As the Voyager probes moved through the outer solar system, they accumulated a huge record of discoveries. Among the newly found objects and phenomena was a large collection of small moons orbiting Jupiter, Saturn, Uranus and Neptune. Most of these were beyond the ability of Earth-based hardware to visualize at the time — we actually had to be there to see them.
Since then, however, improvements in ground optics and the existence of the Hubble Space Telescope have allowed us to find a few small bodies missed by Voyagers, as well as other small objects elsewhere in the solar system, such as the Kuiper Belt. object recently visited by New Horizons. Now researchers have found a way to use advances in computation to do even more with imaging, by spotting a small new moon near Neptune and possibly seeing another one for the first time since Voyager 2.
Since Neptune was visited by Voyager 2 and has been imaged frequently since then, moons we haven’t seen yet will be quite difficult to see, presumably because they are a combination of small and/or dim. The easiest way to see them is to increase the exposure time so that there is a greater chance of weak signals from the noise. This method won’t work if there’s a bright object nearby, which isn’t much of a problem with the outer planets.
What is a problem is the previously discovered moons orbiting close enough to the moons to move noticeably within the span of a single exposure. At some point, this movement creates a “smear” that keeps the signal firmly anchored in the noise.
A small team of researchers from SETI, NASA and Berkeley has devised a way to compensate for this smear. If you know the trajectory of a body, you can predict how much it will move from one image to another. You can then adjust multiple consecutive images to place objects in the same orbit in the same location, making any cues stand out. The transformation can get tricky because the track may be at an odd angle from the imaging device. But it is well within our current computing capabilities.
The problem is that if we don’t know the moon exists, then of course we don’t know its orbit. But the team behind this new research has come up with features that work for any moon moving in a forward, circular orbit around a planet’s equator.
“You can use these features to transform a photo taken when t0 to match the appearance of another image acquired at the time t1 by moving each pixel in the original image to a new location,” the team members write. “After the transformation, each moon will appear on a prograde, circular, equatorial orbit at fixed pixel coordinates.”
The researchers applied this method to Neptune, where Voyager 2 had discovered a series of small moons, using a series of images from an entire orbit from Hubble.
Discovery and rediscovery
When the analysis was done, a small moon emerged. Now named Hippocamp after the mythical sea monster, it is only about 22 miles in diameter and orbits near the outer edge of this cluster of moons. The same analysis picked up the moon Naiad, which had been spotted by Voyager 2. But Naiad was not where we expected it to be; in fact it was 180° off. But an orbit that puts it in the right location is possible with a slight adjustment to the orbit predicted from Voyager data. Unfortunately, that means some claims that they’ve been spotted with ground instruments in recent years are almost certainly false.
With three observations made up to a dozen years apart, it’s possible to calculate Hippocamp’s orbit well enough to determine where it would have been during Voyager 2’s flight. Most images missed that part of space, and the few images that did not all show significant stain. So there was no real way to identify this moon prior to recent imaging campaigns.
Hippocamp’s orbit is just inside that of Proteus, the largest of Neptune’s inner moons. Tidal forces cause Proteus to slowly move away from Neptune, but Hippocamp is so small that these forces have minimal effect on it. And that means the two bodies were once much, much closer. Proteus has a large crater, Pharos, where an impact blasted out 50 times the volume of Hippocamp, suggesting there was more than enough debris nearby to form the small moon.
One of the problems with this idea is that Proteus’ proximity should have thrown Hippocamp into an eccentric orbit. But the researchers suggest this isn’t as much of a problem as it seems. Based on the crater speeds and sizes on Proteus, they estimate that collisions large enough to completely wipe out Hippocamp would have happened at least nine times since its formation. The process of destruction and reform should allow the moon to gradually take on more regular orbits.
However, that model suggests that it and possibly other small moons near Neptune have experienced multiple destructions over the course of their history rather than simply being placed during Neptune’s formation. And there’s a lot of unknown material, suggesting there are some thin rings in the region that would be hard to detect without another visit to the blue planet.
Nature2019. DOI: 10.1038/s41586-019-0909-9 (About DOIs).