Discovery of the Closest Saturnian Irregular Moon, S/2019 S 1, and Implications for the Direct/Retrograde Satellite Ratio

The Ashton et al. (2021a) search consisted of two 1.1 square-degree Canada-France-Hawaii Telescope (CFHT) MegaCam fields on either side of Saturn, with the nearside of the field being ∼7' away from the planet. In total, 120 objects were detected moving near Saturn's on-sky rate, of which 43 have now been linked to previously known moons. The brightest detection that was not linked was internal designation e26r58a12, with w-band magnitude m_w = 25.1 and thus a rough diameter 5 km. This moon was detected (Figure 1) about 20' east of the planet in 2019 July and is about a magnitude brighter than the faintest Saturnians with IAU designations.

Zoom In Zoom Out Reset image size

In an attempt to track some the of the newly discovered moons, the offset observations were repeated in the last week of 2019 August, two months after discovery, and in two dark runs near opposition in 2020. A potential recovery of e26r58a12 was found in the images taken two months after discovery: on 2019 August 23, 26, and 27, and September 3. A preliminary orbit from the 2 months arc allowed us to track the moon to 2020 June 27/28 and July 20, where the moon had moved to the other side of Saturn. Using a dedicated search from CFHT Director's Discretionary Time, the moon was detected on 2021 June 8, 9, 13, and 14 at the predicted position of the updated orbit back on the east side of Saturn, thus confirming without a doubt the legitimacy of this moon and its orbit. Astrometry of e26r58a12 was submitted to the IAU Minor Planet Center, which confirmed the high quality of the astrometric observations (rms residual of 016) and gave the moon the temporary IAU designation S/2019 S 1 (Ashton et al. 2021b).

The fact that this brightest new irregular from the Ashton et al. (2021a) search is a direct moon with small semimajor axis gives additional support to the hypothesis made by Ashton et al. (2020) in the Jovian search: there is some level of discovery bias toward finding retrograde irregulars compared to those with direct orbits. The logical question is thus: Why was this moon not detected during the four oppositions that Subaru was observing Saturnian irregulars, when moons that are approximately a magnitude fainter were found? (We note that the moon is far too faint to have been detected in the original Gladman et al. (2001) search at any point in its orbit.) We suspect that scattered light near the planet is the likely culprit.

We somewhat arbitrarily define three zones that are within 12', from 12' to 18', and from 18' to 24' of Saturn (shown in Figure 1 in red, yellow, and green, respectively). The mosaic CCDs that were closest to Saturn in the CFHT search performed by Ashton et al. (2021a) were mostly in the yellow zone. Through the use of artificially implanted moon signals, Ashton et al. (2021a) found that these CCDs had a limiting magnitude that was 0.3–0.4 mag brighter than the CCDs farthest from Saturn. Thus, for CFHT, the yellow zone probes about 0.3–0.4 mag less faint into the irregular population. The amount the detection efficiency drops when close to Saturn is similar to what Sheppard et al. (2005) found for Uranus, although at a much larger angular separation since Saturn is significantly brighter than Uranus. The green zone likely has a a smaller (≃0.1 mag) depth degradation, while the red zone likely has much more than half a mag of depth loss. This measurement is representative but not easy to generalize to other telescope imaging cameras. This moon detectability loss depends sensitively on a telescope's baffling to minimize stray light from nearby bright sources, i.e., the host planet. Compared to other wide-field imaging telescopes we have used, CFHT has relatively good baffling, which reduces the amount of general scattered light and internal reflection artifacts from bright sources just off the mosaic camera. Other telescopes may have larger limiting magnitude losses in each zones (which in general are only roughly circularly symmetric around the bright source due to specific reflections off of telescopic support structures).

We calculated the ephemeris of S/2019 S 1 from 1990 to 2026 and determined its offset from Saturn at 1 day intervals (see top panel of Figure 2). We then calculated the fraction of time that the moon is within each of the three zones, while Saturn is within 20° of opposition (which is when detection surveys would typically be conducted). We find that S/2019 S 1 spends 12% of its time within 12', 29% within 18', and 39% within 24'. These percentages are cumulative. Depending on how strong the scattered light is in a particular telescope imaging system, these percentages indicate that just detecting the moon will be hampered a significant fraction of the time.

