Tag Archives: numerical methods

9 interstellar asteroids?

Hi there! You may have recently heard of 1I/’Oumuamua, initially known as C/2017 U1, then A/2017 U1 (see here), where C stands for comet, A for asteroid, and I for interstellar object. This small body visited us last fall on a hyperbolic orbit, i.e. it came very fast from very far away, flew us by, and then left… and we shall never see it again. ‘Oumuamua has probably been formed in another planetary system, and its visit has motivated numerous studies. Some observed it to determine its shape, its composition, its rotation… and some conducted theoretical studies to understand its origin, its orbit… The study I present you today, Where the Solar system meets the solar neighbourhood: patterns in the distribution of radiants of observed hyperbolic minor bodies, by Carlos and Raúl de la Fuente Marcos, and Sverre J. Aarseth, is a theoretical one, but with a broader scope. This study examines the orbits of 339 objects on hyperbolic orbits, to try to determine their origin, in particular which of them might be true interstellar interlopers. This study has recently been accepted for publication in The Monthly Notices of the Royal Astronomical Society.


I detail the discovery of ‘Oumuamua there. Since that post, we know that ‘Oumuamua is a red dark object, probably dense. It is tumbling, i.e. does not rotate around a single rotation axis, in about 8 hours. The uncertainties on the rotation period are pretty important, because of this tumbling motion. Something really unexpected is huge variations of brightness, which should reveal either a cigar-shaped object, or an object with extreme variations of albedo, i.e. bright regions alternating with dark ones… but that would be inconsistent with the spectroscopy, revealing a reddish object. This is why the dimensions of ‘Oumuamua are estimated to be 230 × 35 × 35 meters.

Artist's impression of 'Oumuamua. © ESO/M. Kornmesser
Artist’s impression of ‘Oumuamua. © ESO/M. Kornmesser

One wonders where ‘Oumuamua comes from. An extrapolation of its orbit shows that it comes from the current direction of the star Vega, in constellation Lyra… but when it was there, the star was not there, since it moved… We cannot actually determine around which star, and when, ‘Oumuamua has been formed.

Anyway, it was a breakthrough discovery, as the first certain interstellar object, with an eccentricity of 1.2. But other bodies have eccentricities larger than 1, which make them unstable in the Solar System, i.e. gravitationally unbound to the Sun… Could some of them be interstellar interlopers? This is the question addressed by the study. If you want to understand what I mean by eccentricity, hyperbolic orbit… just read the next section.

Hyperbolic orbits

The simplest orbit you can find is a circular one: the Sun is at the center, and the planetary object moves on a circle around the Sun. In such a case, the eccentricity of the orbit is 0. Now, if you get a little more eccentric, the trajectory becomes elliptical, and you will have periodic variations of the distance between the Sun and the object. And the Sun will not be at the center of the trajectory anymore, but at a focus. The eccentricity of the Earth is 0.017, which induces a closest distance of 147 millions km, and a largest one of 152 millions km… these variations are pretty limited. However, Halley’s comet has an eccentricity of 0.97. And if you exceed 1, then the trajectory will not be an ellipse anymore, but a branch of hyperbola. In such a case, the object can just make a fly-by of the Sun, before going back to the interstellar space.

Wait, it is a little more complicated than that. In the last paragraph, I assumed that the eccentricity, and more generally the orbital elements, were constant. This is true if you have only the Sun and your object (2-body, or Kepler, problem). But you have the gravitational perturbations of planets, stars,… and the consequence is that these orbital elements vary with time. You so may have a hyperbolic orbit becoming elliptical, in which case an interstellar interloper gets trapped, or conversely a Solar System object might be ejected, its eccentricity getting larger than 1.

The authors listed three known mechanisms, likely to eject a Solar System object:

  1. Close encounter with a planet,
  2. Secular interaction with the Galactic disk (in other words, long term effects due to the cumulative interactions with the stars constituting our Milky Way),
  3. Close encounter with a star.

339 hyperbolic objects

The authors identified 339 objects, which had an eccentricity larger than 1 on 2018 January 18. The objects were identified thanks to the Jet Propulsion Laboratory’s Small-Body Database, and the Minor Planet Center database. The former is due to NASA, and the latter to the International Astronomical Union.

Once the authors got their inputs, they numerically integrated their orbits backward, over 100 kyr. These integrations were made thanks to a dedicated N-body code, powerful and optimized for long-term integration. Such algorithm is far from trivial. It consists in numerically integrating the equations of the motion of all of these 339 objects, perturbed by the Sun, the eight planets, the system Pluto-Charon, and the largest asteroids, in paying attention to the numerical errors at each iteration. This step is critical, to guarantee the validity of the results.

