An Apple Pie From Scratch, Part I: Time and Place

NASA

The Universe, as we all well know, is vast and old.

Or so it seems to us, anyway; 13.8 billion years since the beginning sounds like a long time,but given the trillions of years that the star-forming epoch is predicted to last, the universe might as well be wearing gigantic cosmic diapers. And the end of stars doesn’t mean the end of the universe either: According to current research, the most likely fate of the universe is neither to stop expanding and collapse back into a singularity (the “Big Crunch” scenario) or expand so fast that it tears even atoms apart (the “Big Rip” scenario), but rather to continue expanding at a moderate rate as entropy causes energy to move to ever more disordered forms (the “Big Freeze” scenario). Some people speak of the universe reaching “heat death” when no matter remains and all energy exists as evenly-spread heat, but really there’s no definitive end to universe. It just…keeps going. Forever.

So, given literally infinite time to work with, it’s a bit curious that the vast majority of science fiction seems to take place about now, with the universe looking much as it does today. If humanity appears in the story at some point then it does make sense, but if not then there’s no particular reason a story has to take place today; or even if it does, it doesn’t have to occur in the same sort of cosmic environment.

Of course, if we’re building a new world from the start, with life appearing naturally and evolving to intelligence, then that does require some conditions—a healthy mix of elements, a planet, the right kind of star—that can only occur within a certain time frame. That’s what I want to explore today: When—and where—can we expect complex, intelligent life to have a reasonable chance of appearing in the universe?

• The Past: Dawn of the Habitable Universe
• The First Stars
• The Green Cosmos
• Sources of Cosmic Radiation
• The Present: Habitability in the Universe Today
• The Future: Decline and Fall of the Habitable Universe
• Stelliferous Era (now-10¹⁴ years)
• Degenerate Era (10¹⁴-10³⁰ years)
• Black Hole Era (10³⁰-10¹⁰⁰ years) and Beyond
• So Anyway
• In Summary
• Notes

The Past: Dawn of the Habitable Universe

So far as we know, it all begins with the Big Bang. The name is often misunderstood to imply an explosion, but that’s not a good description; it was caused not by the movement of objects through space, but by the uniform expansion of space itself. Think of it like increasing the size of an image on your computer: the distance between objects is increasing, but nothing is really moving. 

As the universe expands, the density of matter and energy rapidly drops and the temperature drops with it, allowing progressively larger and more complex types of matter to form without being immediately torn apart. From the start it takes less than 1 second for all the subatomic particles to form, around 10 minutes for protons and neutrons to start joining into small atomic nuclei, and 380,000 years for electrons to start binding to nuclei and forming stable atoms. For us, it was then 9.3 billion years until the sun and earth formed, and another 4.5 billion until humans turned up.

 

But we’re impatient. Do we really have to wait that long?

As it turns out, perhaps not. When the plasma of the early universe cools enough to form atoms, this allows photons to move freely for the first time without being scattered by loose electrons, and so they are emitted all across the universe in all directions. This is an event known as the “last scattering”, and today we can see photons that have traveled undisturbed since then as the cosmic microwave background (CMB).

Because this primordial radiation is ubiquitous, it warms all exposed surfaces in space to a minimum temperature—initially around 3000 Kelvin, but over time the expansion of the universe has caused the photons to redshift, increasing their wavelength (causing visible light to shift from blue to red, hence the name) and decreasing their energy, as well as spreading individual photons apart. Today this background temperature is only 2.7 K, but in the period between 10 and 17 million years after the Big Bang, it would have been between 273 and 373 K, the range for liquid water at 1 atm pressure. This means that if there were any planets with water at that time, they would have been at the correct temperature to form life¹, even with no nearby stars to warm them.

But that’s a big if. When the first atomic nuclei form, the conditions of the early universe² and various forces at work within the nuclei only allow for the formation of a few possible configurations of nucleons (protons and neutrons). The majority of protons remain on their own as simple hydrogen nuclei; all heavier nuclei require the inclusion of neutrons, and at the time nucleosynthesis—the formation of new nuclei—starts, there are only 1/7 as many neutrons as protons (protons can sometimes turn into neutrons by electron capture or beta decay but this happens much too infrequently to matter in this short a time).

 

A minority of protons pair with a neutron to form deuterium, and then another neutron can be added to form tritium, or a proton added to form the universe’s first helium, Helium-3. Two protons and two neutrons form Helium-4, the most common variety of helium today.

"n" = neutron, "p" = proton, "D" = deuterium, "T" = tritium, "y" = gamma radiation. Pamputt, Wikimedia

But then the process hits a wall: There are no stable nuclei with 5 or 8 nucleons, and the only long-term stable nuclei with 7 or 6 nucleons, Lithium-7 and Lithium-6, form in small and vanishingly small amounts, respectively (some Beryllium-7 forms as well, but it’s unstable and later decays to Lithium-7). In stars this gap is bridged by the formation of Carbon-12 from the combination of three ⁴He nuclei, but this happens so infrequently that it takes thousands of years for stars to accrue a significant ¹²C stock. Primordial nucleosynthesis only occurs during a 20-minute window, after which the universe has expanded and cooled too much for spontaneous fusion. A few scattered carbon and heavier nuclei may form, but not enough to have a realistic chance of coming together to perform complex chemistry.

