# How do Rocky Planets Form?

A question in my last post raised the question of how do rocky planets form, and why is Venus so different from Earth? This will take two posts; the first covers how the planets form and why, and the second how they evolve immediately after formation and get their atmospheres.

First, a quick picture of accretion. At first, the gas cloud collapses and falls into the star, and in this stage the star the size of the sun accretes something like 2.5 x 10^20 kg per second. Call that stage 1. When the star has gobbled up most of the material, such accretion slows down, and in what I shall call stage 2 it accretes gas at least four orders of magnitude slower. The gas heats due to loss of potential energy as it falls into the star, although it also radiates heat from the dust that gets hot. (Hydrogen and helium do not radiate in the infrared easily.) In stage 1, the gas reached something like 1600 degrees C at 1 A.U. (the distance from Earth to the sun). In stage 2, because far less gas was falling in, the disk had temperatures roughly what bodies have now. Even in stage 2, standard theory has it that boulder-sized objects will fall into the star within about a hundred years due to friction with the gas.

So how did planets form? The standard explanation is that after the star had finished accreting, the dust very rapidly accreted to planetesimals (bodies about 500 km across) and these collided to form oligarchs, and in turn these collided to form planets. I have many objections to this. The reasons include the fact there is no mechanism to form the planetesimals that we assume to begin with. The calculations originally required one hundred million years (100 My) to form Earth, but we know that it had to be essentially formed well before that because the collision that formed the Moon occurred at about 50 My after formation started. Calculations solved the Moon-forming problem by saying it only took 30 My, but without clues why this time changed. Worse, there are reasons to believe Earth had to form within about 1 My of stage 2 because it has xenon and krypton that had to come from the accretion disk. Finally, in the asteroid belt there is evidence of some previous collisions between asteroids. What happens is they make families of much smaller objects. In short, the asteroids shatter into many pieces upon such collisions. There is no reason to believe that similar collisions much earlier would be any different.

The oldest objects in the solar system are either calcium aluminium inclusions or iron meteorites. Their ages can be determined by various isotope decays and both had to be formed in very hot regions. The CAIs are found in chondrites originating from the asteroid belt, but they needed much greater heat to form than was there in stage 2. Similarly, iron meteorites had to form at a temperature sufficient to melt iron. So, how did they get that hot and not fall into the sun? The only time the accretion disk got sufficiently hot at a reasonable distance from the sun was when the star was accreting in stage 1. In my opinion, this shows the calculations were wrong, presumably because they missed something. Worse, to have enough material to make the giants, about a third of the stellar mass has to be in the disk, but observation of other disks in stage 2 shows there is simply not enough mass to make the giants.

The basic argument I make is that whatever was formed in the late stages of stellar accretion stayed more or less where it was. One of the puzzles of the solar system is that most of the mass is in the star, but most of the angular momentum resides in the planets, and since angular momentum has to be conserved and since most of that was with the gas initially, my argument is any growing solids took angular momentum from the gas, which sends then mass further from the star, and it had to be taken before the star stopped accreting. (I suggest a mechanism in my ebook.)

Now to how the rocky planets formed. During primary stellar accretion, temperatures reached about 1300 degrees C where Mars would form and 1550 degrees C a little beyond where Earth would grow. This gives a possible mechanism for accretion of dust. At about 800 degrees C silicates start to get sticky, so dust can accrete into small stones there, and larger ones closer to the star. There are a number of different silicates, all of which have long polymers, but some, especially aluminosilicates are a little more mobile than others. At about 1300 degrees C, calcium silicate starts to phase separate out, and about 1500 degrees C various aluminosilicates phase separate. This happens because the longer the polymer, the more immiscible it is in another polymer melt (a consequence of the first two laws of thermodynamics, and which makes plastics recycling so difficult.) If this were the only mechanism for forming rocky planets, the size of the finished planet would diminish significantly with distance from the star. Earth, Venus and Mercury are in the wrong order. Mercury may have accreted this way, but further out, stones or boulders would be the biggest objects.