Some perturbed by another star

And here is the result: the authors have found that some of these objects had an elliptical orbit 100 kyr ago, meaning that they probably formed around the Sun, and are on the way to be expelled. The authors also computed the radiants of the hyperbolic objects, i.e. the direction from where they came, and they found an anisotropic distribution, i.e. there are preferred directions. Such a result has been obtained in comparing the resulting radiants from the ones given by a random process, and the distance between these 2 results is estimated to be statistically significant enough to conclude an anisotropic distribution. So, this result in not based on a pattern detected by the human eye, but on statistical calculations.

In particular, the authors noted an excess of radiants in the direction of the binary star WISE J072003.20-084651.2, also known as Scholz’s star, which is currently considered as the star having had the last closest approach to our Solar System, some 70 kilo years ago. In other words, the objects having a radiant in that direction are probably Solar System objects, and more precisely Oort cloud objects, which are being expelled because of the gravitational kick given by that star.

8 candidate interlopers

So, there is a preferred direction for the radiants, but ‘Oumuamua, which is so eccentric that it is the certain interstellar object, is an outlier in this radiant distribution, i.e. its radiant is not in the direction of Scholz’s star, and so cannot be associated with this process. Moreover, its asymptotical velocity, i.e. when far enough from the Sun, is too large to be bound to the Sun. And this happens for 8 other objects, which the authors identify as candidate interstellar interlopers. These 8 objects are

  • C/1853 RA (Brunhs),
  • C/1997 P2 (Spacewatch),
  • C/1999 U2 (SOHO),
  • C/2002 A3 (LINEAR),
  • C/2008 J4 (McNaught),
  • C/2012 C2 (Bruenjes),
  • C/2012 S1 (ISON),
  • C/2017 D3 (ATLAS).

Do we know just one, or 9 interstellar objects? Or between 1 and 9? Or more than 9? This is actually an important question, because that would constrain the number of detections to be expected in the future, and have implications for planetary formation in our Galaxy. And if these objects are interstellar ones, then we should try to investigate their physical properties (pretty difficult since they are very small and escaping, but we did it for ‘Oumuamua… maybe too late for the 8 other guys).

Anyway, more will be known in the years to come. More visitors from other systems will probably be discovered, and we will also know more on the motion of the stars passing by, thanks to the astrometric satellite Gaia. Stay tuned!

The study and its authors

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

A constantly renewed ring of Saturn

Hi there! The outstanding Cassini mission ended last September with its Grand Finale, and it gave us invaluable data, which will still be studied for many years. Today I present you a study which has recently been published in The Astrophysical Journal: Particles co-orbital to Janus and Epimetheus: A firefly planetary ring, by a Brazilian team composed of Othon C. Winter, Alexandre P.S. Souza, Rafael Sfair, Silvia M. Giuliatti Winter, Daniela C. Mourão, and Dietmar W. Foryta. This study tells us how the authors characterized a dusty ring in the system of Saturn, studied its stability, and investigated its origin.

The rings of Saturn

As you may know, Saturn is the ringed planet, its rings being visible from Earth-based amateur telescopes. Actually, the 4 major planets of our Solar System have rings, and some dwarf planets as well, i.e. Chariklo, Haumea, and possibly Chiron. But Saturn is the only one with so dense rings. I summarize below the main relevant structures and distances, from the center of Saturn:

Distance Structure
60,268 km The atmospheric pressure of Saturn reaches 1 bar.
This is considered as the equatorial radius of Saturn.
66,900 – 74,510 km D Ring
74,658 – 92,000 km C Ring
92,000 – 117,580 km B Ring
117,580 – 122,170 km Cassini Division
122,170 – 136,775 km A Ring
133,589 km Encke Gap
140,180 km F Ring
151,500 km Orbits of Janus and Epimetheus
189,000 km Orbit of Mimas
1,222,000 km Orbit of Titan

The A and B Rings are the densest ones. They are separated by the Cassini Division, which appears as a lack of material. It actually contains some, arranged as ringlets, but they are very faint. The Encke Gap is a depletion of material as well, in which the small satellite Pan confines the boundaries. Here we are interested in a dusty ring enshrouding the orbits of Janus and Epimetheus, i.e. outside the dense rings. The discovery of this ring had been announced in 2006, this study reveals its characteristics.

