As the New Horizons science team gathered in Laurel, Maryland, during the last week of 2018, many of us reflected back on a similar gathering three-and-a-half years earlier. In the summer of 2015, we all came together on a voyage of exploration to the planet Pluto and its moons. At that time, we didn’t know very much about our target, and this time we knew even less about “Ultima Thule,” a small, lonely 30-kilometer-wide object known as 2014 MU₆₉ that was orbiting the Sun almost half a billion kilometers beyond Pluto. Unlike what we knew about Pluto, we did not even know Ultima Thule’s surface composition, its shape, or even the length of its day.

We knew Ultima Thule was much smaller than — and hence would be a very different beast from — Pluto. So as the New Horizons team arrived at the Applied Physics Laboratory (APL) in Laurel, Maryland, to watch and enjoy the encounter sequence play out, we were completely in the dark. The only clues we had were stellar occultation observations of Ultima Thule from Argentina in 2017 that indicated it was elongate and might have either two separate or contacting lobes. The lack of any variability in brightness as it rotated hinted that we might be looking at one of its poles, and therefore it must be “spinning on its side” relative to most planets. What we observed at Ultima Thule would be completely new.

The New Horizons engineering and science teams were charged with the task of completing mankind’s first reconnaissance of the outermost known solar system. Twenty years before, the National Science Council had set the reconnaissance of the Kuiper belt as its top planetary priority. The Kuiper belt forms a huge donut-shaped zone of ice-rich objects orbiting the Sun past the last of the major planets, Neptune. Pluto, with its five satellites, is a little wider than the state of Texas and is the largest of these, but thousands of smaller objects, many barely detectable, share this zone with Pluto. Would these smaller objects be frozen remnants of the birth of our solar system?

I arrived at APL on December 28, three days ahead of the encounter, although others had arrived a day or two earlier. (Team members concerned with finding out whether Ultima Thule had moons, rings, or other hazards of its own had been there for a few weeks.) The laboratory was rather quiet, as it was the week between Christmas and New Year’s Day and so many of the staff were away on holiday. We mostly had the building to ourselves and it was almost creepy.

As the encounter observation sequence had been designed months before and had already been uplinked to the spacecraft, and we would not even know the true shape of Ultima Thule until the day before the encounter, the science team did not have much to do except speculate on what we might see. Speculate we did. Any such debates always start with the nature of the Kuiper belt itself and any similar or related objects we might have visited.

The Kuiper belt is similar to the asteroid belt between Mars and Jupiter in that it comprises large numbers of small irregularly shaped objects that did not accrete to form a single planetary mass, but there the similarity ends. Objects in the Kuiper belt (KBOs) formed at great distances from the heat of the Sun and are rich in volatile ices, including methane, nitrogen, carbon dioxide, and of course water ice. Short-period comets such as Wild 2, Temple 1, and 67P/Churyumov-Gerasimenko, each of which have been visited by spacecraft, offered glimpses of what smaller KBO objects might look like. Short-period comets are KBOs whose orbits had been shifted, bringing them close to the Sun. Each of these comets had borne the scars of multiple passes near the Sun’s heat, which had volatilized much of their icy surfaces. The surface of 67P, for example, was revealed by Rosetta as heavily pitted and eroded and coated with dust. Would Ultima Thule betray any hints of this sublimation even though it had never been close to the Sun?

The Kuiper belt is much also larger and more distended than the asteroid belt, with much greater distances between objects. This means the collision rates between objects are much lower and we might have a good chance to look further back in time on their surfaces. In the asteroid belt, most objects there have been heavily bombarded by multitudes of small objects, forming thousands of craters on their surfaces and erasing much of their earliest histories. Would Ultima Thule be a cratered lump with no record of its early history, or would we see evidence of its original construction?

In January, as we entered a new decade, the big day finally arrived. The Maryland winter was cold but not bitter (but sadly, no snow). Media representatives had gathered to share the event with the world, even if not quite in the same numbers as the encounter with Pluto. We gathered in the Geology meeting room to await the scheduled approach images, which would hopefully reveal something of the shape of Ultima Thule.

The New Horizons spacecraft was traveling at ~14 kilometers per second past Ultima Thule, which was only ~30 kilometers across. This gave us only one chance to get it right. As the position of Ultima Thule was uncertain, and the encounter sequence had been designed several months previously, the imaging observations had to be planned with this position uncertainty in mind. As we approached Ultima Thule, that uncertainty shrank and a final update was transmitted to the spacecraft early that day. Given the small size of the object, it was also likely that we would not see geologic features until the spacecraft was only a few hours away from the target, and those images would not start to be downlinked until the morning of January 1.

Because of the tight timeline, New Horizons was programmed, as at Pluto, to ignore Earth during the critical encounter period and go radio silent in order to use all its time for making observations. So December 31 began with a short playback of a few approach images, followed by more than 18 hours of silence as we all awaited word on the success of the encounter. Given that the one-way travel time for radio signals was six hours, there wasn’t anything anyone on Earth could do to help anyway.

There was considerable excitement as the first images of the day were revealed. Even though there was a nominal pixel size of only ~5 kilometers, or ~6 pixels across the entire body, these images clearly showed an elongate object, just as the stellar occultation observations of Ultima Thule from Argentina in 2017 had shown. Although still fuzzy in appearance, they showed an object looking something like a bowling pin. We would have to wait until mid-afternoon the next day, January 1, for the next transmissions to really see Ultima Thule for what it was.

