β PICTORIS AND AB DORADUS MOVING GROUPS: LIKELY NEW LOW-MASS MEMBERS

We have applied our proper motion search in the μ⩾ 70 mas yr⁻¹ subsample of Tycho-2 stars to identify low-mass ABDMG candidates, adopting (U_mg, V_mg, W_mg) = (−8, −27, −14) km s⁻¹ as the mean space motion of the group, based on the known member sample of Zuckerman & Song (2004). In the initial step of the selection technique, candidates were restricted to ϕ ⩽ 12° because the known members of the group also display a range of ϕ values, with $\phi _{\rm max_{ABDMG}} \sim 12^{\circ }$. Our search for ABDMG candidates in the μ ⩾ 70 mas yr⁻¹ Tycho-2 sample resulted in 180 low-mass candidates. This candidate sample is much larger than that of the BPMG when searching in the same Tycho-2 subsample and is due to the older age of the ABDMG. Plotting the known members of the ABDMG on CMD shows that they are clearly older than the members of the BPMG. The absolute magnitudes of ABDMG members of similar SpTy at similar distances are fainter on average (see Figure 2). Because of the position of the ABDMG cluster sequence in the CMD the search algorithm is immediately prone to a higher degree of main-sequence contamination, selecting older stars whose proper motions align with the mean motion of the NYMG by chance (see Figure 2).

Zoom In Zoom Out Reset image size

While the Tycho-2 catalog has been useful in identifying large numbers of candidates of the BPMG and ABDMG, most of the candidates identified are of SpTy's earlier than M3. This bias is due to the V ∼ 12 limit of the Tycho-2 catalog. Using the pre-main-sequence (PMS) isochrones of Siess et al. (2000, hereafter S2000), a 10 Myr, 0.4 M_☉ star at a distance of 40 pc (the median distance to the BPMG) will have V ∼ 12. This means that lower mass members of the group generally will not be detected in the Tycho-2 catalog, and that proper motion catalogs with fainter magnitude limits are needed to detect the members with later SpTy.

The SBK proper motion database is a high-precision astrometric catalog generated by digitally superposing and then blinking images from different epochs of the Digitized Sky Surveys (DSSs) to measure proper motions (Lépine & Shara 2005; S. Lépine & M. Shara 2010, in preparation). SBK is statistically complete to V = 19.0, and has a faint magnitude limit of V = 21.0, allowing access to BPMG candidates with masses down to ∼0.1 M_☉. The bright magnitude limit of the catalog is V ∼ 10.0, complementing the magnitude range covered in Tycho-2. The catalog contains 1.7 million stars above the celestial equator with proper motions ⩾40 mas yr⁻¹ and proper motion errors ⩽8 mas yr⁻¹. While $\phi _{\rm max_{BPMG}}\sim 25^{\circ }$ we have restricted our search to include only those candidates with ϕ ⩽ 12.0°, in order to reduce interlopers from the extremely large initial sample; there are only two known members with larger ϕ which are outliers in the more restricted case. Our search of the SBK catalog for low-mass BPMG candidates returned 195 stars (see Figure 3). The SBK catalog BPMG search returned more candidates than the search in the Tycho-2 subsample even though SBK only covers the northern hemisphere. This is because the 40 mas yr⁻¹ proper motion limit allows us to probe a larger volume, and because the lower magnitude limit makes it significantly more sensitive to stars with lower masses.

Zoom In Zoom Out Reset image size

Table 1 lists the 19 BPMG candidates targeted for follow up in this analysis. For each candidate the columns list the catalog name, in either the Tycho-2 or SBK catalogs, the Hipparcos catalog number if available, the ICRS epoch J2000 coordinates, the proper motions from either the Tycho-2 or SBK catalog, the catalog V magnitude, the K magnitude from the Two-Micron All Sky Survey (2MASS) catalog of point sources (Skrutskie et al. 2006), the kinematic distance (d_kin; see LS09, Equation (7)), the parallactic distance (d_π) if available, the ROSAT X-ray count rate and hardness ratios if available, the Hα equivalent width (EW) if detected, and some notes on the membership of the candidate. There is some cross-contamination between BPMG and ABDMG candidate samples because the groups have very similar U and W space velocities. Table 2 lists the 41 Tycho-2 ABDMG candidates targeted for follow up in the same format as Table 1.