The rings of Saturn seen by Cassini. From right to left: the A Ring with the Encke Gap, the Cassini Division, the B Ring, the C Ring, and the D Ring. © NASA
The rings of Saturn seen by Cassini. From right to left: the A Ring with the Encke Gap, the Cassini Division, the B Ring, the C Ring, and the D Ring. © NASA

Janus and Epimetheus

The two coorbital satellites Janus and Epimetheus are a unique case in the Solar System, since these are two bodies with roughly the same size (diameters: ~180 and ~120 km, respectively), which share the same orbit around Saturn. More precisely, they both orbit Saturn in 16 hours, i.e. at the same mean orbital frequency. This is a case of 1:1 mean-motion resonance, involving peculiar mutual gravitational interactions, which prevent them from colliding. They swap their orbits every four years, i.e. the innermost of the two satellites becoming the outermost. The amplitudes of these swaps (26 km for Janus and 95 for Epimetheus) have permitted to know accurately the mass ratio between them, which is 3.56, Janus being the heaviest one.

Interestingly, Epimetheus is the first among the satellites of Saturn for which longitudinal librations have been detected. As many natural satellites, Janus and Epimetheus have a synchronous rotation, showing the same face to a fictitious observer at the surface of Saturn. For Epimetheus, large librations have been detected around this direction, which are a consequence of its elongated shape, and could reveal some mass inhomogeneities, maybe due to variations of porosity, and/or to its pretty irregular shape.

Janus and Epimetheus seen by Cassini (mosaic of 2 images). © NASA
Janus and Epimetheus seen by Cassini (mosaic of 2 images). © NASA

Images of a new ring

So, Cassini images have revealed a dusty ring in that zone. To characterize it, the authors have first extracted images likely to contain it. Such images are made publicly available on NASA’s Planetary Data System. Since that ring had been announced to have been observed on Sept 15th 2006 (see the original press release), the authors restricted to 2 days before and after that date. The data they used were acquired by the ISS (Imaging Science Subsystem) instrument of Cassini, more precisely the NAC and WAC (Narrow- and Wide-Angle-Camera). They finally found 17 images showing the ring.

The images are given as raw data. The authors needed to calibrate their luminosity with a tool (a software) provided by the Cassini team, and sometimes to smooth them, to remove cosmic rays. Moreover, they needed to consider the position of the spacecraft, to be able to precisely locate the structures they would see.

One of the Cassini images used by the authors. I have added red stars at the location of the ring. © NASA / Ciclops
One of the Cassini images used by the authors. I have added red stars at the location of the ring. © NASA / Ciclops

It appears that the ring presents no longitudinal brightness variation. In other words, not only this is a whole ring and not just an arc, but no density variation is obvious. However, it presents radial brightness variations, over a width of 7,500 km, which is wider than the 5,000 km announced in the 2006 press release.

The next step is to understand the dynamics of this ring, i.e. its stability, its origin, the properties of the particles constituting it… Let us start with the stability.

The ring is removed in a few decades

The authors ran N-body simulations, i.e. numerical integrations of the equations ruling the motion of a ring particle, which would be gravitationally perturbed by the surrounding bodies, i.e. Saturn, and the Janus, Epimetheus, Mimas, Enceladus, Tethys, Dione, and Titan. Moreover, for a reason that I will tell you at the end of this article, the authors knew that the particles were smaller than 13 μm. The motions of such small particles are affected by the radiation pressure of the Sun, in other words the Solar light pushes the particles outward.

The authors simulated 14 times the motion of 18,000 particles equally distributed in the rings. Why 14 times? To consider different particle sizes, i.e. one set with 100 μm-sized particles, and the other sets with sizes varying from 1μm to 13μm.
And it appears that these particles collide with something in a few decades, mostly Janus or Epimetheus. This leaves two possibilities: either we were very lucky to be able to take images of the ring while it existed, or a process constantly feeds the ring. The latter option is the most probable one. Let us now discuss this feeding process.

Renewing the ring

The likeliest sources of material for the rings are ejecta from Janus and Epimetheus. The question is: how were these ejecta produced? By impacts, probably. This study show that Janus and Epimetheus are impacted by the particles constituting the rings, but the impact velocities would not permit to produce ejecta. This is why the authors propose a model, in which interplanetary particles collide with the satellites, generating ejecta.

A firefly behavior

And let us finish with something funny: the ring seems to behave like a firefly, i.e. sometimes bright, and sometimes dark, which means undetectable while present.
To understand what happens, figure out how the light would cross a cloud of particles. If the cloud is dense enough, then it would reflect the light, and not be crossed. But for dust, the light would be refracted, i.e. change its direction. This depends on the incidence angle of the Solar light, i.e. on the geometrical configuration of the Sun-Saturn-ring system. The Solar incidence angle is also called phase. And this phase changes with the orbit of Saturn, which results in huge brightness variations of the ring. Sometimes it can be detected, but most of the time it cannot. This can be explained and numerically estimated by the Mie theory, which gives the diffusion of light by small particles. This theory also explains the creation of rainbows, the Solar light being diffracted by droplets of water.

