Observations of Binary Stars with the Differential Speckle Survey Instrument. IX. Observations of Known and Suspected Binaries, and a Partial Survey of Be Stars

To gauge the uncertainties that should be associated with the measures in Table 5, we first study the repeatability of the results for the measures obtained in the two filters during the same observation. Differences in separation and in position angle between the two channels are shown in Figures 2(a) and (b), respectively. For position angle, an average difference of 013 ± 011 is obtained from the entire data set, with a standard deviation of 146 ± 008. However, as is well known, the position angle uncertainty grows at smaller separation owing to the fact that the same linear measurement error will subtend a larger position angle; if only separations larger than 005 are considered, the mean difference becomes 004 ± 006 with a standard deviation of 0.82 ± 0.04. In terms of the separation measures, a slight offset between the channels is noted at the 2σ level; the average difference is 0.4 ± 0.2 mas, and although small, this may indicate that a better system of scale calibration should be used in the future. The standard deviation of these differences is 2.4 ± 0.1 mas. If only separations larger than 0.05 are included, the mean remains unchanged and the standard deviation is slightly lower, 2.3 ± 0.1 mas.

Zoom In Zoom Out Reset image size

The standard deviation numbers for both position angle and separation are very similar to the previous large group of measures from the LDT reported in Horch et al. (2015). As discussed there, because these are derived from the difference between two independent measures that presumably have similar random uncertainties, we may conclude that the 1σ internal precision of the data set in Table 5 is given by these standard deviations divided by $\sqrt{2}$. Thus, subject to the caveat that the position angle uncertainty is a function of separation as discussed above, we may conclude that the internal precision of the measures in Table 5 is generally 058 ± 003 in position angle and 1.6 ± 0.1 mas in separation over the magnitude range represented by the sample.

Because we have observed a number of well-known binary stars in the data set presented, we also have the opportunity to compare to the ephemeris positions of those stars with orbits in the literature (excluding, of course, those objects used in the determination of the scale and orientation for each run). A list of such objects with orbits of Grade 2 or better in the Sixth Catalog of Orbits of Visual Binary Stars (Hartkopf et al. 2001a; hereafter Sixth Orbit Catalog¹⁶ ) is shown in Table 6. Also included there are the ephemeris position angles and separations, with uncertainties as calculated from the uncertainties in the orbital elements, and the observed minus ephemeris residuals when comparing to the observed values in Table 5. In Figure 3, we plot these results. To make the comparison, we have averaged the astrometric results in both channels of the instrument, whereupon the precision obtained should be that of a single measure divided by $\sqrt{2}$. As determined by the comparison between the channels given above, the precision of a single measure is 1.2 mas in separation and 042 in position angle. We show these values in Figure 3, using $\tan (1.2/\rho )$ (where ρ is in mas) to convert the linear measurement precision into an angular one; for the mean separation of this group of objects, that results in a median angular precision of 020. Together, the plots indicate that there are no obvious offsets in position angle or separation and that the accuracy of these measures relative to the orbital ephemerides is comparable to the internal precision.

Zoom In Zoom Out Reset image size

Table 6.  Ephemeris Positions and Residuals Used in the Astrometric Accuracy Study

WDS	Discoverer	HIP	Date	${\theta }_{\mathrm{eph}}$	${\rho }_{\mathrm{eph}}$	${\rm{\Delta }}\bar{\theta }$	${\rm{\Delta }}\bar{\rho }$	WDS Orbit Grade
	Designation		($2000+$)	(deg)	(arcsec)	(deg)	(mas)	and References
07480 + 6018	HU 1247	38052	15.1848	3.3 ± 1.5	0.1654 ± 0.0018	−2.2	−1.0	2, Hartkopf et al. (1996)
07518 – 1354	BU 101	38382	15.1849	294.5 ± 0.1	0.5554 ± 0.0027	+0.1	+3.8	1, Tokovinin (2012)
07528 – 0526	FIN 325	38474	15.1849	182.0 ± 0.3	0.3506 ± 0.0036	−0.3	+2.4	2, Hartkopf et al. (1996)
09036 + 4709	A 1585	44471	15.1825	286.5 ± 0.2	0.2830 ± 0.0004	+0.1	+0.3	2, Muterspaugh et al. (2010)
10083 + 3136	KUI 48	49658	15.1827	169.8 ± 0.2	0.2171 ± 0.0009	+0.5	−3.1	2, Hartkopf et al. (1996)
12060 + 6842	STF 3123	59017	15.1857	197.1 ± 3.7	0.2991 ± 0.0072	−0.8	−10.9	2, Hartkopf et al. (1996)
12417 – 0127	STF 1670AB	61941	15.1829	5.6 ± 0.6	2.2955 ± 0.0055	−0.1	−23.8	2, Scardia et al. (2007)
12417 – 0127	STF 1670AB	61941	16.0720	3.7 ± 0.5	2.4245 ± 0.0058	−0.3	−10.7	2, Scardia et al. (2007)
13198 + 4747	HU 644AB	65026	15.1828	86.1 ± 0.4	0.7996 ± 0.0301	+0.1	+36.5	2, Hartkopf & Mason (2015)
13203 + 1746	A 2166	65069	15.1829	6.8 ± 5.3	0.1313 ± 0.0292	−1.3	+4.5	2, Zasche & Uhlar (2010)

