The Old Moving Groups in the Field of Taurus

The spectroscopic data used in this work are taken from the Large Sky Area Multi-Object Fibre Spectroscopic Telescope (LAMOST) DR5. LAMOST is a 4 m Schmidt telescope of the National Astronomical Observatories of China, located at Xinglong Observing Station, China. With 4000 fibers on board the focus, LAMOST can observe nearly 4000 spectra simultaneously in optical bands of ∼3900–9000 Å, at a resolution of ∼1800 (Cui et al. 2012). In this work, we use the LAMOST DR5 data set.

In order to construct the spectral energy distribution of each source and estimate its extinction, we used optical photometry in the g, r, i, z, and y bands from Pan-STARRS (Chambers et al. 2016) and G, G_BP, and G_RP bands from Gaia DR2 (Gaia Collaboration et al. 2016, 2017, 2018), near-infrared photometry in the J, H, and K_S bands from the Two-Micron All Sky Survey (2MASS; Skrutskie et al. 2006), and near- and mid-infrared photometry in the W1 (3.4 μm), W2 (4.6 μm), W3 (12 μm), and W4 (22 μm) bands from the WISE (Wright et al. 2010) all-sky survey.

In this work, we use DBSCAN to search for groups in the Taurus field. DBSCAN is a density-based clustering method. Its principal idea is if a point belongs to a group, then it should be surrounded by members of the same group in the multidimensional space. The idea is realized by finding neighborhoods of data points that exceed a given density threshold, which is defined as the minimum number of neighbors or data points (minPts) within a given search radius (). The algorithm can be summarized as:

• 1.  
  With the given threshold (minPts and ), starting from a random data point, the code will find all the points inside the radius (e.g., neighborhoods of the data point). If the number of data points is greater than minPts, then all these data points will be regarded as a part of a "cluster."
• 2.  
  From each of these data points identified in Step 1, the code will repeat Step 1 to search for new data points upon the defined threshold above until no more new data points can be found. All the data points found in this step are also regarded as members of the cluster revealed in Step 1.

For DBSCAN, large minPts values and small values are sensitive to those highly concentrated parts of the stellar groups, and on the contrary small minPts and large values will include too much contamination from field stars, and misrecognize groups. In this work, we perform the DBSCAN algorithm with the python package Sklearn (Pedregosa et al. 2012) to a five-dimensional normalized astrometric space that consists of x, y, z, $\mu {{\prime} }_{\alpha }$, and $\mu {{\prime} }_{\delta }$, where x, y, and z are in Cartesian space and $\mu {{\prime} }_{\alpha }$ and $\mu {{\prime} }_{\delta }$ refer to the tangential velocities in the R.A. and decl. direction ⁷ (the distances used to derive these quantities are extracted from Bailer-Jones et al. 2018). We vary minPts from 5 to 15, and from 0.01 to 0.1 to look for the reasonable values which can rediscover the YSO groups identified in Taurus. Roccatagliata et al. (2020) perform a group search among the spectroscopically confirmed members in Taurus, and find six populations, Groups A–F, with well-defined parallax and proper motions. We find that by setting minPts = 9 and = 0.035, the DBSCAN algorithm can rediscover these groups. With minPts = 9 and = 0.035, the DBSCAN algorithm found 22 groups in the whole region. However, we must stress that our group searching could ignore smaller and more sparse groups than the YSO groups since we have set the minPts and to be efficient for searching the YSO groups.

In order to minimize the influence of the projection effect and to remove the contamination in the group member from the field stars, we further refine the result in the above using a more rigorous criteria via two steps:

• 1.  
  For each group, we apply the DBSCAN algorithm to the stars within 15 pc from the center of the group by setting minPts = 9 and varying between 0.1 and 0.2. The normalization scale in this step is much smaller than the above one, and the used here is corresponding to 0.015 to 0.03 if using the same normalization scale as the above. In this step, we reduce the search radius (smaller ) to mitigate the contamination in the group members from field stars. However, we could lose some members which are far from the group center.
• 2.  
  For each group, we only include the sources with proper motions within 2σ from the mean proper motions of the group (see Figures 3 and 4), where σ is the standard deviation of the proper motions for the member candidates of the group.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

During the refinements, we have excluded about 100 sources from our preliminary sample. Finally, we have 630 sources in our refined sample that are grouped into 22 populations (see Figure 5 for their distribution in the sky). The standard deviation of these groups in the tangential velocity space is ∼0.73 km⁻¹. As a comparison, we derive the the median value (∼2.33 km⁻¹) of the standard deviations of tangential velocities derived from the stellar associations in Gagné et al. (2018a) using Gaia DR2 data. This indicates that our group method is very conservative and should have lost some group members. We will further refine their members of individual groups based on their locations in the dereddened ${M}_{{G}_{\mathrm{RP}}}$ versus G_BP–G_RP diagrams (see Section 4.1).