The study and its authors

And 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.

Tides in the lakes of Titan

Hi there! The satellite of Saturn Titan has hydrocarbon seas, i.e. lakes made of liquid ethane and methane. When you have a sea, or a lake, you may have tides, and this is what this study is about. I present you A numerical study of tides in Titan’s northern seas, Kraken and Ligeia Maria, by David Vincent, Özgür Karatekin, Jonathan Lambrechts, Ralph D. Lorenz, Véronique Dehant, and Éric Deleersnijder, which has recently been accepted for publication in Icarus.

The lakes of Titan

The presence of hydrocarbons in such a thick atmosphere as the one of Titan has suggested since the spacecraft Voyager 1 than methane and ethane could exist in the liquid state on the surface of Titan. There could even be a cycle of methane, as there is a hydrological cycle on Earth, in which the liquid methane on the surface feeds the clouds of gaseous methane in the atmosphere, and conversely.

The spacecraft Cassini has detected dark smooth features, which revealed to be these hydrocarbon seas. Here is a list of the largest ones:

Location Diameter
Kraken Mare 68.0°N 310.0°W 1,170 km
Ligeia Mare 79.0°N 248.0°W 500 km
Punga Mare 85.1°N 339.7°W 380 km
Jingpo Lacus 73.0°N 336.0°W 240 km
Ontario Lacus 72.0°S 183.0°W 235 km
Mackay Lacus 78.32°N 97.53°W 180 km
Bolsena Lacus 75.75°N 10.28°W 101 km

I present you only the detected lakes with a diameter larger than 100 km, but some have been detected with a diameter as small as 6 km. It appears that these lakes are located at high latitudes, i.e. in the polar regions. Moreover, there is an obvious North-South asymmetry, i.e. there are much more lakes in the Northern hemisphere than in the Southern one. This could be due to the circulation of clouds of Titan: they would form near the equator, from the evaporation of liquid hydrocarbons, and migrate to the poles, where they would precipitate (i.e. rain) into lakes. Let us now focus on the largest two seas, i.e. Kraken and Ligeia Maria.

Kraken and Ligeia Maria

Kraken and Ligeia Maria are two adjacent seas, which are connected by a strait, named Trevize Fretum, which permit liquid exchanges. Kraken is composed of two basins, named Kraken 1 (north) and Kraken 2 (south), which are connected by a strait named Seldon Fretum, which dimensions are similar to the strait of Gibraltar, between Morocco and Spain.

Kraken and Ligeia Maria. © NASA
Kraken and Ligeia Maria. © NASA

Alike the Moon and Sun which raise tides on our seas, Saturn raises tides on the lakes. These tides cannot be measured yet, but they can be simulated, and this is what the authors did. In a previous study, they had simulated the tides on Ontario Lacus.

They honestly admit that the tides on Kraken and Ligeia Maria have already been simulated by other authors. Here, they use a more efficient technique, i.e. which uses less computational resources, and get consistent results.

Numerical modeling with SLIM

Computational fluid dynamics, often referred as CFD, is far from an easy task. The reason is that the dynamics of fluids in ruled by non-linear partial derivative equations like the famous Navier-Stokes, i.e. equations which depend on several variables, like the time, the temperature, the location (i.e. where are you exactly on the lake?), etc. Moreover, they depend on several parameters, some of them being barely constrained. We accurately know the gravitational tidal torque due to Saturn, however we have many uncertainties on the elasticity of the crust of Titan, on the geometry of the coast, on the bathymetry, i.e. the bottom of the seas. So, several sets of parameters have to be considered, for which numerical simulations should be run.

It is classical to use a finite element method for problems of CFD (Computational Fluid Dynamics, remember?). This consists to model the seas not as continuous domains, but as a mesh of finite elements, here triangular, on which the equations are defined.
The structure of the mesh is critical. A first, maybe intuitive, approach would be to consider finite elements of equal size, but it appears that this way of integrating the equations is computationally expensive and could be optimized. Actually, the behavior of the fluid is very sensitive to the location close to the coasts, but much less in the middle of the seas. In other words, the mesh needs to be tighter at the coasts. The authors built an appropriate mesh, which is unstructured and follow the so-called Galerkin method, which adapts the mesh to the equations.

The authors then integrated the equations with their homemade SLIM software, for Second-generation Louvain-la-Neuve Ice-ocean Model. The city of Louvain-la-Neuve hosts the French speaking Belgian University Université Catholique de Louvain, where most of this study has been conducted. The model SLIM has been originally built for hydrology, to model the behavior of fluids on Earth, and its simulations have been successfully confronted to terrain measurements. It thus makes sense to use it for modeling the behavior of liquid hydrocarbons on Titan.