Download table as:  ASCIITypeset image

As a further check, we also found that six of our systems with separations greater than 1'' had astrometry for both components in the Gaia DR2 (Gaia Collaboration 2018). In comparing our measures to the position angles and separations implied by Gaia, we find that the average position angle difference was 044 ± 052 and the average separation difference was 8.3 ± 16.7 mas. Although the sample is small and there may be some relative motion of the system in between the mean Gaia epoch of observation and our measures, this nonetheless gives some further confidence that the astrometry has been properly calibrated.

We turn next to the photometric precision and accuracy of the measures in Table 5. This is harder to judge than in the case of relative astrometry because the best space-based measures of magnitude differences for the objects in Table 5 in the literature are usually still those obtained by Hipparcos; the DR2 of Gaia does not contain much information for the objects we have observed owing to their mostly subarcsecond separations. The Hipparcos results are in the so-called H_p filter, which has a center wavelength of 511 nm and a width ${\rm{\Delta }}\lambda$ of 800 nm. This is not a particularly good match for the results in Table 5, which are considerably redder, and with narrower bandpasses.

Despite this deficiency, we nonetheless show in Figure 4 a comparison between the ${\rm{\Delta }}{H}_{p}$ value appearing in the Hipparcos Catalogue and the magnitude differences appearing for each filter in Table 5. To make the comparison as valid as possible, we have limited the range of B − V color for the systems we have plotted to $0.0\lt B-V\lt 0.6$ and excluded systems that appear to be giants based on their placement in the color–magnitude diagram. The horizontal axis of Figure 4(a) is the seeing estimate for the observation multiplied by the separation of the components; in previous papers, we have argued that this combination is a measure of the "isoplanicity" of the observation. That is, it is related to the degree to which the primary star and secondary star will have similar speckle patterns. This is because the seeing is inversely proportional to the Fried parameter, while the size of the isoplanatic angle is linearly proportional to it. In dividing the separation by the size of the isoplanatic angle, one obtains a ratio q, with $q\lt 1$ if the secondary falls inside the isoplanatic region of the primary and $q\gt 1$ if it does not. By multiplying seeing times separation, a parameter ${q}^{{\prime} }$ carrying units of arcseconds squared is instead obtained, but the value is proportional to q. Therefore, one expects that there will be a particular value below which the observation is isoplanatic, whereas above it the magnitude difference obtained will be affected by the decreased correlation between primary and secondary speckle patterns.