Zoom In Zoom Out Reset image size

We must stress that the group result could be influenced by the projection effect. It would be more reliable for searching stellar groups if the x, y, z, U, V, W spaces were used, where the U, V, W velocities are the velocities in the x, y, and z directions, respectively. However, this would severely cut down the number of stars as only bright stars (G ≥ 13.0) are released with radial velocities (RVs) in Gaia DR2.

We match the member candidates of each group with the known young stars with spectral types in the literature with a 2'' matching radius, and find 298 ones (Cannon & Pickering 1993; Nesterov et al. 1995; Slesnick et al. 2006; Biazzo et al. 2012; Herczeg & Hillenbrand 2014; Kraus et al. 2017; Gagné et al. 2018b; Esplin & Luhman 2019; Kervella et al. 2019). We also search for the spectra of our targets in LAMOST DR5, and retrieve the spectral data for 267 ones. Among them, 48 have been classified with the LAMOST pipeline, and the results are used in this work. For the other 219 sources without the spectral types from the LAMOST DR5, we classify their spectra with the method described in Fang et al. (2017). To evaluate the reliability of our spectral classification, we compared the spectral types of the 106 common sources with the spectral type derived in this work and in the literature (Wichmann et al. 1996; Slesnick et al. 2006; Kraus et al. 2017; Luhman 2018; see Figure 6). The comparison shows there is no systematic difference between the spectral types in the literature and in this work. We note that two sources, BP Tau and HO Tau, show more than a 2σ difference on the spectral types. The large difference on the spectral types for an accreting young star is very common, and can be due to variable optical veiling on the spectra (Fang et al. 2020), since the spectral classification usually does not consider the veiling effect.

Zoom In Zoom Out Reset image size

In total, we have spectral types for 427 sources in our sample, with 298 ones from the literature and 129 derived from the LAMOST spectra.

Our sample contains about 300 well-studied young stars in Taurus, and their published extinctions are adopted in this work (Herczeg & Hillenbrand 2014; Esplin & Luhman 2019). For those without estimates of extinction in the literature, we derive their extinctions. For the ones with spectral types, we convert their spectral types to effective temperatures (T_eff) using the conversions in Fang et al. (2017), which are from Pecaut & Mamajek (2013) for stars earlier than M4 and from Herczeg & Hillenbrand (2014) for stars later than the M4 type. We then achieve their intrinsic colors for G_RP − J corresponding to their T_eff by interpolating the model colors for the 10 Myr isochrone from PARSEC (Bressan et al. 2012), and derive the extinction of individual sources using the average extinction law with the total-to-selective extinction ratio R_V = 3.1 from Wang & Chen (2019). We verify this method by comparing our results with those in Herczeg & Hillenbrand (2014) for the common sources. The comparison shows that both agree well with each other (see the left panel in Figure 7).

Zoom In Zoom Out Reset image size

For the sources without spectral types, we perform a least-square fit to the observed colors, taking the extinction and T_eff as a free parameter. The colors used in the fitting are a combination of the broadband photometry in Gaia's G_BP, G_RP bands and the 2MASS J, H, K_s bands. For each T_eff we obtain the model colors for the 10 Myr isochrone from PARSEC (Bressan et al. 2012), redden them using the same extinction law with R_V = 3.1 from Wang & Chen (2019), and then compare the reddened colors to the observed ones. The best fit to the observed colors yields the extinction used in this work. The extinctions derived in this method are mostly for the diskless sources. To verify this method, we collect a sample of the sources with spectral types in the old groups and estimate their extinctions using the above two methods. We compare the extinctions from the two methods in the right panel in Figure 7, which shows a good agreement between them.