In this study, the authors used the 2-dimensional shallow water equations, which are depth-integrated. In other words, they directly simulated the surface rather than the whole volume of the seas, which of course requires much less computation time.
Let us now see their results.

Low diurnal tides

The authors simulated the tides over 150 Titan days. A Titan day is 15.95 days long, which is the orbital period of Titan around Saturn. During this period, the distance Titan-Saturn varies between 1,186,680 and 1,257,060 km because the orbit of Titan is eccentric, and so does the intensity of the tidal torque. This intensity also varies because of the obliquity of Titan, i.e. the tilt of its rotation axis, which is 0.3°. Because of these two quantities, we have a period of variation of 15.95 days, and its harmonics, i.e. half the period, a third of the period, etc.

It appeared from the simulations that the 15.95-d response is by far the dominant one, except at some specific locations where the tides cancel out (amphidromic points). The highest tides are 0.29 m and 0.14 m in Kraken and Ligeia, respectively.

Higher responses could have been expected in case of resonances between eigenmodes of the fluids, i.e. natural frequencies of oscillations, and the excitation frequencies due to the gravitational action of Saturn. It actually appeared that the eigenmodes, which have been computed by SLIM, have much shorter periods than the Titan day, which prevents any significant resonance. The author did not consider the whole motion of Titan around Saturn, in particular the neglected planetary perturbations, which would have induced additional exciting modes. Anyway, the corresponding periods would have been much longer than the Titan day, and would not have excited any resonance. They would just have given the annual variations of tides, with a period of 29.4 years, which is the orbital period of Saturn around the Sun.

Fluid exchanges between the lakes

SLIM permits to trace fluid particles, which reveals the fluid exchanges between the basins. Because of their narrow geometry, the straits are places where the currents are the strongest, i.e. 0.3 m/s in Seldon Fretum.
The volumetric exchanges are 3 times stronger between Kraken 1 and Kraken 2 than between Kraken and Ligeia. These exchanges behave as an oscillator, i.e. they are periodic with respect to the Titan day. As a consequence, there is a strong correlation between the volume of Kraken 1, and the one of Kraken 2. Anyway, these exchanges are weak with respect to the volume of the basins.

The attenuation is critical

The authors studied the influence of the response with respect to different parameters: the bathymetry of the seas (i.e., the geometry of the bottom), the influence of bottom friction, the depth of Trevize Fretum, and the attenuation factor γ2, which represents the viscoelastic response of the surface of Titan to the tidal excitation. It appears that γ2 plays a key role. Actually, the maximum tidal range is an increasing function of the attenuation, and in Seldon and Trevize Fretum, the maximum velocities behave as a square root of γ2. It thus affects the fluid exchanges. Moreover, these exchanges are also affected by the depth of Trevize Fretum, which is barely constrained.

Another mission to Titan is needed to better constrain these parameters!

The study and its authors

And 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. And let me wish you a healthy and happy year 2018.

Does Neptune have binary Trojans?

Hi there! Jupiter, Uranus and Neptune are known to share their orbits with small bodies, called Trojans. This is made possible by a law of celestial mechanics, which specifies that the points located 60° ahead and behind a planet on its orbit are stable. Moreover, there are many binary objects in the Solar System, but no binary asteroid have been discovered as Trojans of Neptune. This motivates the following study, Dynamical evolution of a fictitious population of binary Neptune Trojans, by Adrián Brunini, which has recently been accepted for publication in The Monthly Notices of the Royal Astronomical Society. In this study, the author wonders under which conditions a binary Trojan of Neptune could survive, which almost means could be observed now.

The coorbital resonance

The coorbital resonance is a 1:1 mean-motion resonance. This means that the two involved bodies have on average the same orbital frequency around their parent one. In the specific case of the Trojan of a planet, these two objects orbit the Sun with the same period, and the mass ratio between them makes that the small body is strongly affected by the planet, however the planet is not perturbed by the asteroid. But we can have this synchronous resonance even if the mass ratio is not huge. For instance, we have two coorbital satellites of Saturn, Janus and Epimetheus, which have a mass ratio of only 3.6. Both orbit Saturn in ~16 hours, but in experiencing strong mutual perturbations. They are stable anyway.

