SPECKLE INTERFEROMETRY AT SOAR IN 2012 AND 2013^*

The observations reported here were obtained with the high-resolution camera (HRCam)—a fast imager designed to work at the 4.1 m SOAR telescope (Tokovinin & Cantarutti 2008). For practical reasons, the camera was mounted on the SOAR Adaptive Module (SAM; Tokovinin et al. 2008). However, the laser guide star of SAM was not used; the deformable mirror of SAM was passively flattened and the images are seeing-limited. The SAM module corrects for the atmospheric dispersion and helps to calibrate the pixel scale and orientation of HRCam, as explained below. The transmission curves of HRCam filters are given in the instrument manual.⁶ We used mostly the Strömgren y filter (543/22 nm) and the near-infrared I (774 nm) filter.

Observation of an object consists of accumulation of 400 frames of 200 × 200 pixels each with exposure time of 20 ms or shorter. Frames of 400 × 400 pixels were recorded for pairs with separation larger than 15. Each object was normally recorded twice and these two image cubes were processed independently. Parameters of resolved binary and triple systems are determined by fitting a model to the power spectrum, as explained in TMH10.

The star light reaches HRCam after reflections from two non-rigid mirrors—the thin active primary mirror of SOAR and the deformable mirror of SAM. Both can affect the plate scale. We calibrated the transfer optics of the SAM instrument by imaging a single-mode fiber located at the telescope focus (mounted on the SAM guide probe). The fiber was translated by a micrometer stage, allowing us to accurately determine the detector orientation relative to the instrument frame and the pixel size in microns projected from the detector to the SOAR focal plane. The pixel size at focus is 5.01 μm, stable throughout the runs to better than 0.5%. Considering this stability, we adopted a fixed pixel scale of 15.23 mas pixel⁻¹ as in the previous papers of this series.

The same internal calibration was applied to the CCD imager, SAMI, which is part of the SAM instrument. It has 4096 × 4112 pixels and covers a 3' square field. Thus, both HRCam and SAMI are inter-calibrated. Selected sky images taken with SAMI during the same runs were corrected for the optical distortion in SAM and calibrated against known positions of stars using the autoastrometry tool⁷ and the Two Micron All Sky Survey (2MASS) catalog. The pixel scale of SAMI and its orientation were thus determined and then translated to the orientation of HRCam. We noticed that the orientation of SAM was stable during each run to 01, but changed between the runs, although the instrument was not dismounted from the SOAR Nasmyth rotator. Therefore, the angle of the SOAR rotator was not very stable. Moreover, it should depend on the telescope orientation if the telescope pointing model is not perfect. With this caveat in mind, we estimate the accuracy of the angle calibration to be about 05.

Table 1 lists the observing runs, the angular offsets θ₀ determined for each run, and the number of objects covered. In 2013 February and September, we used the telescope time allocated for this program. The time allocation in 2013 June was lost to clouds almost entirely. The remaining data were collected during commissioning runs of the SAM instrument, as a backup program. On three of those occasions, we used only few available hours, but a whole night was spent in 2012 December.

Previous work with HRCam, SAM, and other SOAR instruments established that the telescope vibrates with the power-line frequency of 50 Hz and the rms amplitude reaches 40 mas in the worst cases. The optical axis moves on an elliptical path with variable elongation and amplitude; the amplitude increases with the zenith distance. The vibrations are not stationary. Attempts to locate their source (e.g., compressors of the closed-cycle coolers) have, so far, not produced definitive results.

During the standard exposure time of 20 ms, one full vibration period is sampled, the speckles are blurred, and the resolution and sensitivity are seriously degraded. Using a shorter exposure time of 5 ms helps to recover the signal, at the cost of reduced flux. Figure 1 illustrates the effect of shortening the exposure time by two observations of the same binary star obtained sequentially under the same conditions. The elliptical motion of SOAR's optical axis produces characteristic "disks" which sometimes blur the speckle completely.

Zoom In Zoom Out Reset image size

We used an exposure time of 5 ms and went as short as 2 ms when possible. Fainter stars were observed with the longer exposure of 20 ms, and in this case the success depended on the presence and amplitude of the vibrations. The 2013 February run was particularly affected, while in 2013 September the vibrations were less intense.