Zoom In Zoom Out Reset image size

Figure 4(a) shows a result similar to previous papers in this series; the difference between space-based relative photometry of the pairs and the speckle result clusters near zero if the seeing times separation is less than approximately 0.6 arcsec², but it trends upward for larger values of this parameter. This is understandable given that, as the value increases and therefore the speckle patterns of the primary and secondary are less and less identical, the derived magnitude difference from the speckle observation is larger and larger as the speckles fail to correlate at the same vector separation in the autocorrelation. Because of this, for observations shown in Table 5 with ${q}^{{\prime} }\gt 0.6$ arcsec², the magnitude difference is shown as less than the value obtained in the fit. In Figure 4(b), we show a plot of the ${\rm{\Delta }}{H}_{p}$ values as a function of the magnitude difference at 692 nm for observations with seeing times separation of less than 0.6 arcsec². Here the data are essentially linear, with standard deviation of 0.21 mag if we consider $0.1\lt {\rm{\Delta }}{H}_{p}\lt 4;$ some of this scatter is due to the uncertainty in the Hipparcos values themselves; if we subtract the average value of the stated uncertainty in ${\rm{\Delta }}{H}_{p}$ in quadrature, which is 0.16 mag, then the uncertainty left over and presumably due to the speckle measurement is 0.13 mag. This is slightly larger than in some previous papers in this series, particularly measures taken at the WIYN telescope, and if the program of speckle observations continues at the LDT, then this will have to be studied in more detail. Perhaps further data releases of the Gaia Collaboration could help resolve this issue.

The Hipparcos suspected doubles and the Be stars observed in the data set presented here were not known to have companions. Likewise, for the double-lined spectroscopic binaries observed from the Geneva–Copenhagen Catalog, it was not known if the resolution limit would permit the detection of the secondary at the LDT. In general, we search for companions visually using reconstructed images made from the data. If none are clearly identified, then we use the following methodology to establish the detection limit as a function of separation from the primary star. We identify pixels in annuli centered on the primary star with center width in increments of 01. We find local maxima within each annulus and compute the average and standard deviation of the peak values of these features. The detection limit for that annulus is then the average peak value plus five times the standard deviation, converted to a magnitude difference. For the purpose of the robust discovery of new companions, we require conservatively that the detection limit at the diffraction limit corresponds to a zero magnitude difference. Once values of the detection limit are obtained for the entire set of annuli, a cubic spline routine is used to interpolate between these values in order to obtain a continuous curve. Examples of curves obtained for both an unresolved star and a companion previously unknown but appearing in Table 5 are shown in Figure 5.

Zoom In Zoom Out Reset image size

In all, 76 stars in the data set were observed in good conditions and were found to be unresolved, to the limit of detection of the DSSI camera at the LDT. These are listed in Table 7. We provide there an estimate of the magnitude difference that represents a 5σ detection at both 02 and 10 from the primary, in both the 692 and 880 nm filters. The detection limit curve may be roughly reconstructed from these numbers, as it is usually fairly linear between 02 and 10, and again roughly linear (though with different slope) between the diffraction limit and 02.

Table 7.  5σ Detection Limits for High-quality Nondetections

(α, δ J2000.0)	Hipparcos	Date	5σ Det. Lim., 692 nm		5σ Det. Lim., 880 nm		List, Notes
(WDS Format)	Number	(Bess. Yr.)	02	10	02	10
00064 + 6412	531	15.8338	3.80	5.96	3.76	6.23	Be star (10 Cas)
00447 + 4817	3504	15.8338	3.77	5.59	3.83	6.01	Be star (o Cas)
00567 + 6043	4427	15.8338	3.66	5.66	4.38	6.12	Be star (γ Cas)
01557 + 5916	8980	15.8340	4.39	6.14	4.20	6.38	Be star (V777 Cas)
03365 + 4812	16826	15.1846	4.23	6.34	4.27	5.99	Be star (ψ Per)
03365 + 4812	16826	15.8342	4.35	7.00	3.64	7.06	Be star (ψ Per)
03423 + 1942	17309	15.8342	4.79	7.21	4.38	7.49	Be star (13 Tau)
03449 + 2407	17499	15.8342	4.24	7.49	4.22	7.80	Be star (17 Tau)
03463 + 2357	17608	15.8343	4.33	7.22	4.12	7.65	Be star (23 Tau)
03475 + 2406	17702	15.8343	4.31	7.23	4.38	7.66	Be star (η Tau)

Only a portion of this table is shown here to demonstrate its form and content. A machine-readable version of the full table is available.

Download table as:  DataTypeset image

We give some further information regarding the 15 new discoveries here; we take basic stellar data for each system from SIMBAD (Wenger et al. 2000).¹⁷

LSC 116Aa, Ab = HIP 32887. Our discovery of a small-separation component of this previously known binary MLR 688 reveals this system to be a hierarchical triple system. The composite spectral type is F8, and the system is at a distance of approximately 114 pc based on the Hipparcos parallax result. (No Gaia result is available at present.) If the new component detected here is indeed a physical member of the system, the current projected separation is 7 au. If this is comparable to the semimajor axis, the period would likely be in the range of 10–15 yr.

