Were There Planets Inside Mercury’s Orbit?

by Paul Gilster on March 3, 2015

With the Mercury Messenger mission now coming to its end, it seems an appropriate time to speculate on why our inner Solar System looks the way it does. After all, as we continue finding new solar systems, we’re discovering many multi-planet systems with planets — often more than one — closer to their star than Mercury is to ours. We have Kepler to thank for these discoveries, its data analyzed in a number of recent papers including one arguing that about 5 percent of all Kepler stars have systems with tightly packed inner planets. The awkward acronym for such systems is STIP.

Well, maybe it’s not all that awkward, and Kathryn Volk and Brett Gladman (University of British Columbia) have good cause to deploy it in their new paper, which focuses on this topic. They’re wondering why our Solar System lacks planets inside Mercury’s orbit, and they point to the paper I mentioned above (Lissauer et al, 2014) as well as another by Francois Fressin and colleagues that concludes that half of all Kepler stars have at least one planet in the mass range from 0.8 to 2 Earth masses with orbits inside Mercury’s distance from our Sun, which is 0.39 AU, or 58.5 million kilometers.


Image: The Caloris basin and adjacent regions on Mercury. Recent exoplanet discoveries raise the question of why our Solar System lacks planets inside Mercury’s orbit. Can instabilities in the early Solar System help us find the answer, while at the same time explaining some of the planet’s peculiarities? Image credit: JHU/APL.

Taking as an hypothesis that nearly all F, G and K-class stars originally form with planets well within Mercury’s orbital distance, Volk and Gladman ask whether the reason we find systems without such planets today is that instabilities have destroyed these worlds through generations of catastrophic collisions and gradual re-formation, leaving (in our case) Mercury as the surviving relic. It is true that STIPs can be dynamically stable over long time-frames (hence we see the Kepler examples), but the absence of tightly packed inner worlds around many stars is here taken as the result of a ‘metastable planetary arrangement’ that leaves one or no short period planets. The Kepler STIPs we see, then, are those that have survived this process.

The authors use the Kepler data to generate systems similar to those we have uncovered, allowing them to ‘evolve’ computationally to study system dynamics, taking some simulations well beyond the first collision to see how the instabilities multiply. An initial collision often produces second collisions at higher speeds. While low-speed impacts can occur in some systems, producing far smaller amounts of debris and subsequent accretion, a fraction of STIPs experience heavy perturbation that can lead to the destruction of their inner worlds. From the paper:

Our experiments show that instability timescales in these systems are distributed such that equal fractions of the systems go unstable (reach a first planetary collision) in each decade in time (Fig. 2). This logarithmic decay is not unknown in dynamical systems (eg., Holman & Wisdom (1993)) and is presumably related to chaotic diffusion and resonance sticking near the stability boundary. After a brief, relatively stable initial period, the systems hit instability at a rate of ∼20% per time decade, with half of the systems still intact at ∼100 Myr. The exact decay rate may be influenced by our usage of the current Kepler STIPs sample (perhaps the most stable); however if this decay rate held, at ∼5 Gyr 5–10% of STIPs would not yet have reached an instability, in rough agreement with the observed STIPs frequency.

Turning the results on our own Solar System, they find that the orbits of the three outer terrestrial planets (Venus, Earth, Mars) remain unaffected on 500 million year timescales by the presence of additional planets totaling several Earth masses, all of the latter inside a distance of 0.5 AU from the Sun. Dynamical instabilities would have initiated a sequence of collisions among these worlds that left Mercury as the sole survivor. The authors argue that it is possible for the orbits of the outer terrestrial planets to remain unperturbed as the inner planets fall victim to these events.