We further verify our derived extinction by repeating the above procedure by using the model colors for the 100 Myr isochrone from PARSEC (Bressan et al. 2012). The derived extinctions are consistent with the above ones using the model colors for the 10 Myr isochrone from PARSEC. The J-band extinction of each source used in this work is listed in Table 1.

In Figures 8 and 9, we show the dereddened ${M}_{{G}_{\mathrm{RP}}}$ versus the G_BP–G_RP CMD of individual groups identified in Section 3. We fit the dereddened ${M}_{{G}_{\mathrm{RP}}}$ versus G_BP–G_RP diagrams with the model isochrones of PARSEC (Bressan et al. 2012). ⁸ The best-fit isochrone for each group is achieved by minimizing the mean distance of its member candidates to the isochrones. The age of the best-fit isochrone is adopted as the age of the group. From the fitting, eight groups (Groups 1–8) have ages of 2–5 Myr, while the other 14 groups have ages ranging from 8–49 Myr.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

In six young groups (Groups 2, 3, 4, 6, 7, and 8), we have found 47 new member candidates in total. We note that many of them locate below the best-fit isochrone for each young group and could be contaminators from "older" populations in the field. This can be even clearer in Figure 10 where we plot the 47 member candidates together in the dereddened ${M}_{{G}_{\mathrm{RP}}}$ versus G_BP–G_RP diagram. In the figure, about 50% of them locate near the 100 Myr old isochrone. In this work, we only include the ones above the 10 Myr old isochrone as the probable members of young groups. This criterion is defined as a compromise between reducing the contamination from the "older" populations and the large spreads in the dereddened ${M}_{{G}_{\mathrm{RP}}}$ versus G_BP–G_RP diagrams for young groups. According to this criterion, we include the 17 sources as the members of the young groups, 13 of which are in Group 7, and exclude 30 the other ones in the further analysis and discussion.

Zoom In Zoom Out Reset image size

After excluding the contamination, we refit the dereddened ${{M}}_{{G}_{{\rm{RP}}}}$ versus G_BP–G_RP diagrams of the five young groups using the PARSEC isochrones. For Groups 2, 3, 6, and 8, we derive similar ages as before, and for Groups 4 and 7, we obtain younger ages than before, 2 Myr versus 3 Myr, and 3 Myr versus 5 Myr. After excluding the contaminators, in Figure 8 we can still see that some sources are located below the best-fit isochrones and look like old populations. All of these sources are known young stars in the literature and harbor disks (see Section 4.2). The locations of these sources in CMDs could be due to accretion activities, the disk orientation, etc. (Fang et al. 2021). Compared with the young groups, the "old" groups show well-defined loci in the CMDs, and their ages can be constrained very well with the PARSEC isochrones. The age of each group is listed in Table 2.