Once primary stellar accretion ends, temperatures were similar to what they are now. Stones collide, but with temperatures like now, they initially only make dust. There is no means of binding silicates through heat. However, if stones can come together, dust can fill the spaces. The key to rocky planet formation is that calcium silicate and calcium aluminosilicates could absorb water vapour from the disk gases, and when they do that, they act as cements that bind the stones together to form a concrete. The zone where the aluminosilicates start to get formed is particularly promising for absorbing water and setting cement, and because iron starts to form bodies here, lumps of iron are also accreted. This is why Earth has an iron core and plenty of water. Mars has less water because calcium silicate absorbs much less water, and its iron is mainly accreted as fine dust.

Finally, Mars is smaller because the solids density is less, and the disk is cleared before it has time to fully grow. The evidence for the short-lived disk is from the relatively small size of Jupiter compared with corresponding planets around similar sized stars that our sun cleared out the accretion disk sooner than most. This is why we have rocky planets, and not planets like the Neptune-sized planets in the so-called habitable zone around a number of stars. Venus is smaller than Earth because it was harder to get going, through the difficulty of water setting the cement, which is partly why it has very little water on its surface. However, once started it grows faster since the density of basaltic rocks is greater. Mercury is probably smaller still because it formed a slightly different way, through excessively mobile silicates in the first stage of the accretion disk, and by later being bombed by very large rocky bodies that were more likely to erode it. That is somewhat similar to the standard explanation of why Mercury is small but has a large iron core. The planets grow very quickly, and soon gravity binds all dust and small stones, then as it grows, gravity attracts objects that have grown further away, which perforce are large, but still significantly smaller than the main body in the zone.

Next post: how these rocky planets started to evolve to where they are now.

# A Giant Planet Around a Dwarf Star

The news here, at least, has made much of the discovery of NGTS-1b, described as a giant planet orbiting a dwarf star. It is supposed to be the biggest planet ever found around such a small star, and it is supposed to be inexplicable how such a big planet could form. One key point that presumably everyone will agree with is, a small star forms because there is less gas and dust in the cloud that will form the star than in the cloud that forms a big star. Accordingly there is less total material to form a planet. Missing from that statement is the fact that in all systems the amount of mass in the planets is trivial compared to the mass of the star. Accordingly, there is nothing at all obscure about an unexpectedly big planet if the planet was just a bit more efficient at taking material that would otherwise go into the star.

So, a quick reality check: the star is supposed to be about 60% the size of the sun, and the planet is about 80% the mass of Jupiter, but has a somewhat larger radius. Planets up to twenty times the size of Jupiter are known around stars that are not more than about three times the size of our sun, so perhaps there is more being made of this “big planet” than is reasonable.

Now, why is it inexplicable how such a large planet could form around a small star, at least in standard theory? The mechanism of formation of planets in the standard theory is that first gas pours in, forms the star, and leaves a residual disk (the planetary accretion disk), in which gas is essentially no longer moving towards the star. That is not true; the star continues to accrete, but several orders or magnitude more slowly. The argument then is that this planetary accretion disk has to contain all the material needed to form the planets, and they have to form fast enough to get as big as they end up before the star ejects all dust and gas, which can take somewhere up to 10 million years (10 My), with a mean of about 3 My. There is some evidence that some disks last at least 30 My. Now the dust collides, sticks (although why or how is always left out in the standard theory) and forms planetesimals, which are bodies of asteroid size. These collide and form bigger bodies, and so on. This is called oligarchic growth. The problem is, as the bodies get larger, the distance between them increases and collision probability falls away, not helped by the fact that the smaller the star, the slower the orbiting bodies move, the less turbulent it will be, so the rate of collisions slows dramatically. For perspective purposes, collisions in the asteroid belt are very rare, and when they occur, they usually lead to the bodies getting smaller, not bigger. There are a modest number of such families of detritus asteroids.