Various issues are explained by this scenario. A series of collisions concentrates iron into the surviving remnants, which accounts for Mercury’s high density. The authors also ask whether instabilities in the inner system approximately 4 billion years ago could account for the Late Heavy Bombardment (sometimes called the ‘lunar cataclysm’), when Mercury, Venus, Earth and Mars experienced a high number of impacts. From the paper:

Gladman & Coffey (2009) estimated that 10–20% of large (m to 100 km) debris originating near current Mercury would strike Venus, with 1–4% impacting Earth (∼0.1% strikes the Moon). The Earth’s impact rate would peak ∼1–10 Myr after the event and decay on ∼30 Myr timescales as Mercury and Venus absorb most of the debris; this is a plausible match for the cataclysm’s final stages (Cuk et al. 2010). A bottom-heavy size distribution for the 1–100 km debris could explain the recent finding (Minton et al. 2015) that a main-belt asteroid source would produce too many impact basins during the cataclysm.

Such an event might be studied by future sampling missions, for:

STIP debris would likely be mostly silicate-rich mantle material similar but not identical to main-belt asteroid compositions, consistent with cataclysm impactor compositions inferred via cosmochemical means (Joy et al. 2012). The smallest dust (being blown out hyperbolically) could impact the Earth-Moon system. We estimate that 10−11 of the departing dust would strike the Moon, at vimp ∼30 km/s. If any dust or meteoroid projectiles were retained, fragments might be found in regolith breccias compacted during the cataclysm epoch.

This is an interesting model, and the authors point out that it gets us out of the difficulty of assuming an inner protoplanetary disk edge that has to be adjusted to account for Mercury and the lack of worlds interior to its orbit. We have a model that would leave the orbits of the existing terrestrial-class worlds unaffected by the series of collisions and disintegrations that Mercury emerged from, although the recipients of catastrophic amounts of early debris. The model also accounts for the apparent instability of Mercury’s orbit on a 5 to 10 billion year time-frame.

The paper is Volk and Gladman, “Consolidating and Crushing Exoplanets: Did it happen here?,” submitted to Astrophysical Journal Letters (preprint). Thanks to Andrew Tribick for the pointer to this paper.


Tom Baty March 3, 2015 at 13:47

Long time ago some astronomers actually thought there was another planet inside Mercury’s orbit; turned out it was only sunspots, but they even named it…. Vulcan. Well timed article with the passing of “Mr. Spock”.

Chakat Firepaw March 3, 2015 at 15:45

Actually a big reason for that belief was that Mercury’s orbit seemed to be wrong, it was doing something that implied another planet was periodically tugging on it.

Then along came Einstein and General Relativity which said that orbits really are supposed to do that.

Andrew LePage March 3, 2015 at 16:43

@Tom Baty March 3, 2015 at 13:47

While there may have been some mistaken reports of “Vulcan” where observers confused sunspots with a planet as you state, there were also other unconfirmed sightings of “Vulcan” that did not involve transits reported over the last couple of centuries (e.g. unexplained “stars” observed during total solar eclipses). By far one of the more convincing arguments for the existence of Vulcan was a dynamical one to explain the anomalous precession of the nodes of Mercury’s orbit. As it turned out, this precession was not caused by a planet but was the result of relativistic effects discovered by Einstein in 1915 (in fact this precession was early proof of the validity of Einstein’s theory).

RobFlores March 3, 2015 at 17:07

There is apparently no room for the idea that the solar system initial
formation was not same as for the solar systems with additional rocky
worlds inside the Orbit Venus-Mercury. I can’t see how one would ignore the composition of the solar Nebula & chaotic influence of close by solar siblings, and even the flow rates of material within the stellar nursery the sun was born in. If the solar system is an outlier at the 10% of the bell curve, Occam’s Razor still does not apply, statistical chance is still plausible.

And as I said previously Kepler lets us make inferences about what we can’t detect, But that is not data.