Zoom In Zoom Out Reset image size

An additional important consideration is that detection in a single dark run does not confirm a moving source to be a moon. Instead, the designation of a moon requires it to be tracked over multiple dark runs before it can be confirmed that an object is in orbit around a planet. Given that moons frequently plunge into glare, multidark run detection is even more difficult. As Figure 1 shows, we were fortunate to have S/2019 S 1 moving away from the planet at our initial detection in 2019 July so that our tracking data 8 weeks later still had the moon available; at a random search epoch where such a moon is far enough away to be found, the probability that two dark runs later it will also still be far enough from the planet is not large. This is difficult to quantify in general, but Figure 2 shows that it will be common that by ±2 dark runs S/2019 S 1 will often be in the planet's glare if the discovery was at opposition.

With e = 0.623, S/2019 S 1 has one of the largest eccentricities for an irregular moon, and thus would typically have a larger maximum separation from Saturn than a moon with the same semimajor axis but smaller eccentricity. Thus we decided to repeat the separation experiment with Kiviuq and Ijiraq, which have more common irregular eccentricities, of about 0.3. As can be seen in Table 2, Kiviuq and Ijiraq spend roughly the same amount of time within each zone compared with S/2019 S 1.

Spending such a large fraction of time within the glare of Saturn presents less of a problem for detecting Kiviuq and Ijiraq because they are bright and were thus easily discovered by Gladman et al.'s (2001) shallow survey. However, if these two moons were instead as faint as S/2019 S 1, then they would be just as hard to detect compared with S/2019 S 1. Again, it takes observations over at least a few dark runs to confirm the validity of a moon. Thus it is likely that a moon with an orbit similar to Kiviuq and Ijiraq and with a magnitude close to the survey limit would often not be observed at all, or not acquire multiple dark runs within a single discovery season. Kiviuq, Ijiraq, and S/2019 S 1 have the smallest semimajor axes of the known Saturnian irregulars, and are therefore the most extreme cases.

Ashton et al. (2021a) stated that S/2019 S 1 is at the completion limit for Saturnian irregular moons, since it was the brightest previously unknown moon that they found. If this is true, then even in the worst-case scenario the population of Saturnian irregulars down to this limit is unaffected by Saturn's glare, and would thus have no biases. The ratio of direct-to-retrograde for the Saturnian system down to the limit is 10:17, which strongly indicates that the imbalance in orbital direction is intrinsic. If this ratio of 10:17 continues down to the smallest known Saturnian irregular then there are 14 direct moons that are larger which have yet to be detected; this is assuming that the retrograde moons are complete down to the smallest one detected.

Since the ice giants are significantly fainter than the gas giants, the effect of the host planet on the detection bias of moons of Uranus and Neptune should be smaller, or even negligible. Neptune does appear to have a similar number of direct-to-retrograde moons, with three to four (or 2:3 if Triton and Nereid are excluded). However, there are 8 times as many retrograde Uranian irregular moons as direct (1:8) and, as such, the Uranian irregular system does not support an intrinsic equality between directs and retrogrades. Both ice giants have low numbers of irregular moons, thus better statistics are needed before concrete conclusions can be made.

Looking at the "classical" irregular moons of Jupiter (i.e., those found before the use of CCDs) the ratio of direct-to-retrograde moons is 4:4. In the case of the Saturnian irregulars discovered by Gladman et al. (2001) and Phoebe the ratio is 7:6. Thus the direct-to-retrograde ratio for "big" irregular of moons of Jupiter and Saturn is almost 1:1. This could either be a piece of evidence supporting intrinsic capture equality or that collisions are causing the discrepancy in the ratio. By the latter we mean that the "big" irregulars are original captured objects (or their largest fragments) and the smaller irregulars are collisional fragments preferentially produced for some reason on the retrograde side. There is no obvious reason collisions would favor destruction in the retrograde side, but neither does capture (e.g., Nesvorný et al. 2014). If one thinks only about Saturn and Jupiter, then the chance that the biggest recent collisions occurred on the same sense for both planet is 50% (since if both had heavily populated direct spaces, the same question would be raised). We would again like to remind the reader that the numbers of Uranus and Neptune irregulars, along with the "big" irregulars of Jupiter and Saturn, are low. Thus the above discussion is more speculation than firm conclusions.

We conclude that while observational bias is likely not the sole contributor to the imbalance between the already-known numbers of retrograde-to-direct irregular moons observed, it certainly plays some role as a factor that should be taken into consideration. Repeated deep searches close to the gas giants will likely yield more direct moons that have previously escaped detection.