In the specific problem of the restricted (the mass of the asteroid is negligible), planar (let us assume that the planet and the asteroid orbit in the same plane), circular (here, we neglect the eccentricity of the two orbits) 3-body (the Sun, the planet and the asteroid) problem, it can be shown that if the planet and the asteroid orbit at the same rate, then there are 5 equilibriums, for which the gravitational actions of the planet and the Sun cancel out. 3 of them, named L1, L2 and L3, are unstable, and lie on the Sun-planet axis. The 2 remaining ones, i.e. L4 and L5, lag 60° ahead and behind the planet, and are stable. As a consequence, the orbits with small oscillations around L4 and L5 are usually stable, even if the real configuration has some limited eccentricity and mutual inclination. Other stable trajectories exist theoretically, e.g. horseshoe orbits around the point L4, L3 and L5. The denomination L is a reference to the Italian-born French mathematician Joseph-Louis Lagrange (1736-1813), who studied this problem.

The Lagrange points, in a reference frame rotating with Neptune.
The Lagrange points, in a reference frame rotating with Neptune.

At this time, 6,701 Trojans are known for Jupiter (4269 at L4 and 2432 at L5), 1 for Uranus, 1 for the Earth, 9 for Mars, and 17 for Neptune, 13 of them orbiting close to L4.

The Trojans of Neptune

You can find an updated list of them here, and let me gather their main orbital characteristics:

Location Eccentricity Inclination Magnitude
2004 UP10 L4 0.023 1.4° 8.8
2005 TO74 L4 0.052 5.3° 8.3
2001 QR322 L4 0.028 1.3° 7.9
2005 TN53 L4 0.064 25.0° 9.3
2006 RJ103 L4 0.031 8.2° 7.5
2007 VL305 L4 0.060 28.2° 7.9
2010 TS191 L4 0.043 6.6° 8.0
2010 TT191 L4 0.073 4.3° 7.8
2011 SO277 L4 0.015 9.6° 7.6
2011 WG157 L4 0.031 22.3° 7.1
2012 UV177 L4 0.071 20.9° 9.2
2014 QO441 L4 0.109 18.8° 8.3
2014 QP441 L4 0.063 19.4° 9.3
2004 KV18 L5 0.187 13.6° 8.9
2008 LC18 L5 0.079 27.5° 8.2
2011 HM102 L5 0.084 29.3° 8.1
2013 KY18 L5 0.121 6.6° 6.6

As you can see, these are faint bodies, which have been discovered between 2001 and 2014. I have given here their provisional designations, which have the advantage to contain the date of the discovery. Actually, 2004 UP10 is also known as (385571) Otrera, a mythological Queen of the Amazons, and 2005 TO74 has received the number (385695).

Their dynamics is plotted below:

Dynamics of the Trojans of Neptune, at the Lagrangian points L4 and L5 (squares).
Dynamics of the Trojans of Neptune, at the Lagrangian points L4 and L5 (squares).

Surprisingly, the 4 Trojans around L5 are outliers: they are the most two eccentric, the remaining two being among the three more inclined Trojans. Even if the number of known bodies may not be statistically relevant, this suggests an asymmetry between the two equilibriums L4 and L5. The literature has not made this point clear yet. In 2007, a study suggested an asymmetry of the location of the stable regions (here), but the same authors said one year later that this was indeed an artifact introduced by the initial conditions (here). In 2012, another study detected that the L4 zone is more stable than the L5 one. Still an open question… In the study I present today, the author simulated only orbits in the L4 region.

Binary asteroids

A binary object is actually two objects, which are gravitationally bound. When their masses ratio is of the order of 1, we should not picture it as a major body and a satellite, but as two bodies orbiting a common barycenter. At this time, 306 binary asteroids have been detected in the Solar System. Moreover, we also know 14 triple systems, and 1 sextuple one, which is the binary Pluto-Charon and its 4 minor satellites.

The formation of a binary can result from the disruption of an asteroid, for instance after an impact, or after fission triggered by a spin acceleration (relevant for Near-Earth Asteroids, which are accelerated by the YORP effect), or from the close encounter of two objects. The outcome is two objects, which orbit together in a few hours, and this system evolves… and then several things might happen. Basically, it either evolves to a synchronous spin-spin-orbit resonance, i.e. the two bodies having a synchronous rotation, which is also synchronous with their mutual orbit (examples: Pluto-Charon, the double asteroid (90) Antiope), or the two components finally split… There are also systems in which only one of the components rotates synchronously. Another possible end-state is a contact binary, i.e. the two components eventually touch together.

At this time, 4 binary asteroids are known among the Trojans asteroids of Jupiter. None is known for Neptune.

