Category Archives: Satellites of Neptune

Triton: a cuckoo around Neptune

Hi there! Did you know that Neptune had a prominent satellite, i.e. Triton, on a retrograde orbit? This is so unusual that it is thought to have been trapped, i.e. was originally an asteroid, and has not been formed in the protoneptunian nebula. The study I present you today, Triton’s evolution with a primordial Neptunian satellite system, by Raluca Rufu and Robin M. Canup, explains how Triton ejected the primordial satellites of Neptune. This study has recently been published in The Astronomical Journal.

The satellites of Uranus and Neptune

We are tempted to see the two planets Uranus and Neptune as kinds of twins. They are pretty similar in size, are the two outermost known planets in the Solar System, and are gaseous. A favorable orbital configuration made their visitation possible by the spacecraft Voyager 2 in 1986 and 1989, respectively.

Among their differences are the high obliquity of Uranus, the presence of rings around Uranus while Neptune displays arcs, and different configurations in their system of satellites. See for Uranus:

Semimajor axis Eccentricity Inclination Radius
Miranda 5.12 Ru 0.001 4.338° 235.8 km
Ariel 7.53 Ru 0.001 0.041° 578.9 km
Umbriel 10.49 Ru 0.004 0.128° 584.7 km
Titania 17.20 Ru 0.001 0.079° 788.9 km
Oberon 23.01 Ru 0.001 0.068° 761.4 km
Puck 3.39 Ru 0 0.319° 81 km
Sycorax 480.22 Ru 0.522 159.420° 75 km

I show on this table the main satellites of Uranus, and we can see that the major ones are at a reasonable distance (in Uranian radius Ru) of the planet, and orbit almost in the same plane. The orbit of Miranda is tilted thanks to a past mean-motion resonance with Umbriel, which means that it was originally in the same plane. So, we can infer that these satellites were formed classically, i.e. the same way as the satellites of Jupiter, from a protoplanetary nebula, in which the planet and the satellites accreted. An exception is Sycorax, which is very far, highly inclined, and highly eccentric. As an irregular satellite, it has probably been formed somewhere else, as an asteroid, and been trapped by the gravitational attraction of Uranus.

Now let us have a look at the system of Neptune:

Semimajor axis Eccentricity Inclination Radius
Triton 14.41 Rn 0 156.865° 1353.4 km
Nereid 223.94 Rn 0.751 7.090° 170 km
Proteus 4.78 Rn 0 0.075° 210 km
Larissa 2.99 Rn 0.001 0.205° 97 km

Yes, the main satellite seems to be an irregular one! It does not orbit that far, its orbit is (almost) circular, but its inclination is definitely inconsistent with an in situ formation, i.e. it has been trapped, which has been confirmed by several studies. Nereid is much further, but with a so eccentric orbit that it regularly enters the zone, which is dynamically perturbed by Triton. You can also notice the absence of known satellites between 4.78 and 14.41 Neptunian radii. This suggests that this zone may have been cleared by the gravitational perturbation of a massive body… which is Triton. The study I present you simulates what could have happened.

A focus on Triton

Before that, let us look at Triton. The system of Neptune has been visited by the spacecraft Voyager 2 in August 1989, which mapped 40% of the surface of Triton. Surprisingly, it showed a limited number of impact craters, which means that the surface has been renewed, maybe some 1 hundred of millions of years ago. Renewing the surface requires an activity, probably cryovolcanism as on the satellite of Saturn Enceladus, which should has been sustained by heating. Triton was on the action of the tides raised by Neptune, but probably not only, since tides are not considered as efficient enough to have circularized the orbit. The tides have probably been supplemented by gravitational interactions with the primordial system of Neptune, i.e. satellites and / or disk debris. If there had been collisions, then they would have themselves triggered heating. As a consequence of this heating, we can expect a differentiated structure.

