6384 KERVIN: A POSSIBLE HUNGARIA BINARY ASTEROID

Abstract

Analysis of CCD photometric observations in late 2015 of the Hungaria asteroid 6384 Kervin indicates that it may be a binary asteroid with a primary lightcurve of P₁ = 3.6194 ± 0.0001 h, A₁ = 0.06 ± 0.01 mag. The secondary lightcurve parameters are P₂ = 15.94 ± 0.01 h, A₂ = 0.03 ± 0.01 mag. No mutual events (occultations or eclipses) were observed. However, other indicators give an estimated diameter ratio on the order of Ds/Dp ~ 0.3, possibly greater.

CCD photometric observations of the Hungaria asteroid 6384 Kervin were made from 2015 Dec 16 thru 2016 Jan 13. The equipment and observing circumstances are given in Tables I and II.

Table I.

List of observers and equipment.

OBS	Telescope	Camera
Warner	0.30-m f/9.6 SCT	FLI ML-1001E
Macías	0.35-m f/10 SCT	SBIG STL-1001E

Table II.

Dates of observation for each observer. a is the solar phase angle at the earliest and latest observation. The last two columns are the average or extreme phase angle bisector longitude and latitude (see Harris et al., 1984).

Obs	2015/16 mm/dd	Sess	α	LPAB	BPAB
Warner	12/16 01/13	1–13 15	5.2 21.1	77	+1 +7
Macías	01/13	14	21.1	77	7

The asteroid had been observed at four previous apparitions by Warner (2006; 2008; 2011; 2014) as part of an on-going study of the Hungarias centering on rotational and binary statistics as well as pole axis orientations. The previous results all found a period of about 3.62 hours and no reported indications of a satellite, save in 2006 when some signs were seen but ultimately rejected.

The 2015 observations were unfiltered. Both authors used MPO Canopus to measure the images. The Comp Star Selector utility in MPO Canopus allowed choosing up to five near solar-color stars to minimize color difference problems. The raw instrumental magnitudes were referenced to V magnitudes from the APASS catalog (Henden et al., 2009) using the simple formula

$$M_t=(m_t-m_c)+M_c$$ (1)

where the subscripts t and c refer to the asteroid and a comparison star, respectively. The lower case m refers to instrumental magnitudes and the upper case M to catalog (apparent) magnitudes. For a measurement on a given image, the asteroid’s apparent magnitude was computed by finding the average of up to five values using Eq. 1 (once for each comp star). No first or second order extinction or color term corrections were applied. Ignoring the last two may have increased the standard deviation of the mean for each observation, but it was less than by using non-solar color comparison stars.

The initial observations from mid-December to mid-January by Warner found a unique period using the FALC algorithm by Harris (Harris et al., 1989) but the resulting lightcurve appeared to have a low amplitude secondary component. The dual period search function in MPO Canopus (which uses the FALC algorithm as well) led to a preliminary result of about 12 or 24 hours for the secondary period.

Since this was commensurate with an Earth day, additional observations were requested of Aznar since his data would cover parts of the long period not covered by the CS3 data. It took only one observing run from Spain to remove the ambiguities and find a period of almost 16 hours for the long period, which is also commensurate with an Earth day. The value of combining data from well-separated longitudes was demonstrated once again.

The two lightcurves show the final results. The primary lightcurve has a period of 3.6194 ± 0.0001 h and amplitude of 0.06 ± 0.01 mag. It does not have a simple bimodal shape, which is entirely possible for objects of low amplitude seen at low phase angles (see Harris et al., 2014). The secondary lightcurve has a period of 15.94 ± 0.01 h and amplitude of 0.03 ± 0.01 mag. Despite this low amplitude, the lightcurve has a well-defined bimodal shape. This is often an indicator of a nearly spheroidal satellite with its rotation period tidally-locked to the orbital period. No signs of mutual events (occultations or eclipses) were seen.

From here, we quote the analysis of Alan Harris (private communications):

… the fact that there are no mutual events implies that you are viewing from off the equatorial plane by a significant amount, hence amplitudes are muted compared to equatorial aspect. Further, you do still see curvature in the secondary lightcurve, indicating that the lightcurve of the secondary is substantial. Assuming the “undiluted” amplitude is no greater than ~1 mag, then the light of the secondary must be at least ~5% of the total light, which implies that it must be ~1/4 the diameter of the primary. It can’t plausibly be much smaller, as that would imply an even larger undiluted amplitude (recall you are viewing off the equator), so I would infer a rather largish secondary. The 16 h period corresponds to a fairly close-in binary separation (around 4 primary radii).

… the short period, with four extrema, looks like it has a significant first harmonic in the solution, and maybe even other odd harmonics. That would argue additionally for a non-equatorial aspect, although if it is only a first harmonic, that could be albedo variegation.

The terms of the fourth-order Fourier model bear this out. The magnitudes of the third order terms are greater than those for the second order while the first and fourth orders have maximum terms of almost identical magnitude.

Taking a Second Look

A check of the previous results shows that the amplitude of the primary lightcurve has ranged from 0.06 to 0.16 mag, implying that the primary also has a nearly spheroidal shape. The lowest amplitude, in 2008, was matched in 2015. The two apparitions occurred at a similar phase angle bisector longitude (L_PAB; see Harris et al., 1984). The L_PAB at other apparitions were significantly different. The 2015 results raised the question of whether or not evidence of a satellite had been overlooked in the earlier analysis.

The data from Warner for the four previous apparitions were reanalyzed using a narrow dual period search centered on the primary and secondary periods found in 2015. The 2006 data showed some indications of the secondary lightcurve, but it was less well-defined than in 2015. This may have been due in part to the data set not being as complete and so the forced secondary lightcurve was poorly or not covered over a span of about 3 hours (about 20% of the period).

We note that Warner et al. (2006) reported signs of a secondary period that were eventually rejected. The scatter (noise) in that data set was considerably greater than in 2015. The other data sets were also hampered by having fewer data points and so it was difficult to find a convincing solution for the secondary period.

Conclusion

We propose that 6384 Kervin is a binary asteroid with, possibly, a significantly-sized satellite. It appears that mutual events are not easily observed since a range of phase bisector longitudes did not reveal any events nor cause a noticeable change in the secondary lightcurve, other than to make it to disappear into the noise in the data. Some well-coordinated and prolonged campaigns in the future may find those mutual events or at least help confirm the periods reported here.

If nothing else, this case reinforces the idea that when observing a small asteroid (D < 10 km) with a period in the range of 2–4 hours, it should be followed for a number of nights, preferably including extended blocks of contiguous nights. As happened here, it took observations covering nearly a month to find a reliable solution. Keep in mind that because the range of reasonable orbital periods for a binary system includes several that are nearly commensurate with an Earth day, help from one or more additional stations may be required.