Numerical simulations

The author considered fictitious binary asteroids close to the L4 of Neptune, and propagated the motion of the two components, in considering the planetary perturbations of the planets, over 4.5 Byr, i.e. the age of the Solar System. A difficulty for such long-term numerical studies is the handling of numerical uncertainties. Your numerical scheme includes a time-step, which is the time interval between the simulated positions of the system, i.e. the locations and velocities of the two components of the binary. If your time-step is too large, you will have a mathematical uncertainty in your evaluation. However, if you shorten it, you will have too many iterations, which means a too long calculation time, and the accumulations of round-off errors due to the machine epsilon, i.e. rounding in floating point arithmetic.
A good time step should be a fraction of the shortest period perturbing the system. Neptune orbits the Sun in 165 years, which permits a time step of some years, BUT the period of a binary is typically a few hours… which is too short for simulations over the age of the Solar System. This problem is by-passed in averaging the dynamics of the binary. This means that only long-term effects are kept. In this case, the author focused on the Kozai-Lidov effect, which is a secular (i.e. very long-term) raise of the inclination and the eccentricity. Averaging a problem of gravitational dynamics is always a challenge, because you have to make sure you do not forget a significant contribution.
The author also included the tidal interaction between the two components, i.e. the mutual interaction triggering stress and strain, and which result in dissipation of energy, secular variation of the mutual orbits, and damping of the rotation.
He considered three sets of binaries: two with components of about the same size, these two samples differing by the intensity of tides, and in the third one the binary are systems with a high mass ratio, i.e. consisting of a central body and a satellite.

Survival of the binaries

The authors find that for systems with strong tides, about two thirds of the binaries should survive. The tides have unsurprisingly a critical role, since they tend to make the binary evolve to a stable end-state, i.e. doubly synchronous with an almost circular mutual orbit. However, few systems with main body + satellite survive.

Challenging this model

At this time, no binary has been found among the Trojans of Neptune, but this does not mean that there is none. The next years shall tell us more about these bodies, and once they will be statistically significant, we would be able to compare the observations with the theory. An absence of binaries could mean that they were initially almost absent, i.e. lack of binaries in that region (then we should explain why there are binaries in the Trans-Neptunian population), or that the relevant tides are weak. We could also expect further theoretical studies, i.e. with a more complete tidal dynamics, and frequency-dependent tides. Here, the author assumed a constant tidal function Q, while it actually depends on the rotation rate of the two bodies, which themselves decrease all along the evolution.

So, this is a model assisting our comprehension of the dynamics of binary objects in that region. As such, it should be seen as a step forward. Many other steps are to be expected in the future, observationally and theoretically (by the way, could a Trojan have rings?).

The study and its author

And 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.

Breaking an asteroid

Hi there! Asteroids, these small bodies in the Solar System, are fascinating by the diversity of their shapes. This is a consequence of their small sizes. Another consequence is their weakness, which itself helps to split some of them into different parts, sometimes creating binary objects, asteroids families… The study I present you today, Internal gravity, self-energy, and disruption of comets and asteroids, by Anthony R. Dobrovolskis and Donald G. Korycansky, proposes an accurate computation of the required energy to provoke this break-up, at any place of the asteroid, i.e. you are more efficient when you hit at a given location. This study has recently been accepted for publication in Icarus.

Shapes of asteroids

Please allow me, in this context, to call asteroid a comet, a comet being a small body, i.e. like an asteroid, but with a cometary activity. The important thing is that the involved bodies are small enough.

Beyond a given size, i.e. a diameter of ~400 km, a planetary body is roughly spheroidal, i.e. it is an ellipsoid with it two equatorial axes almost equal and the polar one smaller, because of its rotation. For a tidally despun body, like the Moon, or a satellite of a giant planet, the shape is more triaxial, since the tidal (gravitational) action of the parent planet tends to elongate the equatorial plane. The same phenomenon affects Mercury.

However, for smaller bodies, the self-gravitation is not strong enough to make the body look more or less like a sphere. As a consequence, you can have almost any shape, some bodies are bilobate, some are contact binaries, i.e. two bodies which permanently touch together, some others are rubble piles, i.e. are weak aggregates of rocks, with many voids.

These configurations make these bodies likely to undergo or have undergone break-up. This can be quantified by the required energy to extract some material from the asteroid.

The energies involved

For that, an energy budget must be performed. The relevant energies to consider are:

  • The impact disruption energy: the minimum kinetic energy of an impactor, to shatter the asteroid and remove at least half of its mass,
  • The shattering energy: the minimum energy needed to shatter the asteroid into many small pieces. It is part of the impact disruption energy. This energy is roughly proportional to the mass of the asteroid. It represents the cohesion between the adjacent pieces.
  • The binding energy: this energy binds the pieces constituting the asteroid. In other words, once you have broken an asteroid (don’t try this at home!), you have to make sure the pieces will not re-aggregate… because of the binding energy. For that, you have to bring enough energy to disperse the fragments.
  • The self-gravitational energy: due to the mutual gravitational interaction between the blocks constituting the asteroids. Bodies smaller than 1 km are strength-dominated, i.e. they exist thanks to the cohesion between the blocks, which is the shatter energy. However, larger bodies are gravity-dominated.
  • The kinetic energy of rotation: the spin of these bodies tends to enlarge the equatorial section. In that sense, it assists the break-up process.