LSC 117 = HIP 35775. The component of the F7V system measured here may well be the spectroscopic component listed in the Geneva–Copenhagen Catalog, given the small separation and the Gaia DR2 distance of 73 pc (Gaia Collaboration 2018). Those data suggest a projected physical separation of 1.8 au and a period on the order of 1 yr. The mass ratio is given in the Geneva–Copenhagen Catalog as 0.800 ± 0.05, and the iron abundance is near solar, which is consistent with the relative photometry we have determined here for a mid-F primary and later-F secondary star.

LSC 118Aa,Ab = HIP 37657. This star has a composite F5 spectrum according to information in SIMBAD and was also discovered to be binary by the Hipparcos satellite, where it is listed in the Hipparcos Catalogue as HDS 1092. We detect a small-separation companion to the primary star of that system. However, the Gaia DR2 parallax is only 2.65 mas, and the apparent V magnitude is 8.79; thus, the primary star would appear to be evolved, as this implies an absolute magnitude of $+0.91$, which is too bright for an F5 dwarf. This comports with the magnitude differences for both the close and wider companions of over 2 mag. Those stars may yet be on or near the main sequence.

LSC 119 = HIP 43519. The primary of this pair appears to be a K0 giant, given the distance (over 300 pc according to the DR2) and apparent V magnitude of 7.872. The component we detect is over 4 mag fainter and at a separation of 03, leading to a projected physical separation of about 100 au. Thus, it is almost certainly not the known companion OCC 382. This could be an optical double.

LSC 120 = HIP 43810. Suspected as a binary by Hipparcos, the primary here is again an early K giant, but this system lies much closer to the solar system at only 135 pc. The separation of the component we detect is large (07) and the magnitude difference is again over 4 mag, so again the physical association cannot be plausibly established at this point. The system was seen as unresolved by Mason et al. (1999), probably due to the large magnitude difference.

LSC 121 = HIP 44908. An F5 pair at a distance of 150 pc, this system has a small magnitude difference. If we assume two F5V stars with a semimajor axis of 15 au separation (which is the current projected physical separation given the angular separation of 01), that would imply an orbital period of less than 40 yr, so follow-up observations of the system would be helpful in the coming years to confirm orbital motion.

LSC 122Ba,Bb = HIP 47196. The secondary of the known binary STF 1372 is revealed here to be a small-separation binary itself; this is the only example in our list of new discoveries of an A-BC architecture among the five triples. This small-separation component may be the spectroscopic one mentioned in the Geneva–Copenhagen Catalog. The mass ratio given there is 0.549, but without an uncertainty, and the [Fe/H] value is listed as −0.08. There is no DR2 parallax, but the revised Hipparcos value is 6.80 ± 0.71 mas (van Leeuwen 2007), indicating a projected physical separation of 1.8 au, and if we use this as an estimate of the semimajor axis, the period would be on the order of 2 yr. However, the magnitude difference between the Ba and Bb components would be on the order of a magnitude given the data in Table 5, and that would most likely give a mass ratio for an early G and late G dwarf of closer to 0.8. An orbit due to Alzner (2005) exists for the AB system; a reanalysis of earlier speckle observations that already appear in the Fourth Interferometric Catalog might yield more astrometry of this pair, and if so, a combined orbit for both the wide- and small-separation pairs could be within reach very soon.

LSC 123 = HIP 48044. This star, suspected of having a companion by Hipparcos, is at a distance of 200 pc, so that the current projected separation is then approximately 18 au. The spectral type is listed in SIMBAD as A5.

LSC 124Aa,Ab = HIP 49495. A subarcsecond component to this G5 system has been known for decades. We detect that companion, now at 035 from the primary, but also find a very small separation component. The DR2 does not give a parallax, but the Hipparcos revised value is 4.69 ± 0.69 mas; thus, the projected separation is about 5.5 au.

LSC 125 = HIP 49677. This F0 star was suspected double by the Hipparcos satellite. We find a component with magnitude difference of about 2.5 at a separation of 038. The Gaia DR2 parallax is 5.4510 ± 0.6343 mas. The system was reported as unresolved by Mason et al. (1999).