All of which is to say that by the end of the universe’s first half-hour, its matter complement consists of 76% hydrogen, 24% helium, a skosh of lithium (and beryllium), and not much else. No water, no rock, and even after the last scattering almost no molecules more than a couple atoms large. It will take until the formation of the first stars for heavier elements to form in any appreciable quantity, and then these stars have to die and disgorge those metals into the cosmos before water or rocky planets can form and life appear on them.

The First Stars

So when do stars first appear? The oldest known star³ is HD 140283, aka the “Methusalah Star”, with an estimated age of 14.46 billion years. You’ll note this is older than the estimated age of the universe, 13.8 billion years, but that’s more likely due to errors in estimating the age of the star than in estimating the age of the universe. Indeed, the reported error on that estimate is 0.8 billion years, meaning it could have formed as much as 140 million years after the big bang (though we should treat even that figure with little confidence).

Methusalah. ESA/Hubble

A better estimate comes from more indirect evidence. When the first stars form, their radiation affects the surrounding hydrogen gas, causing it to absorb some frequencies of light and leaving gaps in the CMB that we can still see today. These gaps are then redshifted with the rest of the CMB, and so based on where they must have started (based on the properties of hydrogen) and the amount of redshift they must have experienced to get where they are today, we can estimate⁴ that they first appear 180 million years after the Big Bang. But really this just tells us when stars first become common enough to have a measurable effect on the scale of the universe; it’s perfectly possible that a few precocious upstarts appear far earlier but are too scattered to leave an observable signal on the CMB.

So forget observational data—when do our theoretical models of the early universe say the first stars could form? In order for stars to form, they require that a large mass of hydrogen and helium cool and condense enough to form a high-density core where fusion can begin. In the current universe the presence of heavy elements aids in this cooling process (they absorb heat from the H/He and radiate it outwards a lot more efficiently), but without them the first stars can form only with the aid of massive dark matter “halos” over 10,000 times the mass of our sun, compressing huge amounts of gas. These halos take some time to form and condense, and a combination of computer modelling and analytical statistics⁵ predicts that the very earliest stars in the currently observable universe likely appear around 30 million years after the Big Bang.

By this point the CMB has cooled to just 175 K, too cold to thaw a planet on its own. Even if we speculated that in the very densest regions of the early universe, way off the end of the bell curve, maybe some isolated stars might form in that earlier “habitable” epoch, it wouldn’t guarantee life. These first stars are quite massive—50 to 100s of times the sun’s mass—and so only last a few million years. They helpfully scatter some heavy elements into the surrounding space when they die, but even if another star formed immediately, the density of metals would have been low enough to make the formation of a large rocky planet exceedingly unlikely (not to mention that formation of such planets likely takes 10s to 100s of millions of years even in ideal circumstances). 

 

Worse, the high temperature of the CMB may actually inhibit the formation of planets (for much the same reason as stars have trouble forming; the gas clouds they form from have to lose heat to condense down to dense objects), such that even gaseous planets may not be possible⁶ until over 100 million years after the Big Bang. And even if we're unreasonably optimistic and assume a planet could form at this point, 7 million years just isn't enough time for much to happen.

 

But let's not lose all hope. For one, the CMB need not be the only source of heat in this scenario; some early planets partially warmed by their star and internal heat may still get additional warming from the CMB long past 17 million years after the Big Bang. For another, life may not be restricted to only water as a solvent (we'll mostly be assuming it is for these early posts, but I'll explore other possibilities later): Hydrocarbon or nitrogen oceans could be kept above their freezing points by the CMB⁷ until over 100 million years after the big bang, and hydrogen oceans could last over 400 million years.

 

So life in the cosmos may have missed that early "habitable epoch" just after the big bang, but an early appearance of life somewhere in the low 100s of millions of years after the Big Bang is not out of the question.

The Green Cosmos

Let’s leave behind unusual, barely-possible edge cases. When would the universe first become favorable to the formation of life? That is, when do we first start seeing Earthlike rocky worlds with atmospheres and liquid water orbiting stable stars?

Well, once again it comes back around to the issue of metals, which in this context refers to all elements other than hydrogen and helium. Hydrogen and helium alone have no complex chemistry, and so cannot form life (lithium isn’t much better). We need, at minimum, carbon, nitrogen, and oxygen in significant quantities to expect life to appear. Alternate biochemistries may be possible with other elements, but these three will be the first metals (other than lithium and beryllium) formed by stars and they remain among the most abundant today, so we can assume that no life could have formed before they became abundant. 

 

However, life also needs pressure to force these elements to gather together and form complex compounds, and pressure requires a planet or similar large body (this is a debatable point—and I'll debate it later—but probably holds for intelligent, Earthlike life at least). It’s possible planets could form⁸ from just these light metals as waterworlds or carbon planets, but not clear if such planets could reasonably be habitable (more on that in a later post; but in short, don't get your hopes up). Our chances both to form planets and to develop life on them improve significantly if we also have heavier metals like silicon, iron, and magnesium.

Recognizing the importance of metals, astronomers often classify stars in terms of their metallicity, assuming that the abundance of all metals are roughly correlated—which isn’t always completely true, but is close enough for our purposes. By this metric stars are divided into three categories, Populations I, II, and III, that counterintuitively first appear in reverse order of their number.