If a candidate actually belongs to an NYMG, then its kinematic distance, d_kin (i.e., the distance calculated assuming group membership), should be consistent with the distance measured by trigonometric parallax, d_π (see LS09). Of the 19 BPMG candidates investigated here, only one, TYC 1766-1431-1 (HIP 11152), has a trigonometric parallax measured by Hipparcos. Its kinematic and parallactic distances match very well and it is targeted for further follow up. In the ABDMG candidate sample, nine stars are Hipparcos stars. One ABDMG candidate has matching d_kin and d_π values: TYC 8513-952-3 (HIP 22738). The remaining eight have inconsistent values of d_kin: five of them have d_kin overestimating their true distances and are likely background objects moving with greater transverse velocities than the ABDMG, the remaining three have d_kin < d_π and must be foreground objects moving slower than the group. After ruling out these candidates as possible ABDMG members, we focus our follow-up observations on the 19 BPMG candidates and the remaining 33 ABDMG candidates.

Following LS09, we survey the remaining NYMG candidates for counterparts in both the ROSAT All-Sky Bright Source catalog (RASS-BSC) and the ROSAT All-Sky Faint Source catalog using 50'' as the largest accepted positional offset between candidate and X-ray source. In the BPMG candidate sample, six stars have ROSAT counterparts, including five stars drawn from the Tycho-2 sample and one drawn from SBK. Nine ABDMG candidates show positional coincidence with X-ray emitters in the ROSAT catalogs.

Of the six BPMG candidates with counterparts in the ROSAT catalogs, five have positions placing them less than 15'' from the X-ray source. TYC 1766-1431-1 (HIP 11152) has a larger positional offset. The proposed counterpart to TYC 1766-1431-1 is 235 from the star with a positional uncertainty of 17''. Inspection of the DSS and 2MASS images shows that there are several optical and IR sources near the star, but none are as close to the X-ray source position as TYC 1766-1431-1, even though it is outside of the quoted positional error. We, therefore, target TYC 1766-1431-1 for RV measurements. All nine of the ABDMG candidates are within 15'' of the ROSAT X-ray source.

Because some known NYMG members do not have ROSAT X-ray counterparts (LS09), a lack of X-ray emission is not sufficient to rule out a star as an NYMG candidate and another indicator of youth must be pursued. Hydrogen line emission may also be used as a diagnostic for youth. We observed a subsample of the low-mass NYMG candidates presented here to identify those with detectable Hα emission using the 1.5 m telescope operated by the SMARTS consortium at the Cerro Tololo Inter-American Observatory (CTIO). We used the RC Spectrograph to obtain low-resolution spectra using the same instrumental setup described in LS09. These data were obtained by service observers and processed by F. Walter following the procedures described in Walter et al. (2004).

Seventeen BPMG candidates and 37 ABDMG candidates were observed at SMARTS. Nine NYMG candidates observed exhibit Hα in emission, three in the BPMG sample, and six in the ABDMG sample. The EWs of the Hα lines in these stars were estimated using software developed in the IDL programming language; the measurements are presented in Tables 1 and 2. As it turns out, all of the stars where Hα was detected also have X-ray counterparts in ROSAT. Conversely, all the candidates with X-ray counterparts were found to have Hα in emission, except TYC 7558-655-1, for which Hα remained undetected (n.d.). The remaining 45 candidates with undetected Hα and no X-ray counterpart are ruled out as possible members of the BPMG or ABDMG, leaving 14 candidates for RV follow up (Section 3.5).

All RV measurements were obtained using the Cryogenic Echelle Spectrometer (CSHELL; Greene et al. 1993) at the NASA Infrared Telescope Facility (IRTF) 3 m telescope using the same instrumental setup as in LS09. The spectra were taken as a series of beam switched 300 s integrations for targets with H ⩽ 7.0 and 600 s integrations for targets with H>7.0 and were extracted using the procedures described in Bender & Simon (2008). RVs were measured by cross-correlation with a sample of low-mass RV standard spectra described in Mazeh et al. (2002) and Prato et al. (2002). These spectra were rotationally broadened to generate a suite of templates at different values of vsin i which can be used to estimate the projected rotational velocity of the target star. The vsin i increment of the templates is 2 km s⁻¹, which sets the uncertainty in our measured values. We used cross-correlation software kindly provided by Chad Bender (2009, private communication) which is capable of measuring velocities of unresolved companions. Stars in our PYC sample may be spectroscopic binaries and introduce contamination, thus, we test the young candidates for which we obtain RV measurements as double-lined systems. No double-lined spectroscopic binaries were identified in the PYCs that we observed.