Table 2. A List of the Groups Identified in This Work

Group	Members	PK a	Other Name b	R.A. J2000 c	Decl. J2000 c	pmR.A. c	pmDecl. c	Distance c	Age d	Age e	${A}_{J}^{c}$
				(deg)	(deg)	(mas yr−1)	(mas yr−1)	(pc)	(Myr)	(Myr)	(mag)
1	55	55	A	65.22	27.91	8.68	−25	129${\pm }_{12}^{7}$	3	3	0.21
2	41	40	F	68.88	17.64	12.15	−18	144${\pm }_{8}^{8}$	3	2	0.12
3	19	19		66.13	26.65	11.16	−17	159${\pm }_{8}^{10}$	3	3	0.57
4	27	26	E	68.69	22.88	9.89	−17	158${\pm }_{10}^{8}$	2	2	0.45
5	49	49	B	68.53	24.79	7.15	−21	130${\pm }_{10}^{13}$	2	2	0.25
6	7	5	D	76.88	25.02	2.78	−17	172${\pm }_{4}^{5}$	4	6	0.14
7	19	6	C	78.03	30.43	4.00	−25	156${\pm }_{4}^{6}$	3	5	0.08
8	30	30	C	73.75	30.08	4.54	−24	158${\pm }_{5}^{10}$	4	4	0.18
9	42	10		81.77	24.42	1.53	−18	176${\pm }_{13}^{15}$	8	7	0.10
10	30	4	Oh17-211(3/3)	81.33	25.27	3.65	−25	162${\pm }_{10}^{10}$	10	9	0.08
11	33	9		84.32	23.31	5.94	−37	109${\pm }_{7}^{5}$	11	13	0.03
12	52	18	Oh17-29(6/9)	62.90	19.51	3.76	−14	120${\pm }_{5}^{6}$	21	23	0.05
13	25	7	Oh17-300(3/3) Oh17-3824(1/2)	74.88	17.27	−2.38	−15	117${\pm }_{6}^{6}$	33	40	0.04
14	30	7	Oh17-28(7/9)	70.03	21.68	−0.45	−15	122${\pm }_{8}^{8}$	29	29	0.07
15	7	0		86.35	25.57	10.38	−20	201${\pm }_{4}^{3}$	29	40	0.03
16	57	5		87.91	20.46	8.46	−19	199${\pm }_{22}^{14}$	32	36	0.04
17	10	0	Oh17-456(1/2)	86.01	25.97	9.18	−19	211${\pm }_{3}^{6}$	32	37	0.07
18	10	0		64.82	30.79	19.57	−25	170${\pm }_{5}^{4}$	37	49	0.16
19	7	0		68.85	29.55	18.52	−23	183${\pm }_{2}^{4}$	33	36	0.20
20	13	1		89.48	17.21	8.46	−20	185${\pm }_{7}^{5}$	37	43	0.05
21	19	1		86.94	23.50	10.27	−21	188${\pm }_{7}^{7}$	38	46	0.02
22	17	1	Oh17-1099(2/2) Oh17-43(2/6)	57.11	13.84	23.97	−24	155${\pm }_{8}^{6}$	49	59	0.08

Notes.

^a Number of stars with known spectral types in the literature (Slesnick et al. 2006; Kraus et al. 2017; Esplin & Luhman 2019). ^b "A"–"F" denote the Groups of Roccatagliata et al. (2020). Oh17 and "Oh17-X(c/t)" indicate Oh et al. (2017) and the star pair of it. "X" denotes the name of the star pair, "t" denotes the total members of the star pair, while "c" denotes how many members are contained in the catalog of this work. ^c The mean position, distance, proper motion, and J-band extinction of each individual group. ^d The ages of the groups that are derived by fitting the dereddened CMDs with the isochrones of PARSEC models. The extinction of each individual star is considered. ^e The ages derived by fitting the observed CMDs with the isochrones of PARSEC models, but assuming that all members of a group share the same extinction and treating the extinction as a free parameter.

A machine-readable version of the table is available.

Download table as:  DataTypeset image

We also fit the observed ${M}_{{G}_{\mathrm{RP}}}$ versus G_BP–G_RP diagram using the PARSEC isochrones with an assumption that all member candidates of each group have the same extinction. ⁹ As a comparison, the ages of the best-fit isochrones are also listed in Table 2. In general, the ages estimated in two ways are consistent with each other.

We verify our age estimate using the strength of the Li i absorption line at 6708 Å which is a good indicator of stellar ages (Soderblom et al. 2014). In Figure 11, we show the Li i absorption line in the LAMOST spectra of the sources with spectral types around M3. Although the spectral resolution is relatively low (R ∼ 1800), we can clearly see the rapid Li depletion around 10–30 Myr for the M3-type young stars, which is consistent with the results in the literature (see Zuckerman & Song 2004 for a review).