The further out the lower the concentration of matter, simply because there is a lot more space. A Jupiter-sized body has to grow fast because it has to get big enough for its gravity to hold hydrogen, and then actually hold it, before the disk gases disappear. Even accreting gas is not as simple as it might sound, because as the gas falls down the planetary gravitational field, it gets hot, and that leads to some gas boiling off back to space. To get going quickly, it needs more material, and hence a Jupiter type body is argued (correctly, in my opinion) to form above the snow line of water ice. (For the purposes of discussion, I shall call material higher up the gravitational potential “above”, in which case “below” is closer to the star.) It is also held that the snow line is not particularly dependent on stellar mass, in which case various planetary systems should scale similarly. With less material around the red dwarf, and as much space to put it in, everything will go a lot slower and the gas will be eliminated before a planet is big enough to handle it. Accordingly, it seems that according to standard theory, this planet should not form, let alone be 0.036 A.U. from the star.

The distance from the star is simply explained in any theory: it started somewhere else and moved there. The temperature at that distance is about 520 degrees C, and with solar wind it would be impossible for a small core to accrete that much gas. (The planet has a density of less than 1, so like Saturn it would float if put in a big enough tub of water.) How would it move? The simplest way would be if we imagined a Jupiter and a Saturn formed close enough together, when they could play gravitational billiards, whereby one moves close to the star and the other is ejected from the system. There are other plausible ways.

That leaves the question of how the planet forms in the first place. To get so big, it has to form fast, and there is evidence to support such rapid growth. The planet LkCa 15b is around a star that is slightly smaller than the sun, it is three times further out than Jupiter, and it is five times bigger than Jupiter. I believe this makes our sun special – the accretion disk must have been ejected maybe as quickly as 1 My. Simulations indicate that oligarchic growth should not have led to any such oligarchic growth that far out. My explanation (given in my ebook “Planetary Formation and Biogenesis”) is that the growth was actually monarchic. This is a mechanism once postulated by Weidenschilling, in 2004 (Weidenschilling, S., 2004. Formation of the cores of the outer planets. Space Science Rev. 116: 53-56.) In this mechanism, provided other bodies do not grow at a sufficient rate to modify significantly the feed density, a single body will grow proportionately to its cross-sectional area by taking all dust that is in its feed zone, which is augmented by gravitation. The second key way to get a bigger planet is to have the planetary accretion disk last longer. The third is, in my theory, the initial accretion is chemical, and the Jupiter core forms like a snowball, by water ice compression fusing. Further, I argue it will start even while the star is accreting. That only occurs tolerably close to the melting point, so it is temperature dependent. The temperatures are reached very much closer to the star for a dwarf. Finally, the planet forming around a dwarf has one final growth advantage: because the star has a lower gravity, the gas will be drifting towards the star more slowly, so the growing planet, while having a less dense feed, also receives a higher fraction of the feed.

So, in my opinion, apart from the fact the planet is so lose to the star, so far there is nothing surprising about it at all, and the mechanisms for getting it close to the star are there, and there are plenty of other “star-burning” planets that have been found.

Why has the monarchic growth concept not taken hold? In my opinion, this is a question of fashion. The oligarchic growth mechanism has several advantages for the preparation of scientific papers. You can postulate all sorts of initial conditions and run computer simulations, then report those that make any sense as well as those that don’t (so others don’t waste time.) Monarchic growth leaves no real room for scientific papers.

# Star and Planetary Formation: Where and When?

Two posts ago, as a result of questions, I promised to write a post outlining the concept of planetary accretion, based on the current evidence. Before starting that, I should explain something about the terms used. When I say something is observed, I do not mean necessarily with direct eyesight. The large telescopes record the light signals electronically, similarly to how a digital camera works. An observation in physics means that a signal is received that can be interpreted in one only certain way, assuming the laws of physics hold. Thus in the famous two-slit experiment, if you fire one electron through the slits, you get one point impact, which is of too low an energy for the human eye to see. Photomultipliers, however, can record this as a pixel. We have to assume that the “observer” uses laws of physics competently.