LSC 126 = HIP 52932. This star is another Hipparcos suspected double that is resolved here for the first time, although the separation is over 1''. The magnitude difference is large, and this is the reason the system was not discovered and measured by the satellite. The distance is approximately 112 pc.

LSC 127Aa,Ab = HIP 53709. This star is listed as a double-lined spectroscopic binary star in the Geneva–Copenhagen Catalog; the mass ratio there is listed as 0.982 (without an uncertainty). A 3'' companion was already known and was outside our field of view. However, it is unlikely to be the spectroscopic component, given the distance of 68 pc; this translates into a projected separation for that component of 200 au. In contrast, the small-separation component we find has projected separation of approximately 1.7 au. Given the primary spectral type of F8V and a magnitude difference of about 1.5, the secondary, if bound, is likely to be in the late G range.

LSC 128 = HIP 53969. We confirm a faint component at a separation of 033 to this Hipparcos suspected double star. The distance is approximately 130 pc; if bound to the primary, the projected separation is 43 au.

LSC 129 = HIP 94068. This is the first of two serendipitous discoveries where we observed the star as a bright unresolved calibration star, in this case, HR 7266. The companion has a modest magnitude difference and is at a separation of 007. Given the relatively small distance to the system of 43.5 pc, this should be an interesting pair to follow up on in the coming years. Data in SIMBAD suggest that the primary star is an evolved star of spectral type F0.

LSC 130 = HIP 105966. This is another serendipitous discovery, but in this case the companion found is faint and at a separation of 036. The primary star has spectral type A1V, and the secondary would appear to be much fainter and redder; the photometry in Table 5 would suggest perhaps something in the late G or early K range. Two previous speckle observations, by McAlister et al. (1987) and DeRosa et al. (2011), did not detect the secondary.

Objects for which we calculate classical visual orbits are shown in Table 8. The method used is that of MacKnight & Horch (2004), which is a two-step process to deriving final orbital elements. After obtaining the previously available astrometry of the system from the Fourth Interferometric Catalog, a grid search is performed to identify the orbital elements that minimize the reduced ${\chi }^{2}$ when comparing ephemeris position to the actual measures. The second step is to use a downhill simplex algorithm to refine those orbital elements to the absolute minimum reduced ${\chi }^{2}$. Uncertainties in the orbital elements are estimated by adding random deviates in ρ and θ of a typical value for large-telescope speckle observations (2.5 mas) and recomputing the orbit. This is done many times and gives a distribution for each orbital element. The uncertainties of each are estimated to be the standard deviation of its distribution. The orbits are shown in Figure 6, and comments on each system follow. Further information regarding the calculation of visual orbits is given in Horch (2013) and references therein.

Zoom In Zoom Out Reset image size

BU 314AB. This mid-F star sits at a distance of 40 pc and had a previous Grade 3 orbit in the Sixth Orbit Catalog due to Söderhjelm (1999). However, since the calculation of that orbit, there has been a sequence of very good quality speckle observations leading up to the present that permitted a redetermination of the orbital elements. This was needed here, because the system was pressed into service as a scale calibration object for the 2016 February run owing to the lack of other suitable observations. In combination with the Hipparcos revised parallax (no Gaia result is available), our orbit gives a total mass of $1.85\pm 0.14\,{M}_{\odot }$.

HDS 1199. The spectral type for this system is listed in SIMBAD as K7V+M0/2V, and the Gaia DR2 parallax is 26.4436 ± 0.5831 mas. No previous orbit exists in the Sixth Orbit Catalog. Using the observations presented in Table 5, together with those in the literature, the data so far span approximately 25 yr of the 133 yr period and trace out approximately one-quarter of the orbit in terms of the position angle coverage, but nonetheless we determine a semimajor axis with only 3% uncertainty. This, together with the parallax and period, yields a mass sum of 2.53 ± 0.38 M_⊙, much higher than expected for two late-type dwarfs. However, Parihar et al. (2009) find that the star is probably an eclipsing binary, and as our orbit does not eclipse, a third star must be present in the system. Parihar et al. (2009) also find that the spectrum of the system exhibits Hα emission. We also note that the magnitude differences for the pair measured so far are in the range of about 2.5, again much larger than expected for a late K plus early M star.