R Kelley March 3, 2015 at 17:50

But how did the hypothetical worlds that were destroyed inside Mercury’s orbit get there in the first place?
1. If they were formed in-situ, that could mean that many of the extra-solar planetary systems we see now could easily harbor Earth-size bodies out in the habitable zone. As you know, Kepler is not sensitive to the presence of Earth-size bodies (or smaller bodies) in the habitable zone.
2. If they migrated in from beyond the snow line, such bodies might have scattered or shepherded the material that would have otherwise condensed into rocky planets in the habitable zone into very tight orbits around the parent star. This was addressed in the article “Terrestrial Planet Formation in the Presence of Migrating Super-Earths” (Izidoro, Morbidelli, Raymond) with the authors favoring a migration of the Super-Earths over in-situ formation. However, if migration of the Super-Earths proceeds quickly (less than 1 million years from the snow-line to their final resting place close to the parent star), then the authors feel that the bulk of the mass in the habitable zone will survive the passage of the Super-Earths and can still condense into rocky planets.
Of course, if the planets passing through the habitable zone are the size of Neptune or larger, survival of planet forming material in the habitable zone is less likely.

Daniel Suggs March 3, 2015 at 19:52

@Andrew LePage could the ‘unexplained stars’ have been the result of the sun’s gravity bending light from stars behind the sun?

Joy March 4, 2015 at 3:38

Agree with RobFlores,
Planetary systems develop in a variety of environments starting from various initial conditions. Variation in stellar metallicity, mass and number of stars in a system are obvious examples of stellar system diversity. Our solar system may just represent its own formation history, no history of additional inferior planets is required.

Ronald March 4, 2015 at 5:51

This just in:
Clouds around exoplanets analyzed

Andrew LePage March 4, 2015 at 9:42

@Daniel Suggs March 3, 2015 at 19:52

The “unexplained stars” can not be the result of gravity bending light from stars behind the Sun. The Sun’s gravity is not strong enough to do something like that on that sort of scale. At best, the Sun bends the light from distant stars only enough to make them appear displaced by a few arc seconds from their normal position – definitely not enough to explain some of the Vulcan sightings during eclipses.

Al Jackson March 4, 2015 at 10:24

@Chakat Firepaw
” Actually a big reason for that belief was that Mercury’s orbit seemed to be wrong, it was doing something that implied another planet was periodically tugging on it.
Then along came Einstein and General Relativity which said that orbits really are supposed to do that.”

Here is an interesting thing. The long , really long, term integrations of the whole motion of the solar system , up to 5Gyr, show the solar system is more stable when General Relativity is included in the celestial mechanics.

Michael March 4, 2015 at 14:36

If there were planets within the orbit of mercury that fell into the Sun they would unlikely to have been gas or large rocky giants as the lithium content of the Sun is very low (1/140) when compared to the solar nebulae. Lithium been a fragile element is broken up at high enough temperatures and is depleted over time, a planet that fell into the Sun would have enriched the solar atmosphere closer to the solar nebulae amount.

andy March 4, 2015 at 15:51

There is apparently no room for the idea that the solar system initial
formation was not same as for the solar systems with additional rocky
worlds inside the Orbit Venus-Mercury.

Well several studies have indicated that the starting condition you need for producing the inner planets is a narrow ring of material spanning roughly the orbits of Venus to Earth: go much outside that range and you end up producing Mars- and Mercury-analogues that are too large. Question then becomes how to arrive at this initial condition: this is one of the reasons for advancing the Grand Tack hypothesis (inward then outward migration of Jupiter and Saturn) which would explain the outer cutoff. But you still need some explanation for the 0.7 AU inner cutoff, rather than running the disc all the way down to the inner magnetically-cleared cavity or the dust sublimation radius which are more “natural” places for the disc to stop, and observations of the exoplanet distribution do suggest that the typical case is that the disc gets down to such close distances from the star.

Mark Zambelli March 6, 2015 at 11:43

We’ve often heard that the 8 planets were built from around 100 or so planetoids (Ceres and Vesta being survivors and Theia being the last one incorporated into a planetary body). If this is roughly correct then surely some would’ve been orbiting within Mercury’s distance. As mentioned, Mercury’s iron ‘excess’ could be explained along these lines.