To assess the scatter expected between the predicted radial velocities (RV_p) and the measured values (RV_m), we measured the RVs and projected rotational velocities for nine known ABDMG members. Table 3 shows that all have predicted values within 3.0 km s⁻¹ of the measurements. Similarly, LS09 observed six known members of the BPMG, finding the largest difference between RV_p and RV_m (ΔRV) to be 4.5 km s⁻¹. We will, therefore, accept ΔRV ⩽ 5 km s⁻¹ as the criterion for designating a PYC an LNM. We note that the predicted RV is calculated assuming that the candidate has the mean UVW space velocity of the group, which introduces an uncertainty in RV_p comparable to the scatter in UVW velocities of known group members. To allow for this uncertainty and the uncertainty in the measured RV, we chose a conservative cutoff for ΔRV which may introduce contaminants to the final sample but leaves less likelihood of rejecting potential bona fide members. The SpTys in Tables 3 and 4 were determined by visual comparison with the standards, and have uncertainties of ±1 spectral subclass.

Of the eight PYCs remaining in the BPMG candidate sample we measured RVs for three, including two from the Tycho-2 sample and one from the SBK sample. Table 4 lists stars for which we have obtained RV measurements plus stars for which accurate RVs were found in the literature. The table is organized into columns designating the star's catalog name, alternate name, IRCS epoch J2000 coordinates, predicted RV, measured RV, projected rotational velocity, d_kin, SpTy, and membership status. Two of the candidates observed, TYC 1766-1431-1 (HIP 11152) and PM I04439+3723W, have RV_m within 4.4 km s⁻¹ of RV_p, comparable to the observed differences in known group members, and are identified as LNMs of the BPMG. The star PM I04439+3723W has a faint common proper motion companion (named PM I04439+3723E) listed in the SBK catalog with an angular separation of ∼9''. Since the two stars appear to form a bound system, we identify PM I04439+3723E as an LNM of the BPMG as well. The star TYC 1158-1185-1N is found to have ∣RV_p − RV_m∣ = 7.5 km s⁻¹, and is thus not retained as an LNM of the BPMG. TYC 1158-1185-1N is however found to be a visual binary with the secondary (TYC 1158-1185-1S) only ∼4'' away, noticed while acquiring the target for RV observations. We measured the RVs of both stars, which differ by only ∼0.1 km s⁻¹, indicating they are probably a bound pair. TYC 1186-706-1, reported to be an LNM of the BPMG in LS09, was re-observed using CSHELL. Direct imaging data also revealed the star to be a visual binary, with an angular separation of ∼2''. We only obtained RV data for the primary but consider the newly identified secondary to be an LNM of the BPMG also.

The two remaining BPMG PYCs, TYC 1208-468-1 and TYC 7558-655-1, were not observed by us but have RVs published in the literature. The RV_m of these stars are noted with superscripts in Table 4. The RV of TYC 1208-468-1, a known visual binary, was reported by Jeffries (1995) and matches well with our predicted value. We accept both components of this system as LNM members of the BPMG. Torres et al. (2006, hereafter T06) measured the RV of TYC 7558-655-1; the RV was also reported by Zwitter et al. (2008, hereafter Z08) in their Radial Velocity (RAVE) survey second data release. The measured values are 16.3 ± 1.1 km s⁻¹ and 14.5 ± 2.4 km s⁻¹, respectively. Both of these values match very well with our RV_p. TYC 7558-655-1 is an LNM of the BPMG using our membership criteria but we note that it does not satisfy T06's criteria (see Section 4.1).

In the ABDMG PYC sample, five of the eight stars were observed using CSHELL (Table 4). We identify TYC 1741-2117-1N and S, TYC 1752-63-1, TYC 5899-26-1, and TYC 523-573-1 as LNM of the ABDMG with ∣RV_p − RV_m ∣ ⩽4.1 km s⁻¹ in each case. TYC 1741-2117-1N and S are a visual binary system with a separation of ∼2'' estimated from CSHELL direct imaging data. The similar RVs of the components suggest that they are a bound system. Our measurements rule out TYC 7002-2219-1 as an LNM of the ABDMG, showing a difference in the predicted and measured RV of ∼15 km s⁻¹.

The three remaining PYCs in the ABDMG candidate sample have RV measurements available in the literature. The RVs of TYC 8513-952-3 (HIP 22738) and TYC 9337-2134-1 were measured by T06. Torres et al. (2008) identify TYC 8513-952-3 (HIP 22738) and TYC 5899-26-1 as new members of the ABDMG. The RVs reported for the two PYCs match well with our measured (in the case of TYC 5899-26-1) and predicted velocities and we reconfirm their membership in the ABDMG. TYC 8513-952-3 (HIP 22738) is a visual binary with a separation of ∼8'' (T06). The RV of TYC 4943-192-1 is reported in Z08 as 1.4 ± 4.6 km s⁻¹, RV_m is in excellent agreement with RV_p, and we identify the star as an LNM of the ABDMG. We will follow up with our own RV measurements to improve the uncertainty of the measurement. The reported RV rules out TYC 9337-2134-1 as an LNM of the ABDMG.

