Spectral Variability of Herbig Ae Star V1295, Aquila

Abstract

This article presents the results of long-term spectral studies of the unusual Herbig Ae star V1295 Aql. The variability in the spectrophotometric parameters of the emission and absorption components of the hydrogen lines Hα, Hβ, and D Na I—and of various metallic lines—is studied. The periodic variability in these parameters over a period of 51.7 days, recently established using mean longitudinal magnetic field measurements, is confirmed. The obtained average value of the radial velocities of metal absorption lines is Vr = 0.75 ± 1.85 km/s. According to our long-term observations, the radial velocities of the star do not change over time. This indicates the absence of a close stellar-mass companion.

1. Introduction

The star V1295 Aql (HD190073, MWC 325, Sp A2IIIe–B9IVep + sh) is located far from the well-known star-forming regions in the constellation Aquila and has a large Infrared (IR) excess due to thermal radiation from cold dust around the star [1]. In the optical range, emission components in the spectral hydrogen lines of Hα, Hβ, He I 5876, D1 and D2 Na I, SiII 6347 and 6371, [OI] 6300, FeII, TiII, MgI FeI, etc., have been recorded in various studies [2,3,4,5,6]. The detection of the MgII resonance lines with the P Cyg profile and the FeII, OI, SiII, SiIII, and SiIV lines in the far-UV range showed that V1295 Cyg is very similar to the typical Herbig AeBe-type stars (HAeBe) AB Aur, HD163296, and HD31648 ([1,7,8]).

The mean longitudinal magnetic field of the order of ∼100 G has been reported and confirmed in several studies [9,10,11,12]. In the work by Aarnio et al. [13], the authors used Magellan Inamori Kyocera Echelle (MIKE) spectra and reported that RV = −1.2 ± 1.3 km/s and v sin i = 3.19 ± 2.45 km/s, which was consistent with [14]. With the physical parameters, an effective temperature of Teff = 9250 ± 250 K, a mass of M = 2.9 ± 0.5 M☉, and a radius of R = 3.6 ± 0.5 R☉—assuming that the stellar rotation axis and the disk share an inclination of 34° ± 2° [15]—the authors estimated a rotation period of ∼32 days.

In the study by [9], the best fit to the dataset from 2011 to 2012 corresponded to a rotation period of 39.8 ± 0.5 d, reproduced in all except the 2012 observations; however, a number of the 2011 data points were discordant with this period. No significant period could be identified in the 2011 dataset, while a period of 40 ± 5 d was clearly detected in the 2012 dataset, which also was observed when analyzing these data separately.

Analysis of spectral energy distribution and parallax data from Gaia EDR3 [16] has recently yielded precise fundamental parameters. The following stellar parameters were derived: spectral type, B9; effective temperature, Teff = 9750 ± 125 K; mass, M = 6.0 ± 0.2 M☉; radius, R = 9.68 ± 0.44 R☉; and age, 0.30 ± 0.02 Myr. The authors of [13] used a radial velocity of Vr = −1.2 ± 1.3 km/s and a v sin i = 3.19 ± 2.45 km/s in their study.

Recently, Jarvinen et al. [6], with numerous spectropolarimetric observations of V1295 Aql, detected a rotation/magnetic period of p = 51.70 ± 0.06 days using the longitudinal magnetic field measurements. The magnetic obliquity angle = 82.9 ± 6.4 and the dipole strength of Bd = 222 ± 66 G were determined using the magnetic phase curve. Importantly, in this study, the construction of dynamical spectra for hydrogen emission lines revealed the first image of a magnetosphere for a Herbig Ae/Be star. Two-dimensional magnetohydrodynamical simulations involving non-isothermal gas showed that the magnetosphere is compact with a radius of about 3 R☉ and that the wind flow extends over tens of R☉. With a reported radius of the accretion disk of 1.14 au around HD190073, the distance between the star and the disk is about 25 R☉ [6].

Interestingly, study [17] revealed the presence of a super-Jupiter-mass candidate around V 1295 Aql at a separation of about 1.1”. Ibrahim et al. [18] reported that their modeling is consistent with a near face-on disk with an inclination of ≤20° and an average radius of a ring-like structure around the star of 1.4 ± 0.2 mas (1.14 au), which was interpreted as the dust destruction front. The observations indicated that the disk around HD 190073 has a skewness on a finer spatial scale, possibly due to a low-mass companion.