Moreover, Triton orbits around Neptune in 5.877 days, on a retrograde orbit, while the rotation of Neptune is prograde. This configuration, associated with the tidal interaction between Triton and Neptune, makes Triton spiral very slowly inward. In other words, it will one day be so close to Neptune that it will be destroyed, and probably create a ring. But we will not witness it.

A numerical study with SyMBA

This study is essentially numerical. It aimed at modeling the orbital evolution of Triton, in the presence of Nereid and the putative primordial satellites of Neptune. The authors assumed that there were 4 primordial satellites, with different initial conditions, and considered 3 total masses for them: 0.3, 1, and 3 total masses of the satellites of Uranus. For each of these 3 masses, they ran 200 numerical simulations.

The simulations were conducted with the integrator (numerical code) SyMBA, i.e. Symplectic Massive Body Algorithm. The word symplectic refers to a mathematical property of the equations as they are written, which guarantee a robustness of the results over very long timescales, i.e. there may be an error, but which does not diverge. It may be not convenient if you make short-term accurate simulations, for instance if you want to design the trajectory of a spacecraft, but it is the right tool for simulating a system over hundreds of Myrs (millions of years). This code also handles close encounters, but not the consequences of impacts. The authors bypassed this problem in treating the impacts separately, determining if there were disrupting, and in that case estimated the timescales of reaccretion.


The authors found, consistently with previous studies, that the interaction between Triton and the primordial system could explain its presently circular orbit, i.e. it damped the eccentricity more efficiently than the tides. Moreover, the interaction with Triton caused collisions between the primordial moons, but usually without disruption (hit-and-run impacts). In case of disruption, the authors argue that the reaccretion would be fast with respect of the time evolution of the orbit of Triton, which means that we could lay aside the existence of a debris disk.

Moreover, they found that the total mass of the primordial system had a critical role: for the heaviest one, i.e. 3 masses of the Uranian system, Triton had only small chances to survive, while it had reasonable chances in the other two cases.

Something frustrating when you try to simulate something that happened a few hundreds of Myrs ago is that you can at the best be probabilistic. The study shows that a light primordial system is likelier to have existed than a heavy one, but there are simulations with a heavy system, in which Triton survives. So, a heavy system is not prohibited.

The study and its authors

  • The study, which is available as free article. The authors probably paid extra fees for that, many thanks to them! You can also look at it on arXiv.
  • A conference paper on the same study,
  • The ResearchGate profile of Raluca Rufu,
  • The Homepage of Robin M. Canup.

Before closing this post, I need to mention that the title has been borrowed from Matija Ćuk (SETI, Mountain View, CA), who works on this problem as well (see these two conference abstracts here and here).

That’s it for today! Please do not forget to comment. You can also subscribe to the RSS feed, and follow me on Twitter and Facebook.

Where is Triton?

Hi there! Today’s post is on the location of Triton. Triton is the largest satellite of Neptune and, of course, we know where it is. The paper I present you, entitled Precise CCD positions of Triton in 2014-2016 using the newest Gaia DR1 star catalog, aims at assisting an accurate modeling of its motion. This is a Chinese study, by Na Wang, Qing-Yu Peng, Huan-Wen Peng, H.J. Xie, S. Ma and Q.F. Zhang, which presents observations made at the Yunnan Observatory. This study has recently been published in The Monthly Notices of the Royal Astronomical Society.

Triton’s facts

Triton is by far the largest satellite of Neptune, with a mean radius of 1,350 km. It was discovered almost simultaneously with Neptune, i.e. 17 days later, in 1846. It orbits Neptune in a little less than 6 days, and rotates synchronously.

Surprisingly, its orbit is very inclined with respect to the equator of Neptune (some 24°), and it is retrograde… which means that the inclination should not be given as 24° but as 180-24 = 156°. Usually the satellites have a very small inclination, since they are supposed to have been formed in a nebula which gave birth to the planet. Such a large inclination means that Triton has probably not been formed in situ but was an asteroid, which has been trapped by Neptune. Since then, it loses some orbital energy, which has resulted in a circularization of its orbit and a fall on Neptune.