The accretion of a star almost certainly starts with the collapse of a cloud of gas. What starts that is unknown, but it is probably some sort of shock wave, such as a cloud of debris from a nearby supernova. Another cause appears to be the collision of galaxies, since we can see the remains of such collisions that are accompanied by a large number of new stars forming. The gas then collapses and forms an accretion disk, and these have been observed many times. The gas has a centre of mass, and this acts as the centre of a gravitational field, and as such, the gas tries to circulate at an orbital velocity, which is where the rate of falling into the star is countered by the material moving sideways, at a rate that takes it away from the star so that the distance from the centre remains the same. If they do this, angular momentum is also conserved, which is a fundamental requirement of physics. (Conservation of angular momentum is why the ice skater spins slowly with arms outstretched; when she tucks her arms in, she spins faster.

The closer to the centre, the strnger gravity requires faster orbital velocity. Thus a stream of gas is moving faster than the stream just further from the centre, and slower than the stream just closer. That leads to turbulence and friction. Friction slows the gas, meaning it starts to fall starwards, while the friction converts kinetic energy to heat. Thus gas drifts towards the centre, getting hotter and hotter, where it forms a star. This has been observed many times, and the rate of stellar accretion is such that a star takes about a million years to form. When it has finished growing, there remains a dust-filled gas cloud of much lower gas density around it that is circulating in roughly orbital velocities. Gas still falls into the star, but the rate of gas falling into the star is at least a thousand times less than during primary stellar accretion. This stage lasts between 1 to 30 million years, at which point the star sends out extreme solar winds, which blow the gas and dust away.

However, the new star cannot spin fast enough to conserve angular momentum. The usual explanation is that gas is thrown out from near the centre, and there is evidence in favour of this in that comets appear to have small grains of silicates that could only be formed in very hot regions. The stellar outburst noted above will also take away some of the star’s angular momentum. However, in our system, the bulk of the angular momentum actually resides in the planets, while the bulk of the mass is in the star. It would seem that somehow, some angular momentum must have been transferred from the gas to the planets.

Planets are usually considered to form by what is called oligarchic growth, which occurs after primary stellar accretion. This involves the dust aggregating into lumps that stick together by some undisclosed mechanism, to make first, stone-sized objects, then these collide to form larger masses, until eventually you get planetesimals (asteroid-sized objects) that are spread throughout the solar system. These then collide to form larger bodies, and so on, at each stage collisions getting bigger until eventually Mars-sized bodies collide to form planets. If the planet gets big enough, it then starts accreting gas from the disk, and provided the heat can be taken away, if left long enough you get a gas giant.

In my opinion, there are a number of things wrong with this. The first is, the angular momentum of the planets should correspond roughly to the angular momentum of the dust, which had velocity of the gas around it, but there is at least a hundred thousand times more gas than dust, so why did the planets end up with so much more angular momentum than the star? Then there is timing. Calculations indicate that to get the core of Jupiter, it would take something approaching 10 million years, and that assumes a fairly generous amount of solids, bearing in mind the solid to gas ratio. At that point, it probably accretes gas very quickly. Get twice as far away from the star, and collisions are much slower. Now obviously this depends on how many planetesimals there are, but on any reasonable estimate, something like Neptune should not have formed. Within current theory, this is answered by having Neptune and Uranus being formed somewhere near Saturn, and then moved out. To do that, while conserving angular momentum, they had to throw similar masses back towards the star. I suppose it is possible, but where are the signs of the residues? Further, if every planet is made from the same material, the same sort of planet should have the same composition, but they don’t. The Neptune is about the same size as Uranus, but it is about 70% denser. Of the rocky planets, Earth alone has massive granitic/feldsic continents.

