Great Filters, Part 3: Planetary Filters

I've discussed galactic and stellar-scale Great Filters. Now it's time to look at them on a planetary level.

The first is the question of how stable planetary systems are. This is a problem which goes back to Isaac Newton. When Sir Isaac compiled his great work the Principia, he was bothered by one implication of his discoveries. There didn't seem to be any general solution to the "Three-Body Problem" — that is, how the interactions of multiple bodies in a planetary system would work. A general solution would show obvious regimes of stability; otherwise you'd just have a constant iterative process of constant perturbations with planets shuffling around and occasionally getting flung out into interstellar space.

The Italo-French mathematician Joseph-Louis Lagrange discovered one set of special conditions for the three-body problem, the famous "Lagrange Points" of stability around a moon or planet orbiting a larger body.

Before the discovery of exoplanets, the question of stable planetary systems was very much up in the air (beyond it, really). With only one sample, our own Solar System, we really had no way of knowing if we lived in a freak example of a stable planetary system, or if such things were commonplace.

The answer seems to lie between those two poles. There are lots of planetary systems, so many that it appears to be the norm. Only stars too young to have formed planets at all may lack them.

But the study of the orbits of those exoplanets has revealed a surprising amount of instability. Apparently planets migrate all over the place in stellar systems, shifting orbits, switching places, and in all likelihood getting kicked out into interstellar space. The stable systems that remain are simply the aftermath of chaos, smoothed down by billions of years.

So how is this a filter?

Well, we still need to figure out what proportion of planetary systems are unstable in the long term. Even if a potentially lifebearing world doesn't get tossed into space, jostling around its home system can't be a good thing for struggling life forms on its surface. We need to know how long planets are likely to remain in the habitable zones of their parent stars, where temperatures are warm enough for liquid water but not hot enough to boil away all the lighter elements.

This isn't an easy number to pick. Astronomers are still figuring out how planets form and how they interact over billions of years. Since we see lots of planetary systems, with plenty of worlds, instability can't be too great a problem. And many of our best methods for detecting exoplanets are biased toward weird systems — it's easier to detect a large world orbiting close to its parent star, and it's precisely those systems which are the product of lots of planetary migration and interaction.

I'm going to make a completely wild-assed guess and assume ten percent of star systems have chaotic interactions among their planets which make it unlikely for any of those worlds to develop life. In last week's post I winnowed the number of star systems which might be home to lifebearing worlds to about 10 billion in the Milky Way. This lops a billion off that figure.

Next comes the big one: how many Earthlike planets are there? By "Earthlike" I mean a rocky planet with liquid water and an atmosphere density somewhere between 1 percent and 200 percent of Earth's, orbiting in the habitable zone (sometimes called the "Goldilocks Zone" of its star system.

At present we know of some 4,000 exoplanets. Of those between 20 and 50 are in their star's habitable zone, depending on whether you take the optimistic or pessimistic view of limits on that zone. For simplicity's sake we'll say 1 percent of planets are in the habitable zone. For the Milky Way, that means 90 million worlds.

Now, some of those bodies are not exactly what you'd call "Earthlike," with masses more comparable to Neptune or even Saturn. It's hard to see how a giant planet like that, even with a rocky core, could avoid a lethal greenhouse effect. Roughly half the worlds on the list of exoplanets in the habitable zone of their parent stars have masses greater than 10 times that of Earth, which I consider beyond the pale. So our 90 million becomes 45 million.

After that comes a grab-bag of filters based on what we know of Earth's history, with absolutely no hard numbers to base our estimates on. These potential planetary filters include:

Giant Impacts: We believe the Moon was formed when Earth collided with a Mars-sized body early in its history. This may have affected Earth's chances of developing life, by blowing off the planet's original atmosphere and preventing Earth from winding up like Venus. That's highly speculative, and it's also unclear how common such impacts might be. A lot of airless bodies with ancient surface features show really big impact basins, so it's quite possible that almost planet gets walloped by something big in its early years. For lack of any hard numbers I'll call that one a coin toss: 50 percent.

Moons: It's been suggested that the Moon makes Earth a home for life by stabilizing the planet's rotation and axial tilt. I'm not sure I buy it — it sounds a lot like correlation (Earth has a big moon, Earth has life) being turned into causation (Earth's big moon is responsible for Earth having life). Since moon formation appears to be tied to impacts, I'm just going to fold this one into the previous filter.

Snowball Earth: When plant life evolved on Earth and began transforming the planet via toxic waste products, one side effect was a drastic drop in temperature as the plants sucked up all of the atmospheric carbon dioxide, reducing the greenhouse effect keeping Earth warm. Earth's surface was covered by ice, with life surviving only in the oceans under the icecap. The "Cryogenian" era may have lasted up to 200 million years, ending only when the Sun's gradual increase in temperature overcame the Earth's cooling.

But what if Earth had been at the outer edge of the habitable zone, or the Sun's output wasn't increasing as rapidly? Snowball Earth could have lasted hundreds of millions of years longer, or been permanent. To me, this one seems like an important filter, one likely to repeat elsewhere. How to estimate the outcome of two independent processes is very hard, so I'm just going to make this one another coin-flip: 50 percent.

Waterworlds: Earth's surface is 75 percent ocean. Raise that sea level by just a couple of miles and the figure would be close to 100 percent. If we assume that more massive worlds retain more light elements, then any world more than, say, two or three times Earth's mass might be entirely covered with water. This wouldn't prevent life from evolving, but might well prevent technological civilizations from arising. While it is possible that an aquatic alien civilization might come up with ways to communicate across interstellar distances before they discover fire or electricity, I'd hardly call it likely. That one's another coin-flip.

Summing up these guesswork filters gives us 1 in 8, which I'm going to adjust to 1 in 10 for simplicity, and to cover any other planetary filters I didn't think of. Applying that to our existing figure we get 4.5 million worlds in the Milky Way which might be home to life and civilization. That's still a lot, but it's certainly less than the 10 billion figure we had at the end of my last post.

It's also cause for some limited optimism. Our world is already a .0045 percent long shot. But with literally millions of possible inhabited worlds, we're certainly not out of the (dark) woods yet. Where is everybody?

Next time: Life Filters.

Comments

[MadRocketSci]

How much do we really know about other planetary systems? When I last looked into it, most of those planets were discovered via the transit method by the Kepler satellite. That means we're looking at periodic dips in the light from another star. Kepler didn't operate for very long, and so we only have information about very close-orbiting planets.

Also, I think we only called something a planet if the dips could be measured as having a definite period. (Otherwise it could easily be a sunspot or other chaotic event). That by definition restricts our detections to planets that have unvarying orbits (to within the precision of the transit method)!

There are a few weird stars where we see fluctuations that look random, but the "norm" (tongue in cheek given the limitations of our methods) seems to be very closely packed systems around class M stars.

[Cambias]

You're quite correct, MadRocketSci. We still don't know much about other planetary systems. As I said above the bear picture, our best methods are biased toward weird systems.

Expanding on your remarks a little: obviously dips in brightness can only be attributed to a planet if they show a regular cycle. This means a strong bias in favor of worlds close to the parent star. You'd have to observe for several centuries to detect Saturn that way. Even Jupiter would take a few decades to nail down.

However I don't think there are many planets with variable periods as you suggest. It's hard to see how that arrangement could last very long on a stellar time scale. I'm sure a physicist could model some system where a planet has a resonant orbit with some other body in the solar system such that it alternates long and short years, but that would likely be an almost unique phenomenon.