 

The Pop III stars are those first formed by dark matter halos in the early universe, containing almost no metals. Regardless of when they first appear, by 500 million years after the Big Bang they're abundant enough to reionize the universe’s hydrogen with their light (by this point the gas is too thin to block all light like it did before last scattering) and form the first small galaxies. As mentioned they are both extremely massive and very short-lived (none have been observed today), which helps seed the cosmos with the metals expelled when they die. But because of their great size and the fact that many form black holes that consume their cores on dying, they produce relatively low amounts of iron and other heavy metals compared to carbon, nitrogen, and oxygen.

However, the production of those light metals eventually catalyzes the formation of smaller Pop II stars (including Methusalah), which then produce metals in proportions closer to those we observe today. Some of these early Pop II stars may have rocky planets, but at first they're both rare and deficient in heavy metals, making them poor candidates for life. Both the rate of star formation and metallicity of these stars gradually rises⁹ until 3 billion years after the big bang, at which point the universe has filled with large galaxies hosting a mix of Pop II and Pop I (metal-rich stars like our sun) systems.

At around this time star formation peaks and then begins to decline and metallicity plateaus, so the likelihood of an Earthlike planet forming is broadly the same as today. In fact, by some estimates the average age for rocky planets in the current universe should be 2 billion years older than Earth (though intriguingly, the average age of gas giants should be about the same¹⁰ as our solar system).

Chopra and Lineweaver 2018

But just because there are potentially habitable planets, that doesn’t mean it's a habitable universe. As stars are forming at high rates at this time, they're also dying at high rates, often with spectacular denouements that pour high-energy radiation into the surrounding cosmos. If we assume that any life must be made of volatile materials (as these will be the only ones available in solution that can be easily gathered and assembled into complex molecules) and that it will require the precise use of complex, delicate molecules (as would be necessary for the actions that define life, like reproduction and response to stimuli) then we can assume that all such life will be vulnerable to radiation to some extent.

Now, cosmic radiation events would not necessarily sterilize a planet—any life living beneath more than a few hundred meters of water or rock would be largely unaffected, so long as the atmosphere was not completely lost—but it could certainly be a setback, and a major impediment to the formation of intelligent life. If any spacefaring civilizations do emerge in this period, they could probably spread out and colonize other stars about as well as they could today—cosmic radiation events weren’t that frequent on the timescale of civilizations—but they’d be less likely to encounter other life compared to today, and in the long term more at-risk of catastrophes or even extinction caused by nearby radiation events.

But assuming these cosmic radiation events were potent enough to preclude complex life, and given that they are more common in the past—and still occur more commonly in certain regions of space—then this places some constraints on when and where we’d expect life and eventually complex life to occur.

Unfortunately, the data is not yet clear on where exactly these constraints are. Some models predict that the universe is inhospitable until shortly before the Earth appears¹¹, while others allow for a small number of hospitable galaxies¹² appearing at around the same time the first Earthlike worlds are forming. To understand why, let's look a bit closer at the causes and impacts of cosmic radiation.

Sources of Cosmic Radiation

NASA/JPL-Caltech

 

The biggest single sources of cosmic radiation are the supermassive black holes at the center of most galaxies, including the Milky Way. That black hole (Sagitarius A*) is, like most in the current universe, fairly quiescent at the moment. But when a large amount of mass approaches the black hole, it tends not to fall straight in but instead form a dense accretion disk of material outside the event horizon. This material is superheated by the compression and mixing caused by the intense gravity and so releases huge amounts of energy. The black hole becomes the core of an active galactic nucleus (AGN) that outputs intense radiation far greater than any star.

 

This happened more frequently when the universe is young and dense, and there were more collisions between galaxies that could have thrown matter towards a black hole. The most intense events form quasars, which produce as much as 1,000 times more light than the entire Milky Way does today. Their formation peaks around 4 billion years after the Big Bang.

The AGN at the heart of M87 (left) and the galaxy-scale jet of particles ejected from it by electromagnetic interactions in the accretion disk (right). ESO/EHT and NASA/ESA

Sagitarius A* may have passed through such an event 8 billion years ago, during which time the radiation produced would have been intense enough to remove the atmospheres of any Earthlike planets¹³ within 1 kiloparsec (kpc; 3,262 light-years) of the center of the galaxy and possibly affect habitability of planets as far as 10 kpc away (the sun is currently 8 kpc from the center). This means the worst effects should typically be limited to the central bulge of a large galaxy, but during mergers between galaxies¹⁴ there can be two AGN travelling through their now-shared galaxy.

Atmospheric loss of planets (in multiples of Earth’s atmosphere) by distance (in kiloparsecs) from Sag A* during its AGN phase. Different lines correspond to different estimates of the efficiency of atmospheric escape (ε) and placement within the galactic plane (τ = 1) or outside it (τ = 0). Balbi and Tombesi 2017

On the other hand, some researchers have suggested that AGN may actually help life appear¹⁵; beyond the range where radiation is high enough to strip away atmosphere, the UV light may help catalyze the formation of water and complex molecules¹⁶, and thus increase the chance of life appearing. If so, an AGN of just the right size could conceivably create a "pulse" of abiogenesis in its galaxy, causing many biospheres on different worlds to appear at roughly the same time (though this doesn't necessarily imply they'll develop at the same rate thereafter).