Stronger evidence comes from the star called LkCa 15 that apparently has a gas giant forming that is already about five times bigger than Jupiter and about three times further away. The star is only 3 million years old. There is no time for that to have formed by this current theory, particularly since any solid body forming during the primary stellar accretion is supposed to be swept into the star very quickly.

Is there any way around this? In my opinion, yes. I shall put up my answer in a later post, although for those who cannot wait, it is there in my ebook, “Planetary Formation and Biogenesis”.

# Theory and planets: what is right?

In general, I reserve this blog to support my science fiction writing, but since I try to put some real science in my writing, I thought just once I would venture into the slightly more scientific. As mentioned in previous posts, I have a completely different view of how planets, so the question is, why? Surely everyone else cannot be wrong? The answer to that depends on whether everyone goes back to first principles and satisfies themselves, and how many lazily accept what is put in front of them. That does not mean that it is wrong, however. Just because people are lazy merely makes them irrelevant. After all, what is wrong with the standard theory?

My answer to that is, in the standard theory, computations start with a uniform distribution of planetesimals formed in the disk of gas from which the star forms. From then on, gravity requires the planetesimals to collide, and it is assumed that from these collisions, planets form. I believe there are two things wrong with that picture. The first is, there is no known mechanism to get to planetesimals. The second is that while gravity may be the mechanism by which planets complete their growth, it is not the mechanism by which it starts. The reader may immediately protest and say that even if we have no idea how planetesimals form, something had to start small and accrete, otherwise there would be no planets. That is true, but just because something had to start small does not mean there is a uniform distribution throughout the accretion disk.

My theory is that it is chemistry that causes everything to start, and different chemistries occur at different temperatures. This leads to the different planets having different properties and somewhat different compositions.

The questions then are: am I right? does it matter? To the first, if I am wrong it should be possible to falsify it. So far, nobody has, so my theory is still alive. Whether it matters depends on whether you believe in science or fairy stories. If you believe that any story will do as long as you like it, well, that is certainly not science, at least in the sense that I signed up to in my youth.

So, if I am correct, what is the probability of finding suitable planets for life? Accretion disks last between 1 to even as much as 30 My. The longer the disk lasts, the longer planets pick up material, which means the bigger they are. For me, an important observation was the detection of a planet of about six times Jupiter’s mass that was about three times further from its star (with the name LkCa 15) than Jupiter. The star is approximately 2 My old. Now, the further from the star, the less dense the material, and this star is slightly smaller than our sun. The original computations required about 15 My or more to get Jupiter around our star, so they cannot be quite correct, although that is irrelevant to this question. No matter what the mechanism of accretion, Jupiter had to start accreting faster than this planet because the density of starting material must be seriously greater, which means that we can only get our solar system if the disk was cleared out very much sooner than 2 My. People ask, is there anything special regarding our solar system? I believe this very rapid cleanout of the disk will eliminate the great bulk of the planetary systems. Does it matter if they get bigger? Unfortunately, yes, because the bigger the planets get, the bigger the gravitational interactions between them, so the more likely they are to interact. If they do, orbits become chaotic, and planets can be eliminated from the system as other orbits become highly elliptical.

If anyone is interested in this theory, Planetary Formation and Biogenesis (http://www.amazon.com/dp/B007T0QE6I )

will be available for 99 cents  as a special promo on Amazon.com (and 99p on Amazon.co.uk) on Friday 13, and it will gradually increase in price over the next few days. Similarly priced on Friday 13 is my novel Red Gold, (http://www.amazon.com/dp/B009U0458Y  ) which is about fraud during the settlement of Mars, and as noted in my previous post, is one of the very few examples of a novel in which a genuine theory got started.