Zoom In Zoom Out Reset image size

In this work, we use the infrared photometry from 2MASS and the AllWISE Catalog to search for objects with disks. In both catalogs, we extract photometry for 609 common stars. In Figure 12, we show their infrared color–color diagrams used to identify the sources with disks. In the figure, the stars to the bottom-right side of the reddening vector are too red to be explained by the reddening of diskless stars, suggesting there is substantial excess emission above the photospheric level in these wavelengths, which is most likely due to emission from the disks surrounding them. Thus, we identify these sources as stars with disks. In this way, we identify 123 reliable disk-bearing sources (see Figure 12), including 104 known ones confirmed in the literature (Maheswar et al. 2002; Eisner et al. 2004; Rebull et al. 2011; Esplin et al. 2014; Menu et al. 2015; Esplin & Luhman 2019). ¹⁰ We compare our disk classification with the result in Esplin & Luhman (2019), and notice that we have missed 20 disk-bearing stars. Among them, 14 ones only show the infrared excess emission in the WISE W4 band or Spitzer 24 μm band, and thus are not identified in this work. ¹¹ Among the other six sources, three of them (Sources 63, 250, and 255) cannot be cross-matched with the sources in the AllWISE catalog within 2''. Another two stars (Sources 19 and 118, red rectangles in Figure 12) are located near the boundary in the left two panels in Figure 12, which we use to identify the sources with infrared excess emission, and might have weak infrared excess emission in those WISE bands. Unfortunately the two sources have no photometry in the WISE W3 band. For the remaining one source (Source 91 or V710 Tau A), it is one component of a binary system (V710 Tau), and the WISE data cannot resolve the system. In this work, we list the 20 sources harboring disks in Table 1. The other 19 disk-bearing sources are first revealed in this work. In Figure 13, we show the spectral energy distributions (SEDs) of these sources. Among these sources, eight are in young groups (Groups 6, 7 and 8). Interestingly, we also find 11 disks in the relatively older Groups (8–11 Myr), two in Group 9, four in Group 10, and five in Group 11.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Classical T-Tauri stars (CTTSs) and weak-line T-Tauri stars (WTTSs) can be distinguished from the Hα emission (White & Basri 2003). CTTSs usually show strong and broad Hα emission lines due to the accretion process, and WTTSs present weak and narrow Hα emission lines due to the chromospheric activity. In this work, we use the criteria from Fang et al. (2009) to divide the stars with LAMOST spectra into WTTSs and CTTSs based on their Hα EWs. Among the 224 stars with Hα emission in their spectra, 38 are CTTSs. The results are listed in Table 1 (column 11). Among the CTTSs, 37 sources are in the groups with ages younger than 5 Myr, and one (Source 332) is in Group 10 at an age of 10 Myr (see the discussion in Section 5.2).

In the relatively older groups, we discover 11 sources with circumstellar disks (see Figure 13). Among the 11 sources, we have LAMOST spectra for four of them, Sources 331 (Group 10), 332 (Group 10), 355 (Group 11), and 368 (Group 11). Based on their Hα EWs, Sources 331, 355, and 368 are classified as WTTSs and Source 332 is a CTTS. The Hα EW of Source 332 is about −53 Å, which is much stronger than the threshold (−18 Å) used to distinguish CTTSs from WTTSs. Figure 14 compares its Hα line with that of a WTTS star (Source 374) with a similar spectral type. The Hα line profile of Source 332 is well resolved in the LAMOST spectra. We fit its Hα line profile using a Gaussian function and derive an FWHM of 243 km s⁻¹. With an assumption that the intrinsic Hα line profile is a Gaussian function and after being deconvolved from the spectral resolution (R ∼ 1800), the intrinsic full width at 10% maximum (FWHM_10%) of the line profile is 343 km s⁻¹, which gives an accretion rate of ∼3 × 10⁻¹⁰ M_⊙ yr⁻¹ using the relation between the FWHM_10% and accretion rates from Natta et al. (2004). Source 332 shows an SED, typical for an evolved disk, with no infrared excess in the WISE W1 and W2 bands, and weak excess emission in the WISE W3 and W4 bands. Thus, we might witness the accretion process at the latest stage of disk evolution.

Zoom In Zoom Out Reset image size

In this work, we identify eight groups, Groups 1–8, with ages younger than ∼4 Myr. Among them, Group 7 has the largest members newly identified in this work. In this group, there are 19 sources and 13 are new. In Group 7, we have seven M4–M5 type YSOs with LAMOST spectra. In Figure 15 we show the Li i absorption line at 6708 Å for seven sources. As a comparison, we also show one M4.0 type young stars (Source 28) in Group 1 at a similar age (3 Myr) to that of Group 7. The strengths of the Li i absorption lines of the seven members in Group 7 are comparable to the one of Source 28 (see Figure 15). This supports the finding that Group 7 is of a similar age to Group 1.

