An amateur astronomer has recovered four of five “lost” Jovian moons.


The Minor Planet Center has now published circulars announcing the four “lost” moons:
S/2003 J 23
S/2003 J 12
S/2003 J 4

S/2003 J 2
(last updated February 1, 2021)

In a first, an amateur astronomer has found four of five “lost” Jovian moons using images from a publicly available archive. The feat allows a recalculation of their orbits, leaving only one of Jupiter’s 79 known satellites still missing.

The formerly missing moons are among a group of 23 small Jovian satellites that Scott Sheppard (Carnegie Institution for Science) and colleagues reported in 2003. Many of these were later lost, though some were later recovered; as of late November, five lost moons remained. They are so faint that large telescopes can see them for only about a month every year, when Jupiter is closest to Earth. Early observations were limited, leaving their  initial orbits uncertain, too, all of which made them easy to lose as their predicted positions became ever-more inaccurate.

Orbits of Jupiter's "lost" moons
This diagram shows the initial orbits calculated for Jupiter’s “lost” moons vs. the newly calculated ones based on longer data baselines. While the differences in the calculated orbits appear small, they are significant when it comes to recovering the moons.
Kai Ly

On the Hunt for Lost Moons

The amateur, Kai Ly, found inspiration in two Minor Planet Electronic Circulars from November, which reported recovery of two previously lost Jovian moons (S/2003 J 16 in MPEC 2020 V10 and S/2003 J 9 in MPEC 2020 V19). Those reports were submitted by professional astronomers, who found the moons in images dating from 2010 through 2018.

To begin their quest to find other lost moons, Ly turned to the Canadian Astronomy Data Centre’s Solar System Object Image Search (SSOIS), where they found the best images of the small Jovian moons came from the same 3.6-meter Canada-France-Hawaii Telescope used to discover them.

“I simply search on an object's name and the page automatically displays a list of all the raw images that are supposed to contain the object,” they said. Ly started seeking the missing moons in images which covered the area where they should have been (according to their orbits) shortly before the first images used in their discoveries. Ly then used that data to extend the moons’ orbits over longer time periods. With more orbital data, they could then hunt for additional images, and so on.

With each raw image around 300 megabytes, Ly says tongue-in-cheek, “most time spent during moon hunting was simply waiting for the files to finish downloading.” But that was only the first step.

Ly lined up sequential images using the World Coordinate System to help match coordinates of reference stars. Then they spent up to several minutes blinking images, seeking objects that were moving from one frame to the next. The Aladin Sky Atlas helped Ly measure objects’ positions and movement, then they used Find_Orb software to calculate the moons’ orbits around Jupiter.

The feat would not have been possible until recent years, says Sam Deen, another amateur astronomer who helped Ly and has done recovery work of his own. “The main resource we amateurs have nowadays is the sheer amount of data from the world’s largest telescopes [and] observatories taken every night for us to hunt through,” he adds.

But while the data and analytical software are free and publicly available, that doesn’t mean it’s easy going. “The process is complicated, and the infrastructure isn’t always user-friendly,” Deen notes. Few amateurs use such resources now, but he asserts that anyone who puts in the time can do the same kind of work.

Lost Moons Found

On December 6th, Ly started looking for S/2003 J 23 because it had the least orbital uncertainty among the five missing moons. Over three days, they found additional observations between March and December in 2003 and also in February 2017.  

Animation of "lost" Jovian moon J 23
This animation blinks two 300-second exposures. The bright streak in the first exposure comes from a bright satellite, but the moon is still clearly visible.
CFHT / OSSOS / B. Gladman

Then Ly picked two targets they considered more interesting: Originally, S/2003 J 2 was thought to be the moon farthest from Jupiter and S/2003 J 12 was thought to be the innermost moon in a retrograde orbit. But the images Ly found revealed the orbits were more ordinary, putting them both in the Ananke group of retrograde moons. They needed 10 days to recover the fourth moon, S/2003 J 4. Ly has submitted these results to the Minor Planet Center for publication in their circular.

Lost moon of Jupiter J 2
Images of lost moon S/2003 J 2, which appears at magnitude 24.5. Several background galaxies are also pictured.
CFHT / OSSOS / B. Gladman

The fifth lost moon proved more difficult: Ly gave up seeking S/2003 J 10 after searching for more than a dozen days. The images they did find enabled them to plot its path over two months. That data suggests the little moon is part of the compact Carme group of irregular moons. But the uncertainty in the moon’s orbit is too large to predict where the moon is now. 

Meanwhile, unknown to Ly and most of the astronomical community, Sheppard had already recovered S/2003 J 2 and S/2003 J 23. But his submissions to the Minor Planet Center, along with Ly’s, are stuck in a processing backlog.

Faint Needles in a Giant Haystack

Jupiter moon orbits
A diagram of Jupiter's 79 known satellites. The planet’s prograde moons (purple, blue) orbit relatively close to Jupiter while its retrograde moons (red) are farther out. (One exceptions is Valetudo, in green, a prograde-moving body that's far out.)
Carnegie Inst. for Science / Roberto Molar Candanosa

“It was impressive that Kai was able to use the older observations,” says Sheppard. Besides the moons being extremely faint, he notes that the 2001 data were not as good as the 2003 images used in the moons’ discovery.

Faintness is not the only problem in tracking Jupiter’s small moons. Their orbits can extend up to 0.35 astronomical unit away from Jupiter (50 million kilometers, or about 5° in the sky), which means moons can be found across an area of about 80 square degrees. That’s a giant haystack to search for faint needles, especially when the required powerful telescopes have small fields of view.

What makes the discoveries — and recoveries — worthwhile for Sheppard is not bragging rights, but what they teach us about planetary satellite systems and the history of the solar system.

Most of Jupiter’s outer moons are small, with orbits that are retrograde (meaning they move around the planet in the opposite direction of its rotation), highly eccentric (long oval-shaped), and inclined to the plane of the solar system. The planet likely captured these moons long ago.

Most of these moons belong to one of five distinct families, each of which contains one big object and many smaller ones. The smaller objects appear to be fragments, broken off during collisions with passing objects. While collisions are rare now, the number of small objects suggest there used to be many more. Finding and tracking new moons thus helps answer questions about the history of the solar system.


You must be logged in to post a comment.