American Scientist

For a cosmic phenomenon that is arguably the most important thing in the universe, dark energy is surprisingly difficult to study. As far as we can tell, it exists everywhere in the universe, completely uniformly, woven into the fabric of space itself, and its only effect is to stretch space out so gradually that it has no detectable impact on any scale smaller than the vast expanses between distant galaxies.

Dark matter physicists have it much easier—despite being just as invisible as dark energy, dark matter makes its presence very known by clumping around virtually every galaxy or cluster of galaxies we’ve ever seen, dominating the gravitational field, bending light, and altering the course of cosmic history from the very beginning. Dark energy, on the other hand, just . . . expands.

This doesn’t completely prevent us from studying it. There are essentially two handles we have on dark energy: the expansion history of the universe and the way that galaxies and clusters of galaxies have grown over time. For both of these, we’re peering into the distance and the past, tracing out the evolution of the cosmos over time. But no matter how we look, we’re trying to tease out small effects using faint signals and statistics.

As challenging as these kinds of studies are, it’s worth putting in the effort, since dark energy is both the dominant component of the cosmos and a sure sign of some new physics beyond our current understanding.

That, and the fact that depending on what dark energy turns out to be, it might violently and inescapably destroy the universe, much sooner than anyone ever imagined, in a dark energy apocalypse appropriately named Big Rip. Not only would it be a kind of destruction from which there is no escape, it would be one that could tear apart the very fabric of reality, rendering any thinking creatures in the cosmos helpless as they watch their universe being ripped open around them.

This alarming possibility is hardly an outlandish fringe idea. Indeed, the best cosmological data we have not only fails to rule it out, but, from some perspectives, slightly prefers it. So it’s worth spending a bit of time exploring what, exactly, it would do to us.

A Cosmological Nonconstant

Dark energy is often assumed to be a cosmological constant that stretches space out, accelerating cosmic expansion by imbuing the universe with some inherent inclination for swelling. On large scales, this is a pretty good description. But within galaxies, solar systems, or in the close vicinity of organized matter generally, a cosmological constant has no effect. It can be more properly thought of as a force for isolation—if two galaxies are already distant from one another, they get more distant, and individual galaxies, clusters, or groups of galaxies find themselves more and more alone as time goes on. They also form a bit more slowly in the presence of a cosmological constant than they otherwise would. What the cosmological constant cannot do is break apart anything that is already, in any sense, a coherent structure: What therefore gravity hath joined together, let not a cosmological constant put asunder.

The reason for this small mercy of the cosmological constant (which, to be fair, does still destroy the whole universe eventually) lies in the “constant” part of the story. If dark energy is a cosmological constant, its defining feature is that the density of dark energy in any given part of space is constant over time, even as space expands. The expansion rate isn’t constant, just the density of the stuff itself, in any given volume of space. This makes sense in a way, if every bit of space is automatically assigned a set amount of dark energy within it, but it’s still super weird, because it means that as space gets bigger, the amount of dark energy increases to keep the density constant. It also means that if you draw a sphere of a given size anywhere in the universe and measure the amount of dark energy inside the sphere, and then do the same at some future time, you’ll always get the same number, regardless of how much the outside universe has expanded in the meantime. If your original sphere contains a cluster of galaxies and some quantity of dark energy, in a billion years the amount of dark energy in that region will still be the same, so if it wasn’t enough to mess up the galaxy cluster before, it won’t be in the future. The balance between matter and dark energy in that sphere does not significantly change even as the rest of the cosmos seems to inexorably empty out.

The Big Rip could tear apart the very fabric of reality, rendering any thinking creatures in the cosmos helpless as they watch their universe being ripped open around them.

• Facebook
• Twitter

This is reassuring. If you happen to be a clump of matter in the universe, and you would like to form a nice, stable, gravitationally bound galaxy, you can rest assured that once you get enough matter together to build something, dark energy won’t ruin all that hard work.

Unless, that is, the dark energy is something more powerful than a cosmological constant.