 

AGN still occur today (and brief "tidal disruption events" when stars pass close to supermassive black holes may have similar effects¹⁷), but much less frequently, and even when their formation peaks they probably aren’t the greatest danger to most worlds¹⁸; systems outside galactic cores likely have more to fear from less intense but much closer events caused by nearby stars. The same supernovae that distribute vital metals throughout the cosmos also flood neighboring star systems with high-energy radiation that can degrade the protective ozone layer in oxygen-rich atmospheres (increasing surface radiation from other sources until the ozone replenishes months or years later) and bombard the surface with charged particles that can directly damage any life living there.

Radiation produced by different sources throughout the history of the universe; SNIa and SNII are varieties of supernovae. Dayal et al. 2017

 

Only the largest, shortest-lived stars end in supernovae, and their reach is fairly short compared to AGN (significant impacts may only reach out under 10 parsecs in most cases¹⁹), so for the most part they should be a significant threat only in the densest regions of galaxies, with ongoing star formation. But in the worst cases, the radiation can be focused into tight beams, known as gamma ray bursts (GRBs), that can reach much farther (back into the kiloparsecs range) and blast any planet unlucky enough to be caught in their path with enough energy to severely damage the atmosphere in seconds. Supernovae produce more radiation overall, but the type and intensity of radiation produced by GRBs likely makes them more dangerous.

Unfortunately, it’s hard to say what effect GRBs have because they’re poorly understood. They’re divided into two categories, short GRBs and long GRBs, so-called because peak gamma radiation lasts 1 and 10 seconds, respectively—SGRBs are more intense, so both are of roughly equal danger to life. LGRBs are now believed to be caused by the collapse of massive, metal-poor stars²⁰, meaning that the threat from them decreases just as the probability of forming rocky planets increases. However, LGRBs are still observed originating from metal-rich galaxies, meaning that they can’t be counted out entirely.

 

SGRBs are more mysterious; the primary theory blames the collision of neutron stars or black holes, but poor observational data and anomalies in the data we do have plague this model. If these stellar remnants are the primary culprits, then SGRBs remain a danger long after star formation peaks and may be the primary threat to potentially habitable planets in the universe today.

The Present: Habitability of the Modern Universe

At this point the question is not just one of time, but of space as well. The region of the galaxy the sun occupies clearly allows for life today, but other parts of the galaxy—and whole other galaxies—may be less hospitable. Similarly, some regions of the universe may have become habitable far earlier than others, and some now-uninhabitable regions may improve later.

This has led some researchers to define a “galactic habitable zone”, near enough to the metal-rich galactic core to form rocky planets and avoid LGRBs, but far enough from the densest region to not be swamped with radiation from SGRBs and supernovae. For the Milky Way, estimates for the extent of this zone have generally placed it at 4 to 10 kpc from the galaxy’s center.

Theoretical habitable region of the galaxy—based on distance from the center—through time. The green line on the right shows the distribution of ages for currently habitable planets, based on this model. Lineweaver et al. 2004

However, recent modelling has suggested that shear weight of numbers is sufficient to extend the inner boundary into the core²¹, and may even make it the most likely place to find life²²; even though a higher proportion of star systems will encounter sterilizing cataclysms, the small fraction of habitable systems may still outnumber those outside the core.

Whether this extends into the central bulge itself (the innermost 1-2 kpc) is harder to say. Radiation from an AGN could leave the whole region effectively sterilized for some time, and even without one a planet there would be exposed to frequent supernovae; while these probably won't strip a planet's atmosphere away in most cases, they could still cause extinction events²³ as often as once every 10 million years (though perhaps we should expect natural selection for more radiation-tolerant life in such cases) In addition, the higher frequency of encounters between stars may scatter all planetary systems²⁴ before they can develop complex life. On the other hand, short distances between stars may also make "panspermia" events—transfer of life from one body to another—more likely (presuming they're possible at all), possibly making simple life more common and thus giving it more chances to develop to intelligence.

As to the outer boundary, data from the Kepler telescope has indicated that, above a minimum threshold—about that reached 3 billion years after the Big Bang—the appearance of Earthlike rocky planets is not strongly correlated with metallicity²⁵, though the relationship isn’t absent²⁶ and if planets around metal-poor stars prove to have fewer heavy metals, that may compromise their long-term habitability (due to the effect on internal heat production; more on that in another post). At any rate, stars may be able to migrate large distances²⁷ into and out of the inner galaxy, meaning that even if a region isn’t hospitable to the formation of habitable systems they may still pass through there anyway.

Observed exoplanets; black dots are lone exoplanets, green dots largest planets in multi-planet systems, red dots smaller planets in such systems. Buchhave et al. 2012

But why limit ourselves to the Milky Way? If our main criteria for habitability are high metallicity and low rate of potentially dangerous star formation, then the most likely places to find life²⁸ are massive elliptical galaxies, formed mostly of old, stable stars . Such galaxies lack the strong magnetic fields of spiral galaxies, however, and so may be more vulnerable to radiation from the AGN of other nearby galaxies. This would have been a bigger issue earlier in the lifetime of the universe, when galaxies were packed closer together, but there’s enough variation in large-scale densities that some elliptical galaxies may have been safe then as well, and ultimately the habitability of any given galaxy is more dependent on its particular history²⁹ than anything else. But let’s be honest with ourselves here, elliptical galaxies are just kinda boring.