This study addresses bodies, which are far enough from the Sun. This is the reason why I do not mention its influences, i.e. the tides and the thermic effects, which could be relevant for Near-Earth Objects. In particular, the YORP effect is responsible for the fission of some of them. I do not mention the orbital kinetic energy of the asteroid either. Actually the orbital motion is part of the input energy brought by an impact, since the relative velocity of the impactor with respect to the target is relevant in this calculation.

I now focus on the two cases studied by the authors to illustrate their theory: the asteroid Kleopatra and the comet 67P/Churyumov-Gerasimenko.

2 peculiar cases: Kleopatra and Churyumov-Gerasimenko

216 Kleopatra is a Main-Belt asteroid. Adaptive optics observations have shown that is is constituted of two masses bound by material, giving a ham-bone shaped. As such, it can be considered as a contact binary. It is probably a rubble pile. Interestingly, observations have also shown that Kleopatra has 2 small satellites, Alexhelios and Cleoselene, which were discovered in 2008.

Reconstruction of the shape of Kleopatra. © NASA
Reconstruction of the shape of Kleopatra. © NASA

However, 67P Churyumov-Gerasimenko is a Jupiter-family comet, i.e. its aphelion is close to the orbit of Jupiter, while its perihelion is close to the one of the Earth. It has an orbital period of 6.45 years, and was the target of the Rosetta mission, which consisted of an orbiter and a lander, Philae. Rosetta orbited Churyumov-Gerasimenko between 2014 and 2016. The shape of this comet is sometimes described as rubber ducky, with two dominant masses, a torso and a head, bound together by some material, i.e. a neck.

Churyumov-Gerasimenko seen by Rosetta. © ESA
Churyumov-Gerasimenko seen by Rosetta. © ESA
216 Kleopatra 67P/Churyumov-Gerasimenko
Semimajor axis 2.794 AU 3.465 AU
Eccentricity 0.251 0.641
Inclination 13.11° 7.04°
Spin period 5.385 h 12.761 h
Mean radius 62 km 2.2 km
Magnitude 7.30 11.30
Discovery 1880 1969

The irregular shapes of these two bodies make them interesting targets for a study addressing the gravitation of any object. Let us see now how the authors addressed the problem.

Numerical modeling

Several models exist in the literature to address the gravity field of planetary bodies. The first approximation is to consider them as spheres, then you can refine in seeing them as triaxial ellipsoids. For highly irregular bodies you can try to model them as cuboids, and then as polyhedrons. Another way is to see them as duplexes, this allows to consider the inhomogeneities dues to the two masses constituting bilobate objects. The existence of previous studies allow a validation of the model proposed by the authors.

And their model is a finite-element numerical modeling. The idea is to split the surface of the asteroid into small triangular planar facets, which should be very close to the actual surface. The model is all the more accurate with many small facets, but this has the drawback of a longer computation time. The facets delimit the volume over which the equations are integrated, these equations giving the local self-gravitational and the impact disruption energies. The authors also introduce the energy rebate, which is a residual energy, due to the fact that you can remove material without removing half of it. This means that the impact disruption energy, as it is defined in the literature, is probably a too strong condition to have extrusion of material.
The useful physical quantities, which are the gravitational potential, the attraction, and the surface slope, are propagated all along the body thanks to a numerical scheme, which accuracy is characterized by an order. This order quantifies the numerical approximation which is made at each integration step. A higher order is more accurate, but is computationally more expensive.

Once the code has been run on test cases, the authors applied it on Kleopatra and Churyumov-Gerasimenko, for which the shape is pretty well known. They used meshes of 4,094 and 5,786 faces, respectively.


The validation phase is successful. The authors show that with a 3rd order numerical scheme, they recover the results present in the literature for the test cases with an accuracy of ~0.1%, which is much better than the accuracy of the shape models for the real asteroids. Regarding Kleopatra and Churyumov-Gerasimenko, they get the gravity field at any location, showing in particular excesses of gravity at the two lobes.

Such a study is particularly interesting for further missions, which would determine the gravity field of asteroids, which would then be compared with the theoretical determination by this code. Other applications are envisaged, the authors mentioning asteroid mining.

The study and its authors

And 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.

And Merry Christmas!