COU 1258. Ours is the first attempt to calculate the orbit of this system, which is of mid-F spectral type and at a distance of approximately 150 pc. The mass sum we derive at this stage is $2.67\pm 1.51\,{M}_{\odot }$, which is very uncertain but nonetheless consistent with this spectral type and the small magnitude difference that we have measured. The system was reported as unresolved in a 2008 observation by Gili & Prieur (2012); our orbit predicts a separation of 94 mas at the time of their observation. That observation was done with a 0.7 m telescope using a 570 nm filter, so the system would have been below the diffraction limit of 021.

HDS 1542AB. SIMBAD lists the spectral type of this system as M1V, and we measure a modest magnitude difference. The distance is about 40 pc; thus, we derive a mass of ∼0.7 M_⊙ with large uncertainty, consistent at this early stage with the spectral information. If, on the other hand, one uses the Gaia DR2 parallax for the primary (no value exists in the data release for the secondary), then the distance is 31 pc, and the mass obtained is $0.27\pm 0.07\,{M}_{\odot }$, which is much lower than one would expect for an M1V pair.

YSC 156. An F3V star with small magnitude difference, the expected mass for this system would be about 3 M_⊙. We derive $6.8\pm 4.2\,{M}_{\odot }$, so the orbit we present is not particularly useful in terms of stellar astrophysics, but nonetheless it should provide reasonable ephemerides for the coming few years.

WSI 65 = 66 Oph. This is a well-known Be star, and it is interesting to note that the orbit appears to be somewhat eccentric. We obtain a total mass of $10.4\pm 3.8\,{M}_{\odot }$. Given the spectral type of B2Ve and a ${\rm{\Delta }}m$ of ∼3, we would expect that the secondary is an early A star; this combination would have a mass of ∼14 M_⊙, consistent with what we derive from the astrometry. The Hipparcos revised parallax is 5.01 ± 0.26 mas, so that the semimajor axis is 34.9 ± 3.8 au. We note that Draper et al. (2014) show that this star's V-band polarization has been declining over the past 30 years; the last time of periastron passage was in 2003.1 according to our orbital elements, so this has been occurring when the secondary is closest to the primary, suggestive of an effect by the secondary on the disk thought to surround the primary.

The number of velocities from our spectroscopic observations of HD 22451 = YSC 127 and HD 185501 = YSC 135 and their orbital phase coverage are sufficient to obtain spectroscopic orbital elements. We first determined preliminary elements of the components of the two systems with the program BISP (Wolfe et al. 1967), which uses the Wilsing–Russell Fourier analysis method (Wilsing 1893; Russell 1902). We refined those elements with the differential corrections program SB1 (Barker et al. 1967). We compared the variances of the solutions for the primary and secondary of YSC 127. The solution for the primary velocities has the smallest variance, so we assigned unit weights to those velocities. We then set the weights of the secondary velocities to be the inverse of the ratio of the variances from the primary and secondary solutions. Thus, for YSC 127 we assigned weights of 1.0 to the primary velocities and 0.6 to those of the secondary except for the KPNO observation of JD 2456868, which had very blended components measured as a single velocity and so was given a weight of zero. The two components of YSC 135 have lines of very similar width and depth, resulting in nearly identical variances for their solutions, and so all velocities for that system were given unit weights. To obtain a simultaneous solution of the components of the two systems, we adopted the above weights and used the program SB2, which is a slightly modified version of SB1. The spectroscopic orbital elements for both systems are listed in Table 9. The resulting spectroscopic orbit for YSC 127 is shown in Figure 7, and that for YSC 135 is displayed in Figure 8.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Table 9.  Orbital Elements of HD 22451 = YSC 127 and HD 185501 = YSC 135