The faintest resolved binary is HIP 48273B = RAO 90 (WDS J09505+0421) at V = 12.1. It was observed in the I filter with 20 ms exposure time; the signal is affected by vibrations. The next faintest pair is KUI 41 (WDS J09310−1329) at V = 10.7, observed again in I, but with a 5 ms exposure. With a normal 20 ms exposure, stars 15 fainter could be recorded. The 2 × 2 binning of the detector helps to increase the sensitivity for faint stars while under-sampling the speckle, but it was not used here.

Table 2 lists 699 measures of 487 resolved binary stars and sub-systems, including newly resolved pairs. The format did not change from the previous papers. The columns of Table 2 contain (1) the WDS designation, (2) the "discoverer designation" as adopted in the WDS, (3) an alternative name, mostly from the Hipparcos catalog, (4) Besselian epoch of observation, (5) filter, (6) number of individual data cubes, (7 and 8) position angle θ in degrees and internal measurement error in tangential direction ρσ_θ in mas, (9 and 10) separation ρ in arcseconds and its internal error σ_ρ in mas, and (11) magnitude difference Δm. An asterisk follows if Δm and the true quadrant are determined from the resolved long-exposure image; a colon indicates that the data are noisy and Δm is likely over-estimated (see TMH10 for details). Note that in the cases of multiple stars, the positions and photometry refer to the pairings between individual stars, not with photo-centers of sub-systems.

For stars with known orbital elements, Columns 12–14 of Table 2 list the residuals to the ephemeris position and the references to the orbits.

We did not use image reconstruction and measure the position angles modulo 180°. Plausible quadrants are assigned on the basis of orbits or prior observations, but they can be changed if required by orbit calculation. For triple stars, however, both quadrants of inner and outer binaries have to be changed simultaneously; usually the slowly moving outer pair defines the quadrant of the inner sub-system without ambiguity.

Our software does not have capability of fitting models of power spectra with more than three components. Sub-systems in the resolved quadruple stars RST 244Ba,Bb (WDS J07185−5721) and MLO 3Ba,Bb (WDS J13147−6335) were measured crudely using the peaks in the autocorrelation functions.

Table 3 contains the data on 112 unresolved stars, some of which are listed as binaries in the WDS or resolved here in other filters. Columns 1–6 are the same as in Table 2, although Column 2 also includes other names for objects without discoverer designations. Columns 7 and 8 give the 5σ detection limits Δm₅ at 015 and 1'' separations determined by the procedure described in TMH10. When two or more data cubes are processed, the best detection limits are listed.

Table 3. Unresolved Stars

WDS (2000)	Discoverer	Hipparcos	Epoch	Filter	N	5σ Detection Limit
	Designation	or Other	+2000			Δm(015)	Δm(1'')
	or Other Name	Name				(mag)	(mag)
00219−2300	RST 5493 BC	HIP 1732	12.9231	y	2	4.91	6.54
00490+1656	HIP 3810	HIP 3810	13.7367	I	1	4.22	5.57
0			13.7367	y	2	4.94	6.26
01144−0755	WSI 70Aa,Ab	HIP 5799	12.9232	I	1	3.88	6.12
01496−4646	TOK 184	HIP 8498	13.7368	y	1	4.71	5.25
02128−0224	TOK 39 Aa,Ab	HIP 10305	12.9205	y	2	4.47	6.88
02136−0849	TOK 361	HIP 10365	13.7367	y	1	3.88	4.71
			13.7367	I	2	3.12	4.59
02321−1515	TOK 382 Aa,Ab	HIP 11783	13.7478	y	1	3.92	4.47

Only a portion of this table is shown here to demonstrate its form and content. Machine-readable and Virtual Observatory (VO) versions of the full table are available.