TYC 1766-1431-1. We identify TYC 1766-1431-1 (HIP 11152) as an M3Ve dwarf at a predicted distance of 29.4 pc. The Hipparcos distance is 30.8 ± 3.3 pc, matching very well with our predicted distance. The star was first identified as a high proper motion M-type star (LP 353-51) by Luyten (1979) from photometric observations at the Palomar Observatory. Zickgraf et al. (2003, hereafter Z03) identified TYC 1766-1431-1 as the optical counterpart to X-ray source 1RXS J022326.7+224344.

TYC 1208-468-1 and -2. We identify TYC 1208-468-1 and TYC 1208-468-2 (BD+17 232AB) as a young, active, late-type system at a predicted distance of 52.6 pc. The separation of the components is 18. Jeffries (1995) detected Hα and Li and measured vsin i>10 km s⁻¹ in both components. Jeffries lists the primary as SpTy K3Ve and estimates a distance of 63 ± 20 pc.

TYC 7558-655-1. TYC 7558-655-1 (CD-44 753) is a young, late-type star at a predicted distance of 35.7 pc. T06 identify the star as a K5Ve PMS star and measure a 50 mÅ lithium EW. The lithium EW is small compared to that of known BPMG members of similar SpT (see T08, Figure 9) and may be the reason T06 reject the star as a possible member of the BPMG. We include it as an LNM because it satisfies the criteria we apply.

PM I04439+3723W and -E. PM I04439+3723W is an M3Ve dwarf at a predicted distance of 76.9 pc. The short rotational period, 4.3 days (Norton et al. 2007, hereafter N07), may account for the relatively large rotational velocity, vsin i ∼ 8 km s⁻¹. The star has a common proper motion companion (PM I04439+3723E) listed in the SBK catalog which has an angular separation of ∼9'', the position of the star is listed in Table 4. The companion is also listed in the NOMAD catalog (Zacharias et al. 2004). The presumed secondary is quite red, with V − K = 5.26, S2000 PMS evolutionary models predict an M5 SpTy assuming the age of the BPMG (12 Myr). If we accept the predicted distance of PM I04439+3723W to be the true distance to the system, the companion lies at a projected separation of ∼540 AU. The companion should also be considered a likely member of the BPMG.

TYC 1741-2117-1N and -S. TYC 1741-2117-1NS (BPM 84322) is a ∼2'' visual binary with K7Ve/K7Ve components at a predicted distance of 47.6 pc. At this distance, the projected separation of the system is 95 AU. Stephenson (1986a) identified the system (assuming the stars were unresolved) as a K5 dwarf with a high proper motion, consistent with our characterization. N07 identify the system as a photometric variable with a period of 3.1555 days, which is consistent with the relatively large vsin is of the components (see Table 4). Z03 identified TYC 1741-2117-1NS as the optical counterpart to X-ray source 1RXS J003408.7+252342.

TYC 1752-63-1. TYC 1752-63-1 (StKM 1-174), which we identify as a K7Ve dwarf, lies at a predicted distance of 38.5 pc. Stephenson (1986b) first identified the star as a red dwarf based on the strength of its sodium D line, classifying it as SpTy K5. This subtype is broadly consistent with our own CCD classification. The star was also identified by Z03 as the optical counterpart to X-ray source 1RXS J013723.4+265709 in the RASS-BSC. N07 identified the star as a variable with variation at the 5%–6% level in the optical regime and a period of 1.0852 days.

TYC 523-573-1. TYC 523-573-1 (BD+05 4576), which we identify as a K7Ve dwarf, is at a predicted kinematic distance of 38.5 pc. The star also appears in the Stephenson (1986a) catalog of high proper motion dwarf K and M stars. Stephenson classifies the star as a K type from visual inspection of the plate spectrogram.

TYC 4943-192-1. We identify TYC 4943-192-1 as a young, late-type star and estimate its distance to be 45.5 pc. Z08 observed the star measuring an RV of 1.4 ± 4.6 km s⁻¹, matching well with our RV_p. The star is identified as RAVE J121518.4−023728 in the RAVE second data release.

The LNMs we have identified survived our selection cuts but interlopers may still remain in the sample and comparison to group cluster sequences is useful. We have consolidated all LNM systems identified thus far and compared them with known group members in CMDs (Figure 4). The absolute K magnitude has been calculated using d_kin for both known members and LNMs. Figure 4 shows that our LNMs are consistent with the known members of the BPMG and ABDMG. We are pursuing parallax and lithium measurements for the LNMs as a definitive check of NYMG membership.

Zoom In Zoom Out Reset image size