Parameter	HD 22451	HD 22451	HD 185501	HD 185501
	SB2	Joint VB+SB2 Fit	SB2	Joint VB+SB2 Fit
P (days)	2347.7 ± 6.3	2401.1 ± 3.2	434.57 ± 0.16	433.94 ± 0.15
T (HJD)	2458283.0 ± 1.6	2458281.9 ± 1.5	2457132.14 ± 0.78	2457134.26 ± 0.62
e	0.5753 ± 0.0029	0.5804 ± 0.0030	0.2127 ± 0.0020	0.2155 ± 0.0023
a (mas)	⋯	42.0 ± 2.2	⋯	41.04 ± 0.88
i (deg)	⋯	111.29 ± 0.93	⋯	116.40 ± 0.93
Ω (deg)	⋯	3.20 ± 0.82	⋯	337.59 ± 0.47
${\omega }_{A}$ (deg)	110.00 ± 0.33	109.51 ± 0.33	81.10 ± 0.69	82.92 ± 0.56
KA (km s−1)	12.001 ± 0.047	⋯	15.422 ± 0.044	⋯
KB (km s−1)	13.003 ± 0.053	⋯	15.801 ± 0.044	⋯
γ (km s−1)	−19.225 ± 0.027	⋯	-31.948 ± 0.025	⋯
${M}_{\mathrm{tot}}$ (M⊙)	⋯	2.579 ± 0.054	⋯	1.775 ± 0.023
${M}_{A}$ (M⊙)	⋯	1.342 ± 0.029	⋯	0.898 ± 0.012
${M}_{B}$ (M⊙)	⋯	1.236 ± 0.026	⋯	0.876 ± 0.012
Distance (pc)	⋯	114.6 ± 6.3	⋯	33.09 ± 0.74

Download table as:  ASCIITypeset image

We next used our spectroscopic orbital elements as the starting point for combined spectroscopic–visual orbits of the two systems, where the relative astrometry is entirely from our own speckle program at WIYN, Gemini, and the LDT. In this case the same method as in Muterspaugh et al. (2010) was used to do the simultaneous fitting. Results of these calculations are also shown in Table 9, including the individual masses of the stars and the independently determined distance to the system, which can be obtained when both astrometric and double-lined spectroscopic data exist. Figures 9 and 10 show the visual orbits for YSC 127 and YSC 135, which have periods of 2401 and 433.9 days, respectively. The combined fit shows that for YSC 135, the standard deviation of separation residuals is 1.9 mas, comparable to the value estimated for the speckle observations in Section 4.1. However, in the case of YSC 127, the standard deviation in ρ is much higher, 5.4 mas. This would appear to be due primarily to the point included at a position angle of 325° in 2011; these were extremely challenging measurements made with the WIYN telescope at a separation of approximately one-quarter of the diffraction limit and show that the separation was probably overestimated in this case. Otherwise, the residuals for this system appear to be on the same level as for YSC 135.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

The distance obtained from the orbit calculation for YSC 127 is 114.6 ± 6.3 pc, whereas the value implied by the Gaia DR2 parallax is 126.2 ± 0.7 pc. Thus, the 1σ error bars of the two measurements do not overlap at this stage. We attempted to compute an orbit without the 2011 data points, and this resulted in a semimajor axis that was lower than the value shown in Table 9 by 6%. This translates into a distance that is 6% larger than our calculated value, or 121.9 pc. This would reconcile the distances given our level of uncertainty, but it will nonetheless be important to see whether the Gaia value is revised downward at all in future data releases, or, if further astrometric data can be obtained in the coming years, whether a subsequent orbit calculation might confirm an increased distance from what we have obtained. In any case, the masses of 1.34 and 1.24 M_⊙ implied by our orbit agree well with those expected for an F6V+F7V pair, as do the magnitude difference and absolute system magnitude. According to the Geneva–Copenhagen Catalog (Nordström et al. 2004), the system has near-solar metallicity.

For YSC 135, a G5 pair with [Fe/H] = −0.28, the distances agree well between the calculation here and DR2; the former is 33.09 ± 0.74 pc, while the latter is 32.58 ± 0.03 pc. The individual masses are 0.90 and 0.88 M_⊙ and are certainly roughly in line with stars of this spectral type, if slightly low. This may be due to a combination of two factors. As shown in Horch et al. (2019), one would expect the value to be low compared to an equivalent binary of solar metallicity by 5%–10% at [Fe/H] = −0.3. On the other hand, the B − V color of the pair is shown to be 0.77 in SIMBAD, which is redder than expected for a G5 pair.

Torres et al. (2010) compiled a list of eclipsing binary systems with masses and radii known to 3% or better. They also included 23 systems with accurate interferometric and spectroscopic orbits that resulted in component masses determined to 3% or better. Our results for YSC 127 and YSC 135 indicate that the stars of these two systems can now be added to the latter list.