A cosmological constant is just one possibility for dark energy. All we really know about dark energy is that it’s something that makes the universe expand faster. Or, more precisely, it has negative pressure. Negative pressure is a weird concept, because normally one thinks of pressure as something that pushes outward. But in Einstein’s general-relativistic way of thinking about the universe, pressure is just another kind of energy, like mass, or radiation, and is thus gravitationally attractive. And in general relativity, gravitational attraction is just a consequence of the bending of space.

Imagine rolling a tennis ball across the surface of a trampoline. Now imagine putting a bowling ball in the center. The way the tennis ball falls toward the bowling ball, or curves as it goes past it, is a pretty good analogy of how objects move through space in the presence of large masses. If you take general relativity into account, the dent is deeper not only if the ball is more massive, but also if it is hot, or if it has high internal pressure. So pressure, like other forms of energy, acts a lot like mass. From a gravitational perspective, pressure pulls. When you calculate the gravitational effect of a clump of gas, for instance, you have to factor in not just its mass, but its pressure, and both contribute to the gravitational impact that gas has on the stuff around it. In fact, the pressure contributes more to the spacetime curvature than the mass does.

What does that mean for something with negative pressure? If the pressure of some weird substance can be negative, it means that it can effectively cancel out the mass of the stuff, at least as regards its impact on the curving of space-time. If you write down the pressure and density of dark energy in the form of a cosmological constant, in the appropriate units, the pressure is exactly the negative of the density.

We usually talk about the relationship between a substance’s density and its pressure using a number called the equation of state parameter, written as w—it’s equal to pressure divided by energy density, in units in which that comparison makes sense. Here, we’re interested in the equation of state of dark energy, which, given enough time, will be the equation of state of the whole universe, since dark energy becomes more and more important in the expanding universe as everything else dilutes away. If the measured value of w=−1 exactly, that tells you that the pressure and the density are exactly opposite, and dark energy is a cosmological constant. Because the energy density in a cosmological constant is always positive, at first glance it seems as though it should act just like matter and amp up the gravity that slows down the expansion of the universe. But because the negative pressure is given a heavier weight in the equations, all a cosmological constant ends up doing is contributing toward accelerating cosmic expansion.

At least it does so in a predictable way. A cosmological constant, with w=−1, has a total energy density that is exactly constant over time as the universe expands, without increasing or decreasing. For dark energy with any other value of w, this is no longer the case. So it’s important to figure out what we’re really dealing with here.

In the years after dark energy was first discovered, it was clear that something was making the expansion of the universe accelerate, which meant there had to be something out there with negative pressure. It turns out that anything that has a value of w less than −1/3 gives you both negative pressure and accelerated expansion. But knowing the value of w could tell us whether dark energy is a true cosmological constant (w=−1 always), or some kind of dynamical dark energy whose influence on the universe might change over time. So astronomers went looking for a way to determine the value of w exactly. If dark energy turned out not to be a cosmological constant, this would indicate that we had not only discovered a new kind of physics acting on the universe, but one with the added bonus of being something even Einstein hadn’t foreseen (he has to be wrong about something).

For a few years, this was the name of the game: measure w, find out what’s going on with dark energy. Measurements were made, papers were written, and plots were drawn showing which values of w agreed with the data. The cosmological constant case looked like it just might win out.

But in the late 1990s and early 2000s, a small group of cosmologists pointed out a major undiscussed assumption their colleagues were putting into their calculations. It was a perfectly reasonable assumption to make, because neglecting it would violate certain long-held principles of theoretical physics so fundamental that no one wanted to upset them. But these principles weren’t required by the data, and in the end, as scientists, our first loyalty has to be to the data. Even if it means rewriting the fate of the universe.

Off the Edge of the Map

The simple question physicist Robert Caldwell of Dartmouth College and his colleagues asked was, What if w is less than −1? Say, −1.5? Or −2? Up until this point, such a possibility was generally thought too outlandish to be considered. Plots in papers showing the “allowed” region for w based on the data tended to abruptly cut off at −1. The axis might go from −1 to 0, or −1 to 0.5, but −1 was a hard wall, the same way you might put a hard wall at 0 when guessing a person’s height.

