our line of sight towards the source but due to an intrinsic change in the SED (warming of the continuum component). The LC of Gaia 20eae in the quiescent phase is showing a small scale fluctuation of amplitude of 0.2 mag and period of ∼2 days.
In the outburst phase, Gaia 20eae is showing a transition stage during which most of its brightness (∼3.4 mag) has occurred at a short timescale of 34 days with a rise-rate of 3 mag month −1 . This rise-rate of Gaia  Giannini et al. 2017, an EXor source). Such difference in timescales of the rise-rates, possibly, implies a different trigger mechanism in Gaia 20eae resulting in the present luminosity outburst. Once it reached maximum brightness, it slowly started to decay from its maximum brightness with a decay rate of 0.3 mag month −1 . The present decay rate is similar to that of bonafide EXor class of sources, EX Lupi and VY Tau, which returned to its quiescent stages in 1.5-2 years after their maximum brightness state (Herbig 1977). The present decay rate is also similar to that of V899 Mon, which transitioned to a short quiescent state from its 2010 outburst state (Ninan et al. 2015). The decay rate of Gaia 20eae is surprisingly similar to that of V1118 Ori being equal to 0.3 mag/month thus pointing to the fact that a similar relaxing phenomenon is occurring in Gaia 20eae also. During the outburst phase, Gaia 18dvy is also showing small scale fluctuations having amplitude of ∼0.2 mag in all the bands. Such a short scale accretion variability has also been reported by Ninan et al. (2015) for V899 Mon. Similar fluctuations were observed in FU Ori, and may be due to flickering or inhomogeneities in the accretion disk (Kenyon et al. 2000;Siwak et al. 2013;Szegedi-Elek et al. 2020). Few interesting spectral features of Gaia 20eae are tabulated in Table 6 to classify Gaia 20eae by comparing its property with different classes of the episodically accretion low mass young stars Connelley & Reipurth 2018). Most of them are matching with EXor source but the Hβ absorption line and P Cygni profile of Hα line hint towards the FUor source classification. The P Cygni profile in Hα originates from the winds generated due to accretion of matter through accretion funnels by the process of magnetospheric accretion. During our spectroscopic monitoring, the P Cygni profile of Hα line showed substantial variations. We have found that the outflow wind velocity for Gaia 20eae shows a large scatter which may be due to the intrinsic variation of wind velocity during the outburst phase. As Hα line originates from the innermost hot zone powered by accretion, the EW of the emission component of Hα line is an approximate indicator of the accretion rate. Table 5 shows a large variation in EW values of Hα line during our monitoring period. Similar to this, EW of O I λ7773Å also varied upto ∼ 75%. This indicates towards the highly turbulent accretion activities going on in Gaia 20eae during its outburst phase. This is also evident from the short scale fluctuations of the photometric magnitudes observed during the same period. These properties of Gaia 20eae are similar to that of the V899 Mon which also showed heavy outflow activities and in- Absorption Emission Absorption Hα P Cygni Emission P Cygni CO(2-0) and CO(3-1) Absorption Emission Emission crease in disc turbulence as it transitioned to its quiescent state after its first outburst (Ninan et al. 2015). Therefore, similar to V899 Mon, the present outburst of Gaia 20eae might be triggered by the magnetic instabilities in magnetospheric accretion. Gaia 20eae also clearly shows the decaying phase less than 15 months as well as CO band-heads in the emission. These features suggest that Gaia 20eae broadly resembles the EXor family of sources. Our high resolution spectrum shows a very strong blue-shifted He λ10830Å absorption signature, which indicates a very strong outflow activity in Gaia 20eae. We have also detected a red-shifted absorption component in all the Ca II IR triplet lines, which could be due to the hot in-falling gas in the magnetospheric accretion funnel. As far as we know, this is the first reported direct detection of an infall signature in Ca II IR triplet lines in FUors/Exors family of objects. We believe that is strong evidence for the magnetospheric funnel origin of Ca II IR triplet lines in heavily accreting YSOs. Based on this, we have also constrained the viewing angle to be such that it is through the accretion funnel to the footprint on the stellar surface.

ACKNOWLEDGMENTS
We thank the anonymous reviewer for valuable comments which greatly improved the scientific content of the paper. We thank the staff at the 1.3m DFOT and 3.6m DOT, Devasthal (ARIES), for their co-operation during observations. It is a pleasure to thank the members of 3.6m DOT team and IR astronomy group at TIFR for their support during TANSPEC observations. TIFR−ARIES Near Infrared Spectrometer (TANSPEC) was built in collaboration with TIFR, ARIES and MKIR, Hawaii for the DOT. We thank the staff of IAO, Hanle and CREST, Hosakote, that made these observations possible. The facilities at IAO and