# Planets for alien life (2)

My last post gave an estimate of how many stars were suitable for having planets with life, if they had rocky planets in the right place. The answer comes out very roughly as one per every five hundred cubic light years. At first sight, not very common, but galaxies are very big, and we end up with about a hundred billion in this galaxy. The next question is, are there further restrictions? Extrasolar planets are reasonably common, according to recent surveys, however most of these found are giants that are very close to the star, and totally unsuited for life. On the other hand, there is a severe bias: the two methods that have yielded the most discoveries favour the finding of large planets close to the star.

To form stars, a large volume of gas begins to collapse, and as it collapses to form a star, it also forms a spinning disk. Three stages then follow. The first stage involves gas falling into the star from an accretion disk at a rate of a major asteroid’s mass each second. The second involves a much quieter stage, where the star has essentially formed, but it still has a disk, which it is accreting at a much slower rate, about a thousandth as fast. Finally, the star has “indigestion” and in a massive burp, clears out what is left of the disk (technically called a T Tauri event). The standard theory has the planets forming in the second stage or, for rocky planets, even following the T Tauri cleanout.

There are two important issues. As the gas falls into the star, both energy and angular momentum must be conserved. The fate of energy is simple: as the gas falls inwards, it gets hotter, and it is simple gravitation that heats the star initially, until it reaches about 80 million degrees, at which point deuterium starts to fuse and this ignites stellar fusion. However, the issue with angular momentum is more difficult. This is like an ice skater – as she brings her arms closer to herself, she starts spinning faster; put out her arms and the spin slows. As the gas heads into the star, the star should spin faster. The problem is, almost all the mass of the solar system is in the star, but almost all the angular momentum is in the planets. How did this happen?

Either all the mass retained its original angular momentum or it did not. If it did, then the sun should be spinning at a ferocious rate. While it could have lost angular momentum by throwing an immense amount of gas back into space, nobody has ever seen this phenomenon. If the stellar mass did not retain its angular momentum, it had to exchange it with something else. In my opinion, what actually happened is that the forming planets took up the angular momentum from gas that then fell into the star. If that is true, every star with enough heavy elements will form planets of some description because it helps stellar accretion. If so, the number of planet-bearing stars is very close to the number of stars.

There is, however, another problem. In my theory (Planetary Formation and Biogenesis for more details) planets simply keep growing until the stage 3 disk clear-out. If they get big enough, mutual gravitational interactions disrupt their orbits and something like billiards occurs. The planets do not collide, but if they come close enough one will be thrown out of the system (astronomers have already detected planets floating around in space, unattached to any star) and the other will end up as a giant very close to the star. A considerable number of such systems have been found. This would totally disrupt Earth-like planets, so stars with planets suitable for life must have had a shorter stage 2.

How short? Stage 2 can last up to 30 million years, although that is probably an exception, while the shortest stage 2 is less than a million years. The answer is, probably no more than a million years, i.e. our planetary system was formed around a star that had a relatively short secondary accretion. The reason I say that is as follows. The rate of accretion of a gas giant should be proportional to how much gas there is around it, and for how long. The amount of gas decreases as the distance from the star increases, and if you double the distance from the star, the gas density decreases somewhere between a half and a quarter. Now the three million year old star LkCa 15 is slightly smaller than our sun but it still has a second stage gas disk. This star has a planet nearly five times as big as Jupiter about three times further away from the star. This almost certainly means that Jupiter must have stopped growing well within three million years. (As an aside, standard theory requires at least 15 million years to start a gas giant.) Fortunately, it appears that about half the stars have such a short secondary stage. If we then say that about half the stars will be in the wrong part of the galaxy, then the estimate of stars that could be suitable for life reduces to about 25 billion. If we further reduce the total by those that are simply too young, or do not have sufficient metallicity, we could reduce the total to about 10 billion. These numbers are very rough, but the message remains: there are plenty of stars suitable to sustain life-bearing planets in the galaxy. The next question is, how many stars will have rocky planets?