A typical spiral (left) and elliptical (right) galaxy. Hubble

Fortunately, we do have a good analogue closer to home. Globular clusters are dense groups of stars found in many galaxies, and in spiral galaxies in particular they are often found outside the main disk of stars. Like elliptical galaxies they’re formed primarily of old stars, with few young hotheads to ruin everyone’s day. But they’ve traditionally been discounted as potential homes for alien life because of their low metallicity and dense packing of stars that, as with the galactic core, could destabilize planetary systems³⁰.

But the first issue, as we’ve seen, may not be a major impediment—these stars formed well after the minimum metallicity threshold for rocky worlds had been passed, and some may even have metallicity comparable to the sun³¹—and a small, long-lived star with a close habitable zone outside of the cluster’s center may be able to hold onto its planets³². In some cases, a good portion of a globular cluster's stars may be positioned just right³³ such that they are unlikely to lose their planets before life can appear, but still near enough to other stars that any intelligent life will have a good chance of reaching them before anything cataclysmic happens to their home systems.

Messier 80 globular cluster. NASA

Were intelligent life to appear in such a system and develop enough to consider the colonization of other stars, it would find the task much less daunting than we do. The average distance between stars in the clusters are as much as 20 times smaller than in our neighborhood (though perhaps somewhat more if we’re looking outside the core for long-term stability) meaning that travel and communication between them could be achieved on timescales similar to those during the age of sail on Earth (this may also help panspermia events, as with the galactic core). So if you want a story with regular interstellar travel and sprawling star empires but you don’t want to violate the lightspeed limit, a globular cluster may offer a good compromise.

Add to that, a nearby globular cluster could make for a rather impressive sight in the night sky.

Concept of the night sky of a planet in the 47 Tucanae globular cluster. William Harris and Jeremy Webb.

There are a lot of variables to consider here, and the many unknown factors and gaps in observational data make it impossible to come to any certain conclusions. But from all this research we can see an emerging model of the universe’s early evolution with two major consequences: First, the time of life’s emergence in the cosmos is probably constrained more by the threat of cosmic radiation than the availability of metals; and second, the Earth is unlikely to be among the first cohort of habitable worlds. The conditions were right for some kind of life to appear long before our sun formed, and even had it been limited to simple, radiation-tolerant forms through the first few billion years of its development, once the cosmic weather cleared up they would have a strong lead on us.

The Future: Decline and Fall of the Habitable Universe

That answers, as best we can, when life could have first appeared, and roughly where. Moving into the future, we can start looking for the latest point when new life could appear in the cosmos. This discussion must come with the major caveat that as time passes, the probability increases that intelligent civilizations—human or otherwise—will colonize enough of the universe to significantly alter or even eliminate the possibility of new intelligent life appearing. Of course, the lack of such interstellar civilizations in view now has some possible implications for the habitability of the universe before Earth, but that’s an issue we’ll tackle at a later date.

If we stick with the Big Freeze scenario, then we’ve got a long way to go. In contrast to the exciting pace of the universe’s past, the future is a long tale of gradual decline over unimaginably large timescales. This sounds rather grim, and indeed most people tend to focus on the dark, depressing ultimate state of this scenario. But there's still a long way to go before that happens.

Stelliferous Era (now - 10¹⁴ years)

For now, we’re still in the earliest phases of the Stelliferous (star-forming) Era, meaning that bright, star-filled galaxies with potentially habitable planets will continue to dominate the universe for many trillions of years, thousands of times longer than its current age.

But partway through this period “galaxies” will have to be amended to “galaxy”. The Milky Way is part of a local group of about 50 galaxies, all gravitationally bound and gradually collapsing into a single massive galaxy, in a process that takes 100 billion (10¹¹) to 1 trillion (10¹²) years. Other clusters can contain thousands of galaxies. But as the galaxies within each cluster combine, the clusters are being carried away from each other by cosmological expansion. Eventually they cross the cosmological horizon, meaning they are so distant that the space between them and us is expanding faster than the speed of light, and communication with them becomes impossible. If the rate of expansion continues to accelerate as it has been, all other galaxies will pass beyond the cosmological horizon on the same timescale as it takes each cluster to consolidate into one galaxy.

As they spread apart, light passing between galaxies will continue to redshift, passing out of the visible spectrum through infrared and microwaves to radio waves. Even after all other galaxies pass over the cosmological horizon, light they emitted previously will continue to reach us for trillions of years, but with increasing time dilation such that we can never see how they appear beyond the time when they pass the horizon. Eventually the redshift becomes so great that the wavelength of the light is greater than the distance to the cosmological horizon, making it impossible to observe. We will never see those distant galaxies again.

Thus each cluster will eventually form its own island universe, isolated from all of its former neighbors and—after sufficient redshift has been reached—unable to even see that they had ever existed at all. If any new intelligent beings appear in this time, they might believe³⁴—like we once did—that their galaxy constitutes all that the universe is and ever was, and that they live in a steady-state universe with no beginning or end. They might eventually deduce from the principles of thermodynamics that the universe must have a beginning, but without evidence of cosmological expansion would they hypothesize a Big Bang? Perhaps with especially good telescopes they could spot the occasional rogue star or planet ejected from the galaxy and document their redshift, then make some deductions from there. But inferring the existence of other galaxies would be a big step, and without other reference points or a CMB to work from, it would be easy to assume that their galaxy occupied a privileged position and expansion was an external phenomenon.