We gathered at the main APL auditorium late on December 31 to celebrate the official time of encounter (despite the radio silence), which by a fluke of cosmic alignment occurred at half past midnight on New Year’s Eve. We got to celebrate the new year with friends and colleagues at a research lab in the middle of Maryland. As at Pluto three-and-a-half years previously, however, we had to wait another 10 hours before the signal from the spacecraft that it had turned toward home would arrive to let us know how things went. In the meantime, we all went back to our hotels to try and get some sleep. The only remaining question was whether our targeting of Ultima Thule was as accurate as we hoped.

At mid-morning on January 1, we all gathered again at the main auditorium to await the first signals. Although it played out in essentially the same way, somehow it felt a little different than at Pluto. We had had that success and knew our little piano-sized spacecraft could do the job, but still there was the mystery of Ultima Thule and whether it had any surprises in store for us. Several hundred of us gathered and waited as the main video screen went to Mission Operations. A big cheer went up when Alice Bowman called out “In lock with telemetry.” This was the first sign that the spacecraft was following the command sequences. More cheers when telemetry confirmed that all the data that was expected had actually been recorded, another sign that things had gone as planned. Yay!

We spent a little time reveling in the moment before proceeding back to our Geology room across the street to await the first images sometime mid-afternoon. Before leaving the auditorium I was approached by JoAnna Wendel to see if I would be able to give a radio interview for Duncan Wilson’s radio show in Auckland, New Zealand. Of course I agreed; this would be the longest-distance interview I’d ever done! Since it was their morning show, the interview was scheduled for approximately 1:00 p.m. U.S. Eastern Standard Time, shortly before we received the next images, which we hoped would finally reveal what Ultima Thule was. So I gave the interview on my iPhone, and proceeded to have a very pleasant 10-minute interview during which we talked about what we hoped to learn about the solar system from Ultima Thule. (Later, on January 4, I would have a video interview with Chris Ready of Launch Pad Astronomy, discussing the results.)

The real science fun began late in the afternoon on the January 1 when the first fully resolved images arrived. These images would have pixels ~300 meters across, just a bit larger than a football stadium. To our delight, they showed Ultima Thule as a bilobate object, basically two spheres “kissing” each other, and a dramatic confirmation of the stellar occultation results. The two lobes looked to be relatively intact bodies that had very gently merged (rather than collided). Apparently they had formed as separate worlds and had come together shortly thereafter, with almost no resulting damage.

Another surprise was that both lobes, while lumpy, were a lot less lumpy and very much less bashed around by impacts than other small icy bodies closer to the Sun. This object clearly did not have many craters on it. In fact, it was hard to be sure in these first images if there were any true impact craters at all. We would have to await the later images, which would have pixels as small as 35 meters across, a factor of 10 improvement. It was clear from these initial images, however, that we were seeing a body relatively undamaged since it was constructed, a window into the formation of planets.

A few more images came down later that evening, and by 10:00 p.m. we were putting together the first stereo pairs of the object from which we hoped to learn more about the object’s shape, which was proving to be flatter than expected. At first described as two pancakes touching, a more accurate description might be two jelly (or crème-filled) donuts touching.

As the images came down over the next three days we set out to understand what we were seeing. We even brought out the modeling clay and molded miniature models of Ultima Thule to try to replicate what we were seeing in the images. It is rather remarkable how accurate those initial clay models turned out to be! Earlier that week, Dr. Brian May (perhaps more famously known as the guitarist for the rock group Queen) had arrived to share in the encounter. May, an astrophysicist and New Horizons groupie, is also a stereo image enthusiast like myself, and we collaborated on putting together the best stereo pairs of Ultima Thule.

On the afternoon of January 2, I was surprised to get a request to be in the New Horizons press briefing to take place on the following day to show the latest stereo images. Unfortunately, the stereo pairs in hand were not of high enough quality to derive topographic maps (that would come later), but it gave us a chance to talk about the images in this way. After three decades in planetary science, I had never been a part of a mission press conference. It was fascinating to get a “behind the scenes” look at how they work.

There is a lot of preparation before holding a press briefing. The participants meet several times in the days before to discuss what topics will be covered. A consensus is reached on the materials and conclusions that the team is confident in and can be presented to the public. If we were not confident about data or a conclusion, then we would not present it until we were. Graphics are reviewed to make them as clear and understandable as possible. A rehearsal is run a few hours before. It was amusing to look back on it afterward, because all the instructions to look at this camera or that person as I was speaking were completely forgotten when my turn at the microphone came. Better luck next time, I guess!

By January 4, many of the team members were packing up to go back home. We would not be getting much new data for a few weeks as New Horizons worked on other observations and downlinks. The best images would come down in late January and throughout the month of February. We knew from the available images that our targeting looked really good. We also knew that in the highest-resolution images, at ~35 meters pixel scale, Ultima Thule would just fill the entire imaging frame. If the pointing was a little bit off, we might miss half or more of the object. Back home, we were all eagerly anticipating these images in mid-February when they came down, and to our delight, there was Ultima Thule right in the middle of the frames exactly where we hoped it would be. Tiny pits less than 1 kilometer across were visible, as were a few small linear fractures, but otherwise Ultima Thule was a pair of similar merged objects, lumpy but surprisingly undamaged, and a relict of the formation of the solar system. It just doesn’t get any better than that.