The solar system beyond Neptune is full of worlds hosting moons. The dwarf planets Pluto, Eris, and Haumea are all known to have one or moons orbiting them. However, for years and despite no small amount of search effort, the dwarf planet Makemake appeared to be solitary.
In many ways, Makemake is much like Pluto; it is two-thirds the size and has a methane-covered surface with a reddish color that is indicative of a dose of tholin-like material. So why did Makemake lack a moon? Did Makemake made it through the history of the solar system without suffering a moon-forming giant impact like the other distant dwarf planets?
Now we know that the picture was simply incomplete before. Makemake does have a moon!
Earlier this week Marc Buie, Will Grundy, Keith Noll, and myself announced the discovery of a Makemakean moon in Hubble Space Telescope images. We nicknamed “MK2” and the IAU has provisionally designated it S/2015 (136472) 1. We've done a preliminary characterization of MK2's properties and submitted a paper describing its implications, which you can read here. In the following post I'll walk through how we found the moon, a few of the things we've learned about it, and what we can learn with future observations.
NASA, ESA, and A. Parker (Southwest Research Institute)
Artist's impression of Makemake and its moon
Current best estimates for the properties of Makemake's moon, MK2: Diameter: roughly 170 kilometers (for 4% albedo) Semi-major axis: >21,000 km (for circular orbit) Orbital period: >12 days (for circular orbit)
How To Find A Moon
The vast majority of Kuiper Belt companions have been discovered with the Hubble Space Telescope. The combination of its high sensitivity and its very high angular resolution make it able to see very faint objects right next to very bright objects. This is what allowed us to discover MK2 sitting adjacent to the 1,300-times brighter Makemake.
Exoplanet direct imaging also requires the ability to see faint targets very close to bright ones, and ground-based instruments like the Gemini Planet Imager are remarkably good at this. However, they don't quite have the sensitivity to reliably detect moons like MK2. For example, the exoplanet 51 Eri b discovered in Gemini imagery is only a factor of a few times fainter than Makemake itself; MK2 is hundreds of times fainter still.
That said, the images collected by Hubble present their own challenges. Because it is in low Earth orbit, Hubble's detectors are frequently struck by cosmic rays that create bright streaks on the images. These streaks wipe out any real information about the sky behind them. To get around this, we design Hubble observations to re-image the same spot on the sky several times, and use these repeat observations to correct for the transient cosmic ray strikes.
One way to remove the cosmic rays is to lay the images on top of one another, and then sort each pixel by brightness. In this “stack,” things like cosmic rays will float to the top layer, while the image information common to most of the images sinks to about the middle layer. The gif below is an illustration of this sorting in progress. The images are centered on MK2, and are all six of the Hubble images that revealed the moon. The bright cosmic ray streaks float over to the right (top of the stack), and after sorting the selected image is relatively clean of image detects.
Animation of the pixel sorting and stacking
The original images are arrayed left to right and are centered on MK2, and then the pixels are sorted by value, with the brightest pixels at each location moving to the left, and the darkest moving to the right. Cosmic rays and background stars end up isolated in the right-most frame, and the selected frame at the end of the animation is clear of most spurious artifacts.
Since Makemake moves with time, this process also removes background stars. Stars are fixed, and as Hubble tracks Makemake's motion, the stars don't re-appear in the same pixels in subsequent images. They also drift up to the top of the stack like cosmic rays, and their light is suppressed in the lower images.
As you can see in the animation, even before sorting, MK2 is easily visible in every image. The discovery epoch imagery was unambiguous; MK2 is relatively bright in the data and moving in perfect lock-step with Makemake. However, when Makemake was imaged two days later, things weren't so straightforward. Because MK2 wasn't making itself obvious in the data, I had to do somewhat more processing to see if I could tease it out of the glare from Makemake. The trick is to remove the light from Makemake without touching the light from MK2; this can be done by subtracting either a synthetic image of a solitary Makemake, or by subtracting the real image of Makemake from the earlier data (and since MK2 had moved, its light would not be subtracted from itself in the later images). Both approaches have their advantages and drawbacks, and neither revealed MK2.
NASA, ESA, A. Parker and M. Buie (Southwest Research Institute), W. Grundy (Lowell Observatory), and K. Noll (NASA GSFC)
Cleaned and processed images from each visit
April 27 on the left, showing MK2, and April 29 on the right showing no sign of MK2.
Since the imagery was designed to detect extremely faint moons, Makemake saturated the detector. This happens when a pixel is subjected to a bright source for long enough that it stops being able to reliably count up how many photons have struck it, and the counts start to leak into adjacent pixels. On top of that, just due to the statistics of counting, regions around bright sources also get noisier the more photons strike them. Together, these effects lead to a region around any bright source where a fainter source would not be visible. Since we could have seen MK2 far from Makemake even if it somehow got several times fainter between the first and second image set, we think the only place MK2 could hide is in this central bright region.
Just given the discovery data and the knowledge that it disappeared in toward Makemake two days later, we made some preliminary estimates of MK2's orbital properties. The most robust result is that MK2 is on a nearly edge-on orbit, with a high inclination both to the plane of the sky and the plane of Makemake's orbit around the Sun. If the orbit is circular, it takes MK2 12 days at a minimum to orbit around Makemake, and it does so at an average distance of 21,000 kilometers or more.