Within these now-isolated galaxies, star formation rates continue to decline but the accumulation of long-lived stars and increasing brightness of stars as they age means that the total light output of the galaxy remains constant for several trillion years. Rising metallicity favors the formation of smaller stars, and the galaxy is increasingly dominated by dim, long-lived dwarfs. Metallicity also affects the types of stars³⁵ that can form, eventually capping the maximum mass at 30 solar masses and reducing the minimum to as little as 0.04 solar masses—the latter types of stars being so cool on their surface that clouds of water ice could form.

ESO/M. Kornmesser

1 trillion years after the big bang, the first of the very low-mass red dwarfs, less than a quarter the sun’s mass, enter their extended death sequences³⁶ and form heretofore unseen blue dwarf stars, brighter and hotter than the preceding red dwarfs. If we assume that even the smallest stars can host habitable worlds at similar rates to sunlike stars throughout their lives, then the chances of new life emerging³⁷ will actually peak around 10 trillion years after the Big Bang, long after the galaxy has started dimming and the rest of the cosmos has receded out of view (however, I have some doubts about those assumptions—in particular that star systems don't decline in habitability over their lives—for reasons I'll discuss in later posts). 

But eventually the party has to end. Hydrogen is continuously consumed by fusion or trapped in white dwarfs, neutron stars, and black holes. Star formation declines until it finally grinds to a halt³⁸ at most 100 trillion (10¹⁴) years after the Big Bang. The smallest and longest-lived of the red dwarfs burn out 10 to 20 trillion years later.

Degenerate Era (10¹⁴ - 10³⁰ years)

But we’re not done yet. New stars may not be forming by conventional methods, but every now and again brown dwarfs—“failed stars” with too little mass to fuse hydrogen—may collide or accrete enough gas from the interstellar medium to edge over the line into proper red dwarfs, creating an island of light for the next few trillion years. Long-lived though these stars are, they form so rarely that in the entire galaxy there are no more than a few hundred at any one time.

Civilizations that formed around these stars would find themselves in a dark universe, though there are some other sources of light; white dwarfs may collide or accrete enough matter to trigger supernovae or form helium or carbon stars, all too fleeting to provide potential homes for new life, but possibly of interest to interstellar colonists—though there will never be more than a handful at once, and none at all at most times.

Travelling hundreds or thousands of light-years to neighboring stars would require a truly gargantuan effort, but it might be easier to make shorter hops to brown dwarfs, white dwarfs, or rogue planets and found colonies based on the dim light and nuclear fuel found there. Even for the most advanced civilizations, megastructures like Dyson spheres would rarely be worth the effort, but a typical brown or white dwarf could host a respectable civilization for a long time if they managed their resources well.

Jeff Bryant

As time goes on the supply of brown dwarfs to form new stars is dwindling, but perhaps more pertinently the supply of planets is dwindling too. Gravitational encounters eject some planets out of their systems, while others lose kinetic energy through gravitational radiation until they fall into their parent star (or what remains of it). By a quadrillion (10¹⁵) years after the Big Bang, no planetary systems remain.

Perhaps a few rogue planets could be recaptured by other brown dwarf or red dwarf stars later; such events would be exceedingly rare, but we have a lot of time to work with. Some rogue planets could even be (marginally) habitable without stars; the radioisotopes that heat Earth's interior will probably be too dilute by this point to help, but a collision with another body could leave the planet warm for some time, or it could receive tidal heating from a captured moon or parent body. For planets in the densest region of the galaxy, capture of dark matter particles may provide enough heat³⁹ to maintain liquid water at the surface (though this depends on exact nature of dark matter, which is not yet known).

Head much further into deep time, though, and the galaxies themselves fall apart. Gravitational interactions between objects within them gradually eject over 90% of their mass into deep space—where they will be isolated in their own island universes—over the next 10²⁰ years. What remains spirals in towards the center due to gravitational radiation. At most it takes 10³⁰ years for there to be nothing left but a single supermassive black hole.

Black Hole Era (10³⁰ - 10¹⁰⁰ years) and Beyond

We can safely assume that no life in any form we might recognize will arise naturally past this point. Anything alive now would have to be a preexisting civilization that had prepared for this outcome and lived on the meager gruel of Hawking radiation provided by these black holes. This sounds like a marginal lifestyle, but if these civilizations were advanced enough to get the most out of what energy they could get, and if they all lived in virtual realities with simulation timescales that could be adjusted to match available energy, then they could have quite lively societies. Depending on exactly how energy-efficient their simulations are, the majority of subjective time experienced by sentient beings over the full lifetime of the universe could be in these black-hole-fueled civilizations.

How long they could survive depends on whether or not protons decay, a phenomenon not proven by experiment but predicted in many theoretical models (neutrons separated from nuclei quickly decay into protons and electrons, so they’d be lost as well). The estimates for the half-lives of protons range widely between 10³² to 10²⁰⁰ years. At the lower end of that range, these civilizations would have to use the energy from Hawking radiation to synthesize new matter to replace their losses, and once the black holes finally die out after 10¹⁰⁰ years their limited reserves of energy and matter will soon run out (on these timescales anyway).

At the higher end, a civilization might survive longer harvesting the heat of an iron star, a mass of matter that continues to produce heat from the decay of heavy elements and quantum-tunneling cold fusion of light elements to iron-56. If protons never decay, such an object could keep producing a miniscule supply of energy for 10¹⁵⁰⁰ years.