The detection of a magnetosphere around V1295 Aql and the possible presence of lower-mass companions at different distances obviously makes this system a valuable example case for studying the magnetic interaction between a host star, its companions, and an accretion disk.

Using our long-term spectral observations of V1295 Aql, we present in this work our study of various spectrophotometric parameters characterizing numerous spectral lines in the spectrum of the star to verify the newly detected rotation/magnetic period of 51.7 days.

2. Observational Material

The spectral material used in our study is presented in

Table 1. Obtained spectral material.

Data	MJD	N	Range, Å	Spectrograph	R	Mean Exposition (Sec)
June–September 2015	57,177.871–57,275.813	22	4400–6700	CES	14,000	900
June–September 2016	57,581.938–57,640.822	6	3700–8000	SHAFES	28,000	1500
June–August 2017	57,951.866–57,978.781	7	3700–8000	SHAFES	28,000	1500
June–August 2018	58,300.871–58,337.969	5	3700–8000	SHAFES	28,000	2500
August 2019	58,699.727–58,719.802	5	3700–8000	SHAFES	28,000	2500
June–August 2020	59,034.942–59,070.8	3	3700–8000	SHAFES	28,000	2500
August 2021	59,448.809	1	3700–8000	SHAFES	28,000	2500
July 2022	59,780.881	1	3700–8000	SHAFES	28,000	3000
June–August 2023	60,145.789–60,178.855	4	3700–8000	SHAFES	28,000	3600

. Observations were carried out using the 2 m Karl Zeiss telescope of the Tusi Shamakhy Astrophysical Observatory (ShAO) of the Ministry of Science and Education of Azerbaijan. In 2015, we used the Cassegrain Echelle Spectrograph (CES), which was constructed on the basis of the Universal Astronomical Grating Spectrograph (UAGS) [19]. For the detector, we used a CCD matrix at 530 × 580 px in size, with pixels of 24 μm, developed at the SAO RAS. Observations of the obtained spectra covered a wavelength range of λ4700–6700 Å. In total, we obtained twenty-two pairs of spectra in 2015 (two spectra per night), which were suitable for processing with a spectral resolution of 14,000.

Observations in 2016–2023 were carried out using the ShAO Fiber Echelle Spectrograph (ShAFES), developed jointly by specialists from ShAO and SAO RAS. A detailed description of the technical characteristics of this spectrograph is given in [20]. A CCD STA4150A (USA) with a size of 4096 × 4096 px and a 15 μm pixel size was used. The spectral range in the spectrograms was λ3700–8000 Å. Observations had a spectral range of λ3700–8000 Å with a binning of 2 × 2 pixels, which allowed us to obtain a spectral resolution of R = 28,000.

Over the last eight years, from 2015 to 2023, high-resolution Echelle spectra were obtained over 54 nights (see Table 1). In all spectra, the signal-to-noise level reached an average of S/N = 80–100 in the region of the Hα line. The errors in measuring the radial velocities for standard stars with sharp lines were 1–1.5 km/s. For spectra with a resolution of 14,000, positional errors are nearly twice as large. The equivalent widths of the Balmer lines of hydrogen were determined with a precision of over 4–5% and, for the intensity measurements, of about 0.5%. The equivalent widths of absorption lines less than 0.05 Å were determined with errors at 10%.

3. Measurements and Results

For the analysis, we used the results of measurements of spectrophotometric parameters of the hydrogen lines Hα, Hβ, and lines D NaI, HeI, SiII, FeII, TiII, MgI, etc. For numerous spectral lines, we measured the equivalent widths EW, radial velocities RV, central depths I, and the full width at half maximum FWHM. Most spectral line profiles exhibited two emission peaks and a central absorption, therefore, each profile detail was signed as follows: for equivalent widths of blue emission EW1e, central absorption EWa, and red emission EW2e. The same principle was used to designate radial velocity measurements and other parameters of spectral lines. Some results of the study of spectral changes in the star are described in the works [4,5]. The authors showed the presence of spectral variability on the time scales from several days to several years.

As mentioned in Section 1, two different rotation/magnetic periods are reported in the literature, 40 and 51.7 days, respectively [6]. For both mentioned periods, we calculated the distribution of the data points over the rotation cycle (phases) of involving individual parameters of spectral lines. We constructed phase curves separately for the June–September 2015 season (22 days) since this is the densest series of observations over one season, and as can be seen from Table 1, these spectra were obtained with a resolution of 14,000. The data for 2016–2023, in each year, do not represent such a dense series, and were obtained with a resolution of 28,000. Therefore, details for weaker metallic lines in the spectrum of V1295 Aql were measured from such high-resolution spectrograms.