Zoom In Zoom Out Reset image size

In Group 7, the six known members belong to the cyan population in Luhman (2018) or Group C in Roccatagliata et al. (2020). Group 7 locates to the east of Group 8, and is at a similar distance and age to this group. Group 8 is also in the cyan population in Luhman (2018) or Group C in Roccatagliata et al. (2020). Thus, there is one possibility that both Groups 7 and 8 belong to the same group and their separation could be because we have not corrected for the projection effect in our grouping procedure. We search for the RVs for the members in Groups 7 and 8 in Gaia DR2, and find the values for three sources. We correct for the projection effect for Groups 7 and 8 employing the method in Kraus et al. (2017). We derive the median values of the UVW velocities for the three sources and assume that all the members in the two groups share these median values in the UVW velocity space. For a group member, we subtract its measured proper motions with the ones expected at its location. Figure 16(a) shows the residual proper motions for Groups 7 and 8, and the standard deviations of the residual proper motions for a combination of the two groups are both ∼1 mas yr⁻¹, which are similar to the values of other young groups in Taurus. Thus, it is very likely that both Groups 7 and 8 belong to the same group.

Zoom In Zoom Out Reset image size

We further verify our grouping results for the young stars with the results in the literature. Compared with the ones in Luhman (2018), Groups 1 and 5 belong to the red population in Luhman (2018), Groups 2, 3, 4 are in the blue population, and Groups 6, 8 and 7 are in the cyan population. In order to compare with the result in Roccatagliata et al. (2020), we cross-match the members of our young groups with the sources which have a probability of belonging to one of the Groups A–F in Roccatagliata et al. (2020) of more than 80%. We find that Group 1 in this work corresponds to Group A, Group 5 to Group B, Group 6 to Group D, Group 4 to Group E, Group 2 to Group F, and both Groups 7 and 8 to Group C.

In this work, we find 14 groups, Groups 9–22, with ages ranging from 8–49 Myr and distances within ∼110–210 pc. The distributions of these groups are shown in Figure 5. In these groups, there are 353 members and 37 have been cataloged in Kraus et al. (2017). In Group 11, there are 33 members and six of them are in the 118 Tau group (total 12 members) in Gagné et al. (2018a). The age of 118 Tau is estimated to be ∼10 Myr in Gagné et al. (2018a), which is consistent with the age of Group 11. Therefore, it is likely that both Groups 11 and 118 Tau are from the same group.

Oh et al. (2017) searched the comoving star pairs by applying a marginalized likelihood ratio test to 3D velocities of stars of Gaia DR1 (e.g., the Tycho-Gaia Astrometric Solution). And they introduced 13,085 comoving star pairs among 10,606 unique stars. Through comparison with their catalog, we notice that about 20 members of the older groups of this work are also cataloged in Oh et al. (2017). In Group 12 (total 52 members) there are six sources belonging to the star pair 29 (nine members) in Oh et al. (2017), in Group 14 (total 30 members) seven sources are in star pair 28 (9 members), and in another four old groups there are one to three sources which have been included in the star pairs in Oh et al. (2017); see Table 2 for more details. Excluding the ones overlapping with those in the literature, seven old groups are discovered in this work, including Groups 9, 15, 16, 18, 19, 20, and 21.

We investigate the kinematic relations among the identified YSOs in Taurus and old groups in this work. In a similar way to Groups 7–8, we derive the residual proper motions for Groups 9–22. We adopt the UVW velocities for the identified YSOs from Luhman (2018), and take the median values of them as the common values for the YSOs in Taurus. We derive residual proper motions for the old groups. For Groups 12–22, their residual proper motions are far away from those known Taurus YSOs, indicating that they are not kinematically related to the known Taurus members. Furthermore, these groups are old (>20 Myr), meaning that for them the Taurus molecular clouds may not have been present for long (Hartmann et al. 2001, 2012). In Figures 16(b)–(d), we show the residual proper motions for the three relative younger groups (Groups 9–11), compared with the identified YSOs. For Group 9, the distribution of its residual proper motions overlaps with those of known Taurus members, suggesting that there is a kinematic relation between them. In Group 9, there are four members (Sources 290, 303, 312, 316) with RVs in Gaia DR2. The RV of Source 316 significantly deviates from the other three, and is thus not used in this work. In Figure 17 we compare the UVW velocities of the three sources in Group 9 with those of YSOs in Taurus, which indicates they are kinematically correlated. For Groups 10 and 11, the distributions of their residual proper motions are separated from the YSOs, which excludes the possibility that they have any kinematic relations with these YSOs.

Zoom In Zoom Out Reset image size