What Can We Do With A Moon?
Moons are great. The existence of one around a world like Makemake opens up a whole bunch of avenues of study that would have otherwise been impossible. If we can measure the orbital separation and period of the moon, Kepler's laws allow us to directly measure the total amount of mass in the system. Since MK2 is much smaller than Makemake, the vast majority of the system mass resides in Makemake. Additionally, MK2 has to have come from somewhere, and it has to have survived to the present day. MK2's properties will tell us about the processes that lead to moon formation (or capture) and the history of the environment that Makemake is embedded in.
Makemake also has a very well-measured size (from stellar occultations, of about 1430 by 1420 kilometers), so determining its mass via MK2's orbit will mean we can calculate its bulk density. Other dwarf planets with moons have relatively high densities, and it's thought that these worlds have suffered giant collisions in the past. Since these worlds are large enough to be differentiated — that is, dense materials sank to their cores and light materials floated up to their surface — a large collision can preferentially strip off lighter material, and leave behind a denser object. Adding Makemake to the list of objects with measured density will be key to determine if these giant, density-increasing collisions are a universal feature of the history of Kuiper Belt dwarf planets.
Dark Moon, Squashed Makemake
One longstanding mystery about Makemake might be cleared up by the discovery of MK2. Since Makemake is so bright, it has been targeted by many different kinds of investigations in the past. One of these investigations tracked the brightness of Makemake over time, in order to determine how it varies. If Makemake had a mottled surface like Pluto, you would expect its rotation to make its brightness vary with time. However, despite rotating once every 7.77 hours, Makemake hardly varies at all in brightness. This suggests a very uniform surface for Makemake.
Other past observations were made in very long-wavelength light, including in far infrared and sub-millimeter wavelengths by the Spitzer and Herschel space observatories. These wavelengths are sensitive to the thermal emission of objects in the outer solar system. If the surface of Makemake was very uniform, you would expect its thermal emission to look like a single black-body curve. Instead, the thermal emission from Makemake looked more like two distinct black-body curves. Previous analysis indicated that these curves look like most of Makemake is very bright and very cold, while small discrete parts of Makemake are very dark and warmer.
So, how do you get dark patches on Makemake without making its brightness vary as it rotates? Previous studies suggested two options. There might be many very small spots spaced uniformly around Makemake, so that as one disappears, another comes into view. Alternatively, we might be looking down at one of Makemake's poles, so that the spots just go around and around and never leave view.
However, the discovery of the moon has provided a third option: the dark material is not entirely on Makemake. Instead, some or all of it is the dark surface of the moon. If this is the case, MK2 must be very dark, with a surface reflecting only 2-4% of the light that strikes it. This is in contrast to the bright surface of Makemake, which reflects over 80% of the light that strikes it.
The moon also makes it less likely that we are looking down on Makemake's pole. Makemake spins so quickly that it probably squashed into an oblate spheroid. Such a shape would tend to drive Makemake's spin and the orbit of MK2 into the same plane. Since the orbit plane is edge on, if Makemake is aligned with its orbit, we are probably looking down on its equator.
In 2011, Makemake passed in front of a relatively bright star. Astronomers spread across thousands of kilometers on Earth watched as the star disappeared and reappeared, and by mapping the shape of the shadow were able to measure the size and shape of Makemake. What they saw was a shadow that was elongated in the North-South direction. Since it was thought that Makemake was rotating pole on and was oblate, not prolate, subsequent works argued that this North-South elongation was unlikely to be real. However, if Makemake is rotating equator-on and aligned with MK2, this elongation is exactly what we would expect! The long axis of the elongation is even oriented approximately along the plane of MK2's orbit, as it should be. As such, it is plausible that this observation did indeed detect the squashed shape of Makemake.
Learning More With Future Observations
MK2 offers some very exciting avenues for future observations. First and foremost, because of its edge-on orbit, MK2 is probably near a “mutual event season” with Makemake. This is where the moon and the dwarf planet actually pass in front of each other as seen from the Earth. By carefully monitoring these events, we could use them to create detailed maps of Makemake. This is exactly how the first maps of Pluto and Charon were made, using mutual events that occurred between the two during the late 1980s.
However, we can't yet tell if this Makemake-MK2 mutual event season is coming up or if we just missed it. To determine that, we need to measure MK2's orbit much more accurately with follow-up observations from Hubble. These orbital measurements are also what are needed to permit the measurement of Makemake's mass.
In the slightly longer term, the idea that MK2 has a very dark surface can be directly tested with the yet-to-be-launched James Webb Space Telescope. Since it will be able to see in the long-wavelength infrared, and will have sharper vision in these wavelengths than any observatory in history, it could actually separate the thermal emission from Makemake and MK2. With observations like that, we would be able to uniquely determine the surface properties both MK2 and Makemake. These observations require high-precision predictions of the location of MK2, so these also rely on more follow-up from Hubble before they can be planned.
No matter what observations we make in the future, we've already learned one important thing: Makemake is not alone!