Assuming the period of ~40 days, our phase curves constructed for our measurements did not present a periodic character, although for some seasons certain parameters showed a quasi-periodicity. However, the vast majority of parameters for the entire dataset—even for individual seasons—did not show periodical behavior. Therefore, we do not consider this 40-day period in the following presentation of our results. The phases of the period of 51.7 days for our data were calculated with elements, T_max (MJD) = 56,108.06 + 51.70(6) × E, as presented in [6]. The phase curves for individual spectral lines were constructed separately for June–September 2015 and for the years 2016–2023.

3.1. Hα and Hβ Lines

As an example, Figure 1 shows the Hα and Hβ line profiles obtained in 2015. As can be seen, both lines exhibited a p Cyg-type profile: absorption on the blue and weak form in red wings; emission peaks with a dominant peak with a red shift; and central absorption [5]. Phase curves with a period of p = 51.7 days have been calculated for individual components of the Hα and Hβ hydrogen lines.

Our data show that, in general, the parameters of the hydrogen Hα and Hβ lines represent a certain periodicity in a weak form but with clearly distinguished maxima or minima at a rotation phase of 0.5 for different parameters of the spectral lines. As an example, Figure 2 shows the phase curves of the Hα line parameters constructed using 2015 data. The two upper panels from left to right refer to the radial velocity of the second emission peak RV2p and the bisector radial velocity of the emission component RVbis. The lower panels, from left to right, show the phase curves constructed from the equivalent width of the emission component EWe and the absorption component on the red wing EW2a. As can be seen, equivalent widths of the red absorption and emission components are changing in the inverse form: around phase 0.5 we see a maximum of the emission component and a minimum of the absorption component. The radial velocities of the emission component peak RV2p and the bisector velocity RVbis of emission have a well-defined redshift at zero phase.

In Figure 3, we present phase curves for the line Hβ for the 2015 data. At the top, we show phase curves of equivalent widths of blue absorption EW1a and depth Ie of the emission component, and in the bottom panels, equivalent widths of the emission component EWe and its peak radial velocity RVp. The equivalent widths present the periodic dependence more clearly.

In Figure 4, phase curves for the Hβ line parameters for observations from the years 2016–2023 are presented. In the left panel, we show the phase variability for equivalent widths of the blue absorption component EW1a (open circles) and the total emission component EWe. In the panels on the right side, we show radial velocity phase curves for the blue RV1e (at the top) and for the red RV2e components (at the bottom). As can be seen here too, we observe periodical variability in the studied parameters. However, it seems that since the results of long-term observations were used, apparently, due to various non-stationary processes occurring from time to time, the periodic process is distorted in different seasons.

In Figure 4, phase curves for the line Hβ parameters for data are shown. In the left panel, we present the phase variability for equivalent widths of blue absorption component EW1a (open cycles) and total emission component EWe. In the right panels, we present the radial velocity phase curves for the blue RV1e (at the top) and for the red RV2e components (at the bottom). As can be seen, here too we observe periodical variability in the indicated parameters. It is also observed that, since the results of long-term observations were used due to various non-stationary processes occurring from time to time, the periodic process is distorted in different seasons.

3.2. D1, D2 Na I Lines

Na I D1, D2 doublet neutral sodium lines in the spectrum of V1295 Aql are seen as a double peak with deep central absorptions (see [2,4]). In Figure 5, we show data point distributions of some spectral line parameters. The top left panel for the D2 Na I presents radial velocities for blue emission component RV1e, separate from the data for 2015 (open circles) and 2016–2023 (black points); in the right panel, for the D1 Na I line, equivalent widths for the blue emission component EW1e are shown for the data from 2015 (open circles) and 2016–2023 (black points). In the bottom left panel, we show for the line D2 Na I equivalent widths of the blue emission component EW1e and central absorption component EWa for the observations from 2016 to 2023. In the bottom right panel, we show radial velocities RVa of the central absorption component for the line D Na I.

In Figure 5, significant variations in the Na I lines are observed in the equivalent widths and radial velocities of the emission components of the doublet. The equivalent widths of the absorption components do not show variations within the error limits. The radial velocities for the absorption components of the doublet are close to the velocity of the mass center of the star [13]. Despite this, a significant spread of values RVa is observed in individual phases, which indicates the existence of an additional contribution from the circumstellar disk in these lines.