So Anyway

For the purposes of our example world, we won’t do anything too fancy just yet. We’ll assume that they are roughly our contemporaries, and live within the disk of a spiral galaxy at a comfortable distance from the center. Tempted though I am to place them in a globular cluster, I think it would just add complications in the future, and keeping interstellar distances similar to those in Earth’s neighborhood will make it easier to draw analogies between our example civilization and scifi tales about humanity’s future when we reach the point of discussing interstellar travel. Anyway, Earthlike worlds have yet to be found in globular clusters, so let’s not put all our chips down just yet.

Next time we’ll actually get to make some creative choices, and get to know where we’ll be spending the next few billion years.

In Summary

•  Life requires metals, and so could not have appeared before the earliest stars died.
• The absolute earliest stars likely appeared no early than 30 million years after the Big Bang, and did not appear in number until 180 million years after the Big Bang.
• Planets may not have formed until at least 100 million years after the Big Bang.
• Life of some form becomes at least marginally plausible around the same time.
• Metals became abundant enough to favor the appearance of habitable planets 3 billion years after the Big Bang.
• Cosmic radiation may have inhibited the appearance of life before Earth’s formation, and may still do so in the innermost and outermost regions of the galaxy.
• The most habitable regions today are likely elliptical galaxies and middle disk areas of spiral galaxies, but life could plausibly appear in a much larger region including globular clusters and near the galactic core.
• The universe will consist of lone, isolated galaxies in 1 trillion years.
• Most star formation stops after 100 trillion years, but new life may be possible for as long as 10³⁰ years.
• Preexisting life could survive 10¹⁰⁰ to 10¹⁵⁰⁰ years, depending on the nature of proton decay.

Notes

Properly speaking the Big Rip scenario is not yet ruled out by known cosmology, but it’s the less likely of the 2 known options, and even if it does occur it may take so long that the whole scenario I laid out happens first anyway.

Though I tried to link mostly to primary sources, much of the background material for the early habitability of the universe came from the collection Habitability of the Universe Before Earth. It’s a pretty good read—a bit pricey, unfortunately, but if you have ScienceDirect access you can get the full text online⁴⁰.

If  the idea of civilizations clinging to life around black holes long after the disappearance of the last stars interests you, there’s a couple decent videos on the subject by the ever-informative Isaac Arthur that go into more detail.

Also I just wanted to note that this paper⁴¹ includes perhaps the sassiest citation I’ve ever seen, when they cite a mathematics paper from 1904 for the claim that “It is well known that extrapolation can lead to misleading conclusions”.

And finally a special note of thanks to Sean Raymond, author of the Planetplanet blog, for giving me some advice and encouragement to get started.

Buy me a cup of tea (on Patreon)