Triton has been visited by the Voyager 2 spacecraft in 1989, which covered about 40% of its surface. It revealed an atmosphere of nitrogen and evidences of melting, which indicate a geological activity, probably resulting in a differentiated body. It could even harbor or have harbored a subsurface ocean.

Astrometry in the Solar System

The goal of the paper I present today is to give accurate positions of Triton. This is called astrometry. The idea is this: measuring accurately the position of a body at a given date requires to take a picture of the satellite. Stars must be present in the field since the satellite will be positioned according to them. The result will then be compared to the predictions given by dynamical models, called ephemerides. Since we observe only in 2 dimensions, i.e. on the celestial sphere, which is a surface, then an astrometric position consists of two coordinates: the right ascension and the declination.

The positions of the stars surrounding the satellite, actually its image projected on the celestial sphere, since the stars are much further, are known thanks to systematic surveys. The satellite Gaia is currently conducting the most accurate of such surveys ever performed, and the first data were released in 2016, in the Gaia Data Release 1 (DR1, see this post, from October 2016). The accuracy of Gaia lets us hope very accurate future astrometry, and even past, since old observations could be retreated in using Gaia’s catalog.

But why making astrometry? For improving the ephemerides, which would give us a better knowledge of the orbital motion of the satellite. And why improving the ephemerides? I see at least two reasons:

  1. To help future space missions,
  2. To have a better knowledge of the physical properties of the bodies. Some of these properties, like the mass and the energy dissipation (tides), affect the orbital motion.

Potential difficulties

This study presents Earth-based observations, which are affected by:

  1. Diffraction on the CCD chip. When you observe a point as a light-source, you actually see a diffraction disk, and you have to decide which point on the disk is the position of your object. You can partly limit the size of the diffraction disk in limiting the exposure time, this prevents the chip from being saturated. If this results in a too faint object, then you can add several images.
  2. The anisotropy of the light scattering by the surface of the body (here Triton). When we see Triton, we actually see the Solar light, which is reflected by the surface of Triton. Since the surface could be pretty rough, since the limb is darker than the center because of a different incidence angle, and since Triton is not seen as a whole disk (remember the lunar crescent), then the center of the diffraction disk, i.e. the photocentre, is not exactly the location of the body.
  3. The refraction by the atmosphere. The atmosphere distorts the image, which makes the satellite-based images more accurate than the Earth-based ones. This distortion depends on the location of the observatory, and the weather. Some systems of adaptive optics exist, which partly overcome this problem.
  4. The aberration. The relative velocity of the observed object with the observer (the Earth is moving, remember?) alters the apparent position of the objects.
  5. The seeing. When you have some wind, the locations of the stars present some erratic variations.
  6. The inhomogeneity of the CCD chip. An homogeneous lightning will not result in a homogeneous response, because of the positions of the pixel on the image (you have a better sensitivity in the center than close to an edge), and some technical differences between the pixels. A way to overcome this problem is to normalize the light measurement by a flat image, which is the response to a homogeneous lightning. This is usually obtained in taking a picture in the dome, or from the averaging of many images.

This paper

This paper presents 775 new observations, i.e. 775 new positions of Triton at given dates, between 2014 and 2016. The images were taken with a 1-meter refractor at the Yunnan National Observatory, Kunming, Yunnan, China. The residuals, i.e. observed minus predicted positions, are obtained from the ephemerides made by the Jet Propulsion Laboratory in California, USA. The authors obtain mean residuals of a few tens of milli-arcsec, i.e. some thousands of kilometers. Something interesting is the dispersion of these residuals: the authors show that when the stars are positioned with Gaia DR1, the residuals are much less dispersed than with an older catalog. The authors used the catalog UCAC4, released in August 2012 by the US Naval Observatory, for comparison.

These new observations will enrich the databases and permit the future improvements of ephemerides.

Some links

That’s all for today! Please do not forget to comment. You can also subscribe to the RSS feed, and follow me on Twitter.