3.3. Weak Metal Lines

Most of the spectral lines of metals in the spectrum of the star, as a rule, have the same profiles as the lines of the sodium doublet D Na I; namely, two emission peaks are observed in the blue and red wings that are separated by a central absorption ([2,21]). The star’s spectrum also contains purely emission or absorption lines, without additional components. In the work [2], it was shown that if synthetic theoretical spectra are subtracted from the observational spectra of the star, then in the residual spectrum, the profiles of the metal lines are observed as a pure emission peak. This shows that the absorption lines have only a photospheric origin. Knowing this fact, we examined the metal lines for the following tests: (1) to check the phase variations with a period of 51.7 days; (2) to check the variability in the star’s radial velocities using photospheric lines.

To study the variability in weak metallic spectral line components, we used spectra with the spectral resolution of R = 28,000, which were obtained for the 2016–2023 period (see Section 2). We have selected the two most interesting spectral regions, which include a number of emission–absorption lines with various intensity. For 12 selected lines, the spectroscopic parameters of the emission and absorption components were determined.

In Figure 6, we show two selected spectral intervals in the star’s spectra from which the spectroscopic parameters of lines were measured. The parameters of each selected spectral line were calculated for all the available spectra. In the bottom panel of Figure 6, it can be seen that with the increase in emission, the absorption components are weakened and in some lines disappear, for example in the profile of line FeII 5198.

To avoid confusion, the parameters of the emission components of the blue and the red wings were designated by the symbols 1e and 2e, respectively, and the absorption component by the index a. For example, the line parameters were assigned as follows: equivalent widths EW1e, EW2e, and EWa; radial velocities RV1e, RV2e, and RVa. In

Table 2. Equivalent widths and radial velocities of selected spectral lines.

No sp. line	Selected lines	EW1e	rms	EW2e	Rms	EWa	rms
1	TiII 4501.27	0.062	0.023	0.091	0.036	0.012	0.008
2	FeII 4508.29	0.042	0.018	0.065	0.027	0.018	0.009
3	FeII 4515.34	0.031	0.017	0.047	0.022	0.025	0.014
4	TiII 4533.97	0.063	0.023	0.075	0.030	0.021	0.015
5	FeII 4541.52	0.010	0.005	0.022	0.018	0.042	0.013
6	CrII 4558.65	0.030	0.016	0.040	0.022	0.038	0.013
7	TiII 4563.76	0.042	0.022	0.059	0.025	0.019	0.010
8	TiII 4571.97	0.093	0.035	0.129	0.042	0.014	0.007
9	MgI 5172.68	0.055	0.023	0.063	0.025	0.033	0.011
10	MgI 5183.680	0.060	0.024	0.087	0.025	0.021	0.011
11	TiII 5188.611	0.028	0.109	0.042	0.025	0.020	0.009
12	FeII 5197.58	0.021	0.035	0.149	0.043	-	-
No sp. line	Selected lines	RV1e	rms	RV2e	rms	RVa	rms
1	TiII 4501.27	−21.1	3.2	20.8	3.3	−0.9	1.9
2	FeII 4508.29	−22.9	3.6	21.0	2.9	−1.3	2.2
3	FeII 4515.34	−24.0	2.7	21.9	4.5	−1.2	2.0
4	TiII 4533.97	−21.6	2.4	27.5	2.8	0.9	2.0
5	FeII 4541.52	−23.1	3.4	23.0	3.8	−0.5	1.8
6	CrII 4558.65	−23.7	2.7	23.4	3.2	−0.9	2.0
7	TiII 4563.76	−22.0	2.6	20.1	1.9	−1.0	2.2
8	TiII 4571.97	−20.7	3.0	20.3	2.3	−0.5	1.7
9	MgI 5172.68	−24.1	3.3	22.8	2.7	−0.3	1.6
10	MgI 5183.68	−22.6	2.9	21.1	2.3	−0.5	1.3
11	TiII 5188.611	−18.5	3.8	25.2	3.2	3.2	3.1
12	FeII 5197.58	−19.4	4.0	17.1	2.7	−2.1	1.6

Note for Table 2: EW and its root mean square (rms) are presented in Å units, whereas RV and its rms are in km/s units.