Download table as:  Machine-readable (MRT)Virtual Observatory (VOT)Typeset image

For the reader's convenience, we extracted data on the binaries resolved here for the first time into Table 4. There are 11 objects, some of which are newly resolved visual triples. As in the previous runs, we tried to observe nearby solar-type dwarfs that are known to be binary by their variable radial velocity (RV), mostly from the work by Nordström et al. (2004), or by astrometric acceleration (Makarov & Kaplan 2005). This continues the work on resolving astrometric binaries that was done at SOAR and at Gemini (Tokovinin et al. 2012, 2013). Comments on individual stars follow.

02321−1515 = HIP 11783 is an F5V dwarf at 27 pc. It has variable RV and astrometric acceleration. The companion was resolved only in the I band. The star is bright, V = 474; it was observed through clouds when no other program objects could be accessed. There is also a common proper motion companion B = HIP 11579 of spectral type K2.5V located at 345'' and 2525 from A (Tokovinin & Lépine 2012), making the whole system triple. The proper motions of A and B differ slightly, possibly because of the inner system Aa,Ab with an estimated period of ∼10 yr.

04007+2023 = HIP 18719 is a spectroscopic binary in the Hyades with period 16.7 yr, estimated semi-major axis of 012 (Griffin 2012), and astrometric acceleration. The secondary lines were detected by Bender & Simon (2008), who evaluated the mass of the secondary component at 0.4 ${\cal {M}}_{\odot }$. The object was tentatively resolved with Robo-AO (R. Riddle et al. 2014, in preparation) in 2012.757 at 342°, 0142, and Δi = 2.8. The resolution here is secure, at smaller separation and a very different position angle of 96° (or 276°) with ΔI = 3.6.

07435+0329 = HIP 37645 has a variable RV. According to D. Latham (2012, private communication), it is a single-lined binary with a period on the order of 35 yr. Such a period corresponds to a semi-major axis of 027, of the same order as the measured separation. This is a triple system with physical companion B at 96 (AB is STF 1134, discovered in 1832). The more distant companion C listed in the WDS is optical, as evidenced by the motion of AC since 1906. Although the system AC is designated as STF 1134AC, it was actually first measured in 1906 by Burnham (1913).

08391−5557 = HIP 42424 is the known binary HU 1443, measured here at 0917 separation. Its secondary component turns out to be a new pair Ba,Bb with 007 separation. This example shows the power of speckle interferometry at a 4 m telescope. The outer pair AB was observed with speckle interferomety by Horch et al. (2006) at smaller telescopes that are sufficient for its resolution, but the close sub-system escaped detection until now.

09086−2550 = HIP 44874 is similar to the previous object: the known pair RST 2610 at 178 hosts the new close 012 binary Ba,Bb. The A component is itself a close pair with variable RV and astrometric acceleration. It was already observed at SOAR and found unresolved (also unresolved here). The orbital period and mass ratio of the sub-system Aa,Ab thus remain indeterminate so far. In contrast, the new sub-system Ba,Bb has an estimated period of ∼15 yr and component masses of 0.6 and 0.5 ${\cal {M}}_{\odot }$ estimated from their luminosity. The whole system is therefore a 2+2 quadruple.

09383+0150 = TOK 358 was also resolved with Robo-AO (R. Riddle et al. 2014, in preparation) in 2013.0526 at 3451, 0447, and Δi = 2.34.

11192−1950 = RV Crt = HD 98412 is an eclipsing binary of β Lyr type, spectral type F8. The existence of the tertiary companion was suggested from the eclipse timing by T. Armond (2013, private communication). The object was observed on her request and resolved into a 005 pair of nearly equal stars.

11420−1701 = HIP 57078 unexpectedly turned out to be a triple star. This is an F5V dwarf with astrometric acceleration (Makarov & Kaplan 2005), presumably produced by the newly resolved 014 pair Aa,Ab with an estimated period of ∼20 yr. In addition, we see a distant companion B at 111. The physical nature of B is yet to be confirmed by repeated measurement within a year, but is likely, considering the low density of background stars around this target.

12509−5743 = HIP 62699 is another nearby dwarf with astrometric acceleration and variable RV resolved here at 019. The estimated period is ∼25 yr.