Part II

1 Loeb, A. (2014). The habitable epoch of the early Universe. International Journal of Astrobiology, 13(4), 337-339.
2 Rauscher, T., & Patkós, A. (2010). Origin of the chemical elements. arXiv preprint arXiv:1011.5627.
3 "Hubble finds 'birth certificate' of oldest known star", ScienceDaily
4 Bowman, J. D., Rogers, A. E., Monsalve, R. A., Mozdzen, T. J., & Mahesh, N. (2018). An absorption profile centred at 78 megahertz in the sky-averaged spectrum. Nature, 555(7694), 67-70.
5 Naoz, S., Noter, S., & Barkana, R. (2006). The first stars in the universe. Monthly Notices of the Royal Astronomical Society: Letters, 373(1), L98-L102.
6 Johnson, J. L., & Li, H. (2013). Constraints on planet formation via gravitational instability across cosmic time. Monthly Notices of the Royal Astronomical Society, 431(1), 972-977.
7 Lingam, M., & Loeb, A. (2021). The Extended Habitable Epoch of the Universe for Liquids Other than Water. arXiv preprint arXiv:2101.10341.
8 Mashian, N., & Loeb, A. (2016). CEMP stars: possible hosts to carbon planets in the early universe. Monthly Notices of the Royal Astronomical Society, 460(3), 2482-2491.
9 Chopra, A., & Lineweaver, C. H. (2018). The cosmic evolution of biochemistry. Habitability of the Universe Before Earth, 1, 75-87.
10 Behroozi, P., & Peeples, M. S. (2015). On the history and future of cosmic planet formation. Monthly Notices of the Royal Astronomical Society, 454(2), 1811-1817.
11 Piran, T., & Jimenez, R. (2014). Possible role of gamma ray bursts on life extinction in the universe. Physical review letters, 113(23), 231102.
12 Li, Y., & Zhang, B. (2015). Can life survive gamma-ray bursts in the high-redshift universe?. The Astrophysical Journal, 810(1), 41.
13 Balbi, A., & Tombesi, F. (2017). The habitability of the Milky Way during the active phase of its central supermassive black hole. Scientific Reports, 7(1), 1-6.
14 Comerford, J. M., Pooley, D., Barrows, R. S., Greene, J. E., Zakamska, N. L., Madejski, G. M., & Cooper, M. C. (2015). Merger-driven fueling of active galactic nuclei: Six dual and of AGNs discovered with Chandra and Hubble Space Telescope observations. The Astrophysical Journal, 806(2), 219.
15 Lingam, M., Ginsburg, I., & Bialy, S. (2019). Active galactic nuclei: boon or bane for biota?. The Astrophysical Journal, 877(1), 62.
16 Liu, C., Chen, X., & Du, F. (2020). Impact of an Active Sgr A* on the Synthesis of Water and Organic Molecules throughout the Milky Way. The Astrophysical Journal, 899(2), 92.
17 Pacetti, E., Balbi, A., Lingam, M., Tombesi, F., & Perlman, E. (2020). The impact of tidal disruption events on galactic habitability. Monthly Notices of the Royal Astronomical Society, 498(3), 3153-3157.
18 Dayal, P., Ward, M., & Cockell, C. (2016). The habitability of the Universe through 13 billion years of cosmic time. arXiv preprint arXiv:1606.09224.
19 Gehrels, N., Laird, C. M., Jackman, C. H., Cannizzo, J. K., Mattson, B. J., & Chen, W. (2003). Ozone depletion from nearby supernovae. The Astrophysical Journal, 585(2), 1169.
20 Levesque, Emily M., Lisa J. Kewley, Edo Berger, and H. Jabran Zahid. "THE HOST GALAXIES OF GAMMA-RAY BURSTS. II. A MASS–METALLICITY RELATION FOR LONG-DURATION GAMMA-RAY BURST HOST GALAXIES." The Astronomical Journal 140, no. 5 (2010): 1557.
21 Gowanlock, M. G., Patton, D. R., & McConnell, S. M. (2011). A model of habitability within the milky way galaxy. Astrobiology, 11(9), 855-873.
22 Gowanlock, M. G. (2016). Astrobiological effects of gamma-ray bursts in the Milky Way galaxy. The Astrophysical Journal, 832(1), 38.
23 Balbi, A., Hami, M., & Kovačević, A. (2020). The Habitability of the Galactic Bulge. Life, 10(8), 132.
24 Jiménez-Torres, J. J., Pichardo, B., Lake, G., & Segura, A. (2013). Habitability in different Milky Way stellar environments: a stellar interaction dynamical approach. Astrobiology, 13(5), 491-509.
25 Buchhave, L. A., Latham, D. W., Johansen, A., Bizzarro, M., Torres, G., Rowe, J. F., ... & Quinn, S. N. (2012). An abundance of small exoplanets around stars with a wide range of metallicities. Nature, 486(7403), 375-377.
26 Wang, J., & Fischer, D. A. (2014). Revealing A Universal Planet–Metallicity Correlation For Planets Of Different Sizes Around Solar-Type Stars. The Astronomical Journal, 149(1), 14.
27 Roškar, R., & Debattista, V. P. (2015). Radial Migration in Spiral Galaxies. In Lessons from the Local Group (pp. 379-390). Springer, Cham.
28 Dayal, P., Cockell, C., Rice, K., & Mazumdar, A. (2015). The quest for cradles of life: using the fundamental metallicity relation to hunt for the most habitable type of galaxy. The Astrophysical Journal Letters, 810(1), L2.
29 Stanway, E. R., Hoskin, M. J., Lane, M. A., Brown, G. C., Childs, H. J. T., Greis, S. M. L., & Levan, A. J. (2018). Exploring the cosmic evolution of habitability with galaxy merger trees. Monthly Notices of the Royal Astronomical Society, 475(2), 1829-1842.
30 Kane, S. R., & Deveny, S. J. (2018). Habitability in the Omega Centauri Cluster. The Astrophysical Journal, 864(2), 115.
31 Smith, G. H. (2021). Which Milky Way Globular Clusters Might Have Favored an Extraterrestrial Civilization?. Research Notes of the AAS, 5(7), 169.
32 "Globular Clusters Could Nurture Interstellar Civilizations", Harvard-Smithsonian Center for Astrophysics
33 Di Stefano, R., & Ray, A. (2016). Globular clusters as cradles of life and advanced civilizations. The Astrophysical Journal, 827(1), 54.
34 Krauss, L. M., & Scherrer, R. J. (2007). The return of a static universe and the end of cosmology. General Relativity and Gravitation, 39(10), 1545-1550.
35 Adams, F. C., & Laughlin, G. (1997). A dying universe: the long-term fate and evolutionof astrophysical objects. Reviews of Modern Physics, 69(2), 337.
36 Adams, F. C., Bodenheimer, P., & Laughlin, G. (2005). M dwarfs: planet formation and long term evolution. Astronomische Nachrichten: Astronomical Notes, 326(10), 913-919.
37 Loeb, A., Batista, R. A., & Sloan, D. (2016). Relative likelihood for life as a function of cosmic time. Journal of Cosmology and Astroparticle Physics, 2016(08), 040.
38 Adams, F. C., & Laughlin, G. (1997). A dying universe: the long-term fate and evolutionof astrophysical objects. Reviews of Modern Physics, 69(2), 337.
39 Hooper, D., & Steffen, J. H. (2012). Dark matter and the habitability of planets. Journal of Cosmology and Astroparticle Physics, 2012(07), 046.
40 Gordon, R., & Sharov, A. (2017). Habitability of the Universe before Earth: Astrobiology: Exploring Life on Earth and Beyond (series). Academic Press.
41 Marzban, C., Viswanathan, R., & Yurtsever, U. (2014). Earth before life. Biology Direct, 9(1), 1-7.

Previous Versions:

• Before August 2021