, the first two columns present the number of selected spectral lines and their identification. In the first top part of Table 2, we present the mean value of equivalent widths and their rms, and in the bottom part, the same but for radial velocities of separate components.

In Figure 7, in the top panels, we show the data point distribution of radial velocities of the blue emission component (RV1e) of the spectral lines TiII 4533 and FeII 5198, and for the red emission component (RV2e) of the spectral lines MgI 5172 and TiII 4572 over the rotation/magnetic period p = 51.7 days.

Figure 8 shows the data point distribution for equivalent widths of blue emission component EW1e for the lines FeII 4508 and TiII 4572 and for the red emission component EW2e for the lines FeII 4508 and TiII 4572. In the bottom panels of Figure 8, we can see data point distributions for the absorption component EWa for the lines FeII 4515 and FeII 4508.

We observe that at phase 1.0, the biggest red shift of radial velocities of the spectrum is visible for both the blue and red emission components. For the values of equivalent widths of emission components, the maxima are observed at phase 0.5, and the minimum at phase 1.0 (Figure 8). It is interesting that the equivalent widths of the absorption components varied in antiphase with the emission components. This is clearly seen in the lower panels of Figure 8 with the data point distributions for the absorption components of the FeII 4515 and FeII 4508 lines.

Our measurements have revealed that the radial velocities of the absorption lines RVa do not vary within the measurement errors. This is clearly seen in Figure 9, which shows the dependence on the phase of the radial velocities of the absorption components for all measured metal lines in the observations from 2016 to 2023. As can be seen from Figure 9, there is a certain scatter of the RVa values for different lines. We assume that a large scatter in RV may be because the measured absorption components of metal lines in the spectrum of this star are sometimes distorted by weak emission components superimposed on the blue and red wings. This may cause a slight asymmetry in the absorption lines and lead to scatter in the RV of measured lines. For all measurements, the average value of the RV of the center of mass of the system is determined by us as −0.75 ± 1.85 km/s. This is in good agreement with the data of [13].

4. Discussion and Conclusions

In this work, we tried first to check whether a certain periodic variability with a rotation/magnetic period of p = 51.7 days reported by Jarvinen et al. [6] is observed in the characteristics of the star V1295 Aql. For this task, we used our data obtained from the results of long-term spectral observations of the star at the Shamakhy Observatory. For the analysis, the parameters of the hydrogen lines Hα and Hβ, D1, D2 Na I, and weak metallic lines were employed. Our data revealed that the radial velocities of the emission components for almost all the lines under consideration show a certain variability with the rotation phase. The maximum shift to the red part of the line spectrum is observed at the phase of about zero. Further, we performed the most accurate measurement of radial velocities using the absorption components of metallic lines. The average value for the center of mass of the star, RV = −0.75 ± 1.85 km/s, was obtained, which is in good agreement with the data of [22,23]. According to our long-term observations, the radial velocities of the star have not exhibited variation over time. This indicates that the star V1296 Aql lacks a close stellar-mass companion.

The variability in the equivalent widths of both emission and absorption components of the selected spectral lines was analyzed. In the data point distributions for the equivalent widths of the emission line components, the maximum is observed at phase 0.5. The absorption components showed minima in this phase. We believe that the change in the equivalent widths of the absorption lines of metals is due to the fact that with an increase in the intensity of the emission lines, the absorption components are filled-in by the blue and red emission components. This contributes to a decrease in the equivalent widths of the absorption lines.

When the line peak is below the continuum level, the equivalent widths of the absorption lines can be measured, and in this case, we accept the line structure as a double peak emission. However, if the emission intensity decreases, the depths of the absorption components increase. For this reason, we observed a variation in the equivalent widths of the emission and absorption components in antiphase.

Summarizing our results, we obtain the following conclusions:

• The emission (circumstellar disc) spectrum of the star shows significant variability with a magnetic rotational period of 51.7 days. At the same time, the absorption spectrum of the star does not show variations over time;
• The observed variability in the equivalent widths of the absorption lines is a visible effect caused by a change in the emission spectrum of the star. This observational fact allows us to assert that the magnetic field of the star contributes to the formation of the magnetosphere structure, which is revealed by the rotation period;
• Based on the results of long-term homogeneous spectral observations, it was shown that the radial velocities of the absorption lines do not change. Since we are observing the star practically from the pole-on, the question of the existence of a stellar companion of V1295 Aql remains open.