21368−3043 = VOU 35 AB = HIP 106701 is a G5V dwarf with Hipparcos parallax of 16.1 mas. The projected separation of AB indicates a probable orbital period of about 20 yr (see the orbit with P = 19.8 yr in Table 6, Figure 14). We detect another faint component C at 045 and 535 from A. Re-examination of archival speckle data show that the component C was not seen on 2008.54 (it was fainter than the detection limit), but was present on 2008.7669 at approximately 629 and 0475 (in the y filter). However, the next observation on 2012.923 with good signal-to-noise ratio shows no trace of the C component, while it is securely detected now (Figure 2).

Zoom In Zoom Out Reset image size

The multiple system HIP 106701 has been detected in X-rays (see also Torres et al. 2006). The disappearance of C in 2012.92 could be caused by its variability (e.g., eclipses). The projected separation of AC is ∼30 AU and implies a period on the order of 100 yr. The observed motion of AC (10° in 5 yr) does not contradict this estimate. If C were some unrelated background star, the proper motion of A (011 per yr) would have drastically changed the configuration of AC in 5 yr.

Interestingly, there is another star in the 2MASS catalog at 2340, 112 from A. Its infrared magnitudes match a dwarf of ∼0.65 ${\cal {M}}_{\odot }$ at the same distance. This additional component D is visible at similar relative position in the saturated DSS image. Considering also low crowding, we tentatively conclude that the pair AD is physical and that the whole system is at least quadruple. For completeness, we list the AD pair in Table 4.

Potentially spurious binaries. A binary system can become too close and unresolvable when it goes through periastron. It is expected to re-appear after a few years if the estimated orbital period is short. Repeated observations of several promising candidates at SOAR failed to resolve them, however, despite much improved resolution and dynamic range of speckle interferometry in comparison with the visual observations used to discover these binaries. Table 5 lists several such cases, adding to the list of mysterious "ghosts" in Tokovinin (2012). It gives the year of last measurement according to the WDS, the period of speckle non-resolution, and the number N_UR of negative speckle observations. The orbit with 21 yr period computed for B 594 by Norro (1983) predicts its separation between 007 and 012 during the period of non-resolutions at SOAR.

Reversed quadrant of the orbit is evidenced in some resolved triple systems where the orientation of the inner (fast) pair is determined by the outer (slow) binary. In these cases, reversal of the observed quadrant (allowed in classical speckle) is not possible; instead, the orbital node must be changed by 180°. Such triples here are FIN 337BC (WDS J01198−0030), FIN 308AB (WDS J10282−2548), and RBT 1 Aa,Ab (WDS J14038−6022).

02426−7947 = HIP 12654 is the acceleration binary TOK 362 resolved with NICI (Tokovinin et al. 2013). The new measure confirms the hypothesis that the pair opens up. Note that the secondary component is red: ΔV = 4.2, ΔI = 2.6, while ΔK = 0.61. The secondary could itself be a close pair of M dwarfs. The large ΔV and presumably close separation in 1991 help to understand why this pair was not resolved by Hipparcos.

04311−4522 = HIP 21079 is the acceleration binary TOK 208 resolved in Tokovinin et al. (2012). The wide separation implies a long estimated orbital period of ∼600 yr, making it difficult to explain the acceleration, unless the system is triple.

06454−3148 = HIP 32366. We discovered independently the close pair Ba,Bb, converting this nearby solar-type dwarf into a triple system. In fact, this system was resolved earlier by Ehrenreich et al. (2010). They even suggested a preliminary circular orbit of Ba,Bb with a period of 3.5 yr, based on several measurements. The small astrometric acceleration of A could be caused by its attraction to component B.

06573−4929 = RST 5253 AC. The wide companion C was discovered at SOAR on 2010.97 at 105, 2369; the quadrant published in Hartkopf et al. (2012) was chosen wrongly. It is measured here at 106 and 2386. The quadrant is now confirmed from the direct image and matches the orbit of the inner pair AB.

09383+0150 = HIP 47292 is an acceleration binary, also resolved in 2013.05 with Robo-AO (R. Riddle et al. 2014, in preparation).

14038−6022 = VOU 31 and RBT 1 Aa,Ab is the spectacular triple system β Cep.