If planetary debris-discs do descend to near the star’s surface as a rule, then outside the Roche limit and away from the magnetic effects close-in then maybe any planetoids that managed to stay in that zone birthed a planet or two that have subsequently been lost?… perturbed to their demise either mutually or during any migrations that started.

I know planetary migrations can play a major role in shaping planetary systems but how major? There are so many variables when considerring migrations… planetary masses, number, rate of migration, how early or late the migrations started and how early or late they were halted, density of debris disc, magnetic and thermal environments due to the Sun/star-in-question… etc. I can imagine altering any one would give a different arrangement, all other things being equal.

I wonder then if a recipe-book for cooking planetary systems can even be constructed; whether we can find a set of patterns that will let us say, “See this infant system here… that will evolve to be like such and such an arrangement in yay-number of years.” I hope so. Even with the chaotic nature involved, it would be good to know how our solar system was constructed (other than calling Slartibartfast’s team of designers and seeing if they kept the original plans).

Steffen Christensen March 11, 2015 at 15:32

This model seems a lot more plausible when you add in the Bu and Wu research that the spacing of Kepler planets is best explained by assuming that all systems started densely packed (http://arxiv.org/pdf/1502.05449v3.pdf). Too-densely packed systems suffer subsequent dynamic instabilities leading to system repackaging, where the tightly packed planets loosen up over time. They call it dynamical sculpting. This model depends on the planet spacing following a pattern of 12-14 mutual Hill distances between adjacent planets. Intriguingly, it provides a natural explanation for systems like ours: our system, like most (~75-90%) formed too tightly packed for 5 billion year stability. The innermost planets collided, which interior to Mercury’s orbit, happens so fast that the planets got shredded instead of getting bigger and heavier. Other inner terrestrial planets got unstable due to resonances between Jupiter and Saturn, which cleared out the inner to spacings of 30-40 Hill radii per planet (not the initial 8-12). The Nice model and the Grand Tack still works to explain the asteroid belt and Mars’ orbit, and the 12-14 Hill radii spacing is observed for all the planets between Mars and Neptune. The only outlier is Jupiter-Saturn, but there, a clever trick saves us: all the other planets are lighter than Jupiter and Saturn. Therefore, their gravitational neighborhood is dominated by themselves and the Sun, and they can pack more tightly between themselves, since systems with only 2 members are stable down to 4-5 mutual Hill radii.

One more data set supporting this theory: Kepler systems have an innermost detected planet at an orbit of average 60 days, with a factor of 4-5 dispersion. So the Kepler data is telling us that most systems look like our system, and that the densely-packed systems with planets in ~5 day orbits are pretty rare.

George W September 16, 2015 at 9:33

Based on the work of Matthew J. Holman and Paul A. Wiegert (but does not fully address the reinforcement of near orbital resonance effects fully, nor the effects of inclined orbits), there very well could have been a planet interior to mercury however it would have had to have been closer then 0.13 au to the Sun to have had any sort of stability. This hypothetical body would have also been subjected to stong Kozai mechanism, Yarkovsky effect, Tidal dispersion effects from Jupiter and Saturn, and would ave been perturbed further by any resonances with Jupiter, Saturn, and Mercury. The likelihood of long term (age of the solar system) stability is doubtful under those conditions. The hypothetical body could have had a mass up to earth and an eccentricity less then 2.0 and have had short term stablity most likely being ejected (80+%) or colliding (1+%) with one of the inner planets (making it a potential candidate for Theia).

The work of Matthew J. Holman and Paul A. Wiegert also shows why there is no planet where the asteroid belt is located. The effects of Mars and Sun place
the outward limit of stability beyond 3.135 au, and the effects of Jupiter and Sun place an inward limit of stability interior to 2.255 au. Those limits place a pretty good fit +/- 10% for the asteroid belt where the inner edge is ~ 2.06 au and the outer edge is at 3.27 au.

Comments on this entry are closed.

{ 2 trackbacks }