But when Caldwell looked at the problem, all the observations of w pointed to a value of −1 or something very close to it. Which suggested that there might be values below −1 that were also allowed by the data, if only someone were to check. This hypothetical dark energy with w less than −1 was dubbed by Caldwell “phantom dark energy” and would be deeply inconsistent with the Important Theoretical Principles—specifically, the “dominant energy condition,” which says, roughly, that energy can’t flow faster than light. This seems like a completely sensible condition to place on the universe, but it’s subtly different than the usual statement that light (or any kind of matter) has an ultimate speed limit, and it’s currently less of a proven physical principle than a Very Good Idea. Maybe it’s flexible?

Caldwell and his colleagues went ahead and calculated constraints based on a full range of possibilities for w. Not only did they find that values below −1 were perfectly consistent with the data, they also found, through a simple, straightforward calculation, that if w is even infinitesimally lower than −1, dark energy will tear the entire universe apart, and it will do so in a finite, calculable time.

The Big Rip

You can think of it as an unraveling.

The first things to go are the largest, most tenuously bound. Giant clusters of galaxies, in which groups of hundreds or thousands of galaxies flow lazily around each other in long intertwined paths, begin to find that those paths are growing longer. The wide spaces traversed by the galaxies over millions or billions of years widen even more, causing the galaxies at the fringes to slowly drift away into the growing cosmic voids. Soon, even the densest galaxy clusters find themselves inexorably dissipated, their component galaxies no longer feeling any central pull.

From a vantage point within our own galaxy, the loss of the clusters should be the first ominous sign that the Big Rip is in progress. But the speed of light delays this clue until we are already feeling the effects much closer to home. As our local cluster, Virgo, begins to dissipate, its previously languid motion away from the Milky Way begins to pick up speed. This effect is subtle, though. The next one is not.

We already have astronomical all-sky surveys that are capable of measuring the positions and motions of billions of stars within our own galaxy. As the Big Rip approaches, we start to notice that the stars on the edges of the galaxy are not coming around in their expected orbits, but instead drifting away like guests at a party at the end of the evening. Soon after, our night sky begins to darken, as the great Milky Way swath across the sky fades. The galaxy is evaporating.

From this point, the destruction picks up its pace. We begin to find that the orbits of the planets are not what they should be, but are instead slowly spiraling outward. Just months before the end, after we’ve lost the outer planets to the great and growing blackness, the Earth drifts away from the Sun, and the Moon from the Earth. We too enter the darkness, alone.

As the Big Rip approaches, the stars on the edges of the galaxy drift away like guests at a party at the end of thee vening. Soon after, our night sky begins to darken. The galaxy is evaporating.

• Facebook
• Twitter

The calm of this new solitude doesn’t last.

At this point, any structure still intact is straining under the push of the expanding space within it. The Earth’s atmosphere thins, from the top. Tectonic motions within the Earth respond chaotically to the shifting gravitational forces. With only hours to go, the Earth cannot hold together: Our planet explodes.

Even the destruction of Earth could, in principle, be survivable, if, having interpreted the signs, you have already retreated to some compact space-based capsule. (When the danger is space itself, you want to be in a structure that has as little space in it as possible.) But that reprieve is short-lived. Before long, the electromagnetic forces that hold together your atoms and molecules cannot withstand the ever-expanding space within all matter. In the last tiny fraction of a second, molecules crack open, and any thinking beings still holding on are destroyed, torn atom-from-atom from within.

Beyond that point, there is no possibility of watching the destruction, but it carries on nonetheless. Nuclei themselves, the ultradense matter in the centers of atoms, are the next to go. The impossibly dense cores of black holes are eviscerated. And at the final instant, the fabric of space itself is ripped apart.

Unfortunately, we may never be able to say with certainty that we are safe from a Big Rip. If w is at all lower than −1, even one part in a billion billions, dark energy is phantom dark energy, capable of tearing the universe apart. Because it’s impossible to ever measure anything with complete, uncertