Export citation and abstract BibTeX RIS
1. Introduction
The interstellar comet 2I/Borisov is entering the solar system on a strongly hyperbolic trajectory (e = 3.4) with perihelion on 2019 December 8.6 UT at a heliocentric distance r = 2.0 au. Since the August 30 discovery, the comet has remained at low, but steadily increasing, solar elongation in morning skies. This rendered photometric monitoring challenging due to short observational windows, bright skies, and high airmass.
The Canadian Space Agency (CSA) launched NEOSSat into a 110 minute Sun-synchronous polar orbit in 2013 February. The 15 cm telescope is baffled and designed to observe near-Sun asteroids (Hildebrand et al. 2004) down to low (≃45°) solar elongations, making it an excellent platform to monitor Borisov's perihelion approach.
2. Observations
Observations commenced September 13 UT at a cadence of 1–5 orbits day−1 to the time of writing. The spacecraft usually acquired Borisov for 1–2 arcs day−1 (an "arc" being a half-hour period when NEOSSat is on the same side of the Earth–Sun line), intending to monitor daily for bursts/splitting events and brightening while approaching perihelion.
Sets of 100 s duration images are obtained during arcs. Frames are bias and dark subtracted, cosmic-ray cleaned, astrometrically calibrated, and resampled (with flux conservation) to a common world coordinate system. There are effects related to dark current, hot pixels, and instrument drift that make it more challenging than normal to remove instrumental effects; stacking an arc's image sequence suppresses some of these issues. The stack is combined twice: a shift+co-add and a shift+median-filter, where shifting compensates for Borisov's sky-plane motion. Astrometric was sent to the Minor Planet Center with NEOSSat geocentric position the onboard GPS.
Photometry uses a 27'' radius aperture, containing essentially all the cometary light, even after correcting for geocentric distance (≃10'' in late September, Jewitt & Luu 2019). The NEOSSat instrumental photometric zero-point is set using PPM catalog stars, and was stable over the observations at 5%.
Images used for photometry are restricted to those unaffected by the South Atlantic Anomaly, bright stars, or image defects. Because NEOSSat's photometry has no filter, we approximate the R-band magnitude with R − N = 0.9, where N is the NEOSSat magnitude and the 0.9 color matches near-simultaneous (Fitzsimmons et al. 2019) R-band TRAPPIST-North observations (black dots in Figure 1).
Figure 1. Top: light curve of 2I/Borisov using NEOSSat data. The open (blue) circles give photometry from each orbital arc, while the closed (orange) points are boxcar smoothed over 5 data points (averaging out detector issues). A line is fit to the smoothed data. The comet's heliocentric/geocentric distances (au) are shown for three dates (arrows). Bottom: residuals after subtracting the linear trend from the data. A rough periodicity of 13 days is apparent.
Download figure:
Standard image High-resolution image
3. Results
Borisov has steadily brightened over two months of observation. A linear magnitude change in time satisfactorily matches the observations (Figure 1), but with a clear low-amplitude periodicity present. After removing the linear trend, a Lomb–Scargle periodogram reveals peak power with period = 13.2 ± 0.2 day. A nonlinear least squares fit to detrended data using a sine independently recovers this period and an amplitude of 0.06 ± 0.02 mag. Data were assumed to have the 0.14 mag dispersion of the unsmoothed, detrended data.
The comet has risen to magnitude R ≃ 16, the perihelion value predicted by Fitzsimmons et al. (2019). If this trend continues, by perihelion it will be ≃0.5 mag brighter than predicted. Nevertheless, this overall brightening is unremarkable.
The low-amplitude 13 day period is intriguing. Although room for doubt remains, the pattern is conspicuous. The amplitude is near the photometry's quality limit and does not perfectly model the data. We tried several methods for removing the long-term trend, but this periodicity almost always remained. The coma currently obscures the nucleus; if the periodic photometric variation is real, it is likely driven by jets distributing light in the coma. Most comets have spin periods in the 0.2–2 day range so the 13 day period is unlikely the nuclear spin; however, precessional motions are known, with 1P/Halley having 3.7 and 7.1 day periodicities, the latter associated with a non-principal axis precession (see Samarasinha et al. 2014, for a discussion). A plausible scenario is a more rapid nucleus spin, with the 13 day periodicity related to non-principal axis rotation. While the first interstellar object (1I/'Oumuamua) also wobbled (e.g., Fraser et al. 2018), in Borisov's case strong cometary jets could excite this motion as the Sun was approached.
Existing NEOSSat cadence is poorly adapted to detecting hour-scale spin periods. We expect to provide additional photometry in a few weeks.
NEOSSat data are available at https://www.asc-csa.gc.ca/eng/open-data/access-the-data.asp; data acquired via a Cycle 1 NEOSSat Guest Observer allocation (PI Gladman). We acknowledge Stefan Thorsteinson, Viqar Abbasi, and Denis Laurin for enabling the NEOSSat observations, and helpful conversations with Beatrice Meuller. We acknowledge the CRC program, NSERC funding, and the CADC/CANFAR operated by NRC-Canada with CSA support.
Facility: NEOSSat - .