Welcome to Second-Chance Planets: feel-good planets that overcome the odds to get a second chance at life…
Have you ever been lost in the desert?
Stumbling along the sand dunes, the Sun beating down on you, your mouth sticky and dry. All you can think about is water… A tough situation.
Water is pretty key for life. It makes up more than half of your body weight. And an even higher fraction in tasty beverages...
Which is why it’s so sad when a planet forms dry.
And it’s even sadder when a dry planet has other conditions that could be good for life, like an orbit in the habitable zone. The habitable zone is where a planet could potentially have liquid water! (cosmic frowny face).

Earth — our pale blue dot — is not a very wet planet.
Sure, oceans cover 70% of Earth’s surface. But when you consider the entire planet, Earth is only about 1 part in 1000 water.
Here is what Earth would look like if all its water was vacuumed up (maybe by megamaid?):

And Earth is actually the wettest of the four rocky planets in the Solar System.
Thousands of exoplanets have been found close to their host stars. Many are small enough that they could be rocky. Do these planets have water? We don’t know — it’s not something that’s easy to measure from light years away (yet).
We think that rocky exoplanets probably have a range of water contents. Some are probably much wetter than Earth, others much drier, and everything in between.
Today we’ll focus on the dry ones.
How does a rocky planet end up dry?
It can simply form dry, from solids like rock and iron that condensed in the hot parts of the planet-forming disk. We think Earth formed almost entirely from dry material with just a sprinkling of water-rich material.
Or a rocky planet can form wet and then lose its water.
Giant collisions with other growing planets can strip off some water but can’t completely dry out a planet.
What can dry out a planet is atmospheric loss driven by heating from the central star. Water molecules in the upper atmosphere are broken into hydrogen and oxygen atoms by energetic (UV and X-ray) photons from the star. Then some of the hydrogen is kicked hard enough to escape the planet’s gravity.
This is probably how Mars and Venus lost their water. A telltale sign is that any remaining water is dominated by heavy isotopes, as is the case for both planets.
This type of water loss is probably very common among rocky exoplanets.
The habitable zone is the ring of orbits around a given star in which a planet can maintain liquid water (if it has an atmosphere with the right properties). Around small, faint red dwarf stars the habitable zone is much closer-in than around Sun-like stars:

Planets in the habitable zones of small stars are vulnerable to losing their oceans.
Small stars take hundreds of millions of years to quiet down, and there is an early “active” phase during which young planets are baked. Many planets are likely to be completely dried out (see this post in the How planets die series).
Millions — or even billions — of rocky planets are likely to be roasted and dried out. How could they get a second chance at habitability?
Simply put, it needs to rain. Cosmic rain.
Water-rich objects — asteroids or comets — must crash down onto these dry planets and deliver water.
This bombardment of water-rich bodies must happen late in a planet’s history. Because if it happens early then the star will still be hyperactive and the fresh water will escape.

A late bombardment — what I’m poetically calling “cosmic rain” — happens in two stages.
First, an outer belt of planetary leftovers (asteroids and comets) survives long after the planets are done forming.
Second, that belt is destabilized and objects end up on planet-crossing orbits.
Outer belts of planetary leftovers are common. We detect them routinely around other stars from the cold dust that they produce. They are called debris disks and they exist around 20% of Sun-like stars.
One way that outer belts of leftovers are destabilized is when giant planets go unstable. Our own Solar System underwent an instability that cleared out its outer disk of leftovers (as I’ve written about before here, here and here):

The Solar System’s instability was long thought to have happened about half a billion years after the planets formed. It was called the Late Heavy Bombardment. This kind of delayed instability is perfect for cosmic rain (although Earth’s water was delivered much earlier and by a different process).
New analysis has found that the instability almost certainly happened much earlier, probably just after planet formation (or even during). Not as good for cosmic rain.
But some instabilities in systems of gas giant or ice giant planets may happen late.
And other processes can destabilize outer belts of leftovers. My personal favorite? Wide binary stars.
Wide binaries have a separation of 1-10 thousand AU. The two stars are only loosely gravitationally bound. Small gravitational kicks — from stars or dense clouds of gas that happen to pass relatively close-by — change the shapes of wide binary orbits.

On billion-year timescales, some wide binary stars end up on super-stretched out (highly eccentric) orbits. Their closest approach distance gets much closer, sometimes close enough to destabilize planetary systems or outer belts of leftover debris.
Then the leftovers scatter and collide with the planets. Cosmic rain!
Let’s dive into the numbers. Can cosmic rain deliver enough water?
It depends how much water a planet needs.
As a first example, imagine that Earth had formed completely dry (which it didn’t in real life). Let’s assume that all of Earth’s water was delivered during a late giant planet instability (even though we now think the instability happened early).
When the giant planets went unstable, those leftovers were almost entirely ejected from the Solar System, to become interstellar objects (like ‘Oumuamua and Borisov). Most leftovers wind up getting tossed out by gravitational kicks from Jupiter. Before they are ejected, some pass close to the rocky planets and can impact Earth.
Only about one leftover in a million collides with Earth. The Solar System’s outer belt of leftovers contained 20-30 Earth masses. That’s about 0.002% of Earth’s mass in cosmic rain. Assuming these objects have 10% water, that’s about 1% of the water on Earth’s surface delivered during the instability.
What would Earth look like if it had 1% of its surface water?
It’s possible that all the water might be sucked up into hydrated minerals in the mantle. No oceans or lakes.
If the water stayed on the surface, it would still make for a pretty sweet planet. The mean depth of Earth’s present-day oceans is 3.7 km (2.7 km if spread over the whole globe). Having 1% of the total water would be the equivalent of having an ocean-covered planet with a depth of 27 meters. Or, maintaining the same average depth as today, water would be spread over 1% of the surface — that’s the combined surface areas of India and Mexico.
Having relatively little water on a planet’s surface can stabilize its climate. This is called a Dune planet (described in Real-life Sci-Fi world number 5 here). Water vapor is a strong greenhouse gas so having less surface water prevents excess heating. Plus, desert planets cool off more efficiently than water-covered planets (and so do giraffes). These effects cause Dune planets to have wider habitable zones than wetter worlds.
So even a modest sprinkling of water onto a dry planet can make it a Dune world, which is not a bad place for life.

And cosmic rain may be more efficient in other situations.
Jupiter is the reason that only one in a million leftovers hit the Earth. Jupiter is so massive that it ejects leftovers efficiently, tossing them into interstellar space before many collide with Earth.
In systems without a Jupiter, a much higher fraction of leftovers would collide with Earth. No one has done the numbers (to my knowledge) but it could easily be 10-100 times higher. In some cases cosmic rain might deliver a full Earth’s worth of water.
(Of course, just how much cosmic rain also depends on whether a star has an outer belt of water-rich leftovers and how massive it is.)
Let’s estimate how many dry planets in our Galaxy are given a second chance by cosmic rain.
There are more than 100 billion stars in the Galaxy. The fraction with rocky planets in the habitable zone is up in the air, but let’s say it’s 10% (it could easily be a few times higher or lower). That’s 10 billion stars with Earths.
About 3/4 of all stars are small red M dwarfs, some of which have an early very active phase that may dry out their planets. That’s 1-5 billion dried-out rocky planets.
About 20% of Sun-like stars have detected debris disks, but the fraction is lower for M dwarfs. It’s uncertain whether there are outer leftovers that are just too faint/cold to detect. Let’s assume that 1-10% of all stars have long-lived disks of outer, comet-like leftovers from planet formation. That makes 10 to 500 million dry planets that are waiting for cosmic rain.
Now on what fraction of these systems will cosmic rain actually be late enough that water is retained? Most instabilities probably happen early, but the ~10% of stars with wide binaries have a delayed trigger. Let’s say 1-10% of instabilities happen later than 100 million years after the planets have stopped forming.
Put together, that makes between 100 thousand and 50 million dry worlds on which cosmic rain delivers water and gives the planet a second chance.
Boom!
A philosophical note: As we saw in How planets die, late giant bombardments are often destructive to life-bearing worlds. Here we have turned this on its head to show how late giant bombardments can breathe new life into previously dried-out worlds.
A final note: My all-time favorite song is Bob Dylan’s A hard rain’s a-gonna fall. If we’re lucky, that’s cosmic rain he’s singing about.
Resources
- The How planets die series, featuring How planets die: pulverized in a deluge of comets and asteroids! and How planets die: roasted, toasted and swallowed by their evolving stars!
- Blog post: Where did Earth’s (and the asteroids’) water come from?
- Real-life Sci-Fi World #5: a Dune planet (Arrakis)
- Some of the issues about planets orbiting small stars are discussed in Real-life Sci-For World #3: Earth around a brown dwarf.
- Review paper: Origin of Earth’s water — sources and constraints by Karen Meech and myself.
I have thought some about just how little water a life-bearing world could do with. Chris Wayan’s Tharn has been mentioned in connection with Arrakis. It has 25 % of the Earth’s mass and 0.1 – 0.2 % the amount of water. This is far from enough for the continent-sized bodies of water I would consider oceans. Yet the geological equivalents of ocean floors are spotted by lakes of various sizes, shapes and salinity.
I suppose the relative abundance of lakes on Tharn is due to its relatively small size. In fact, Chris Wayan made up Tharn to see how just how small and dry he could make a world which would still have a full ecosystem. With five to ten times as much as water Tharn could have seas covering much of its equvalent of ocean basain. But a world the size of the Earth with that little water would probably be limited to overzied lakes. A planet significantly larger than Earth could have continent-sized areas as barren as the McMurdo Dry Valleys. Not because it is too cold and windy, but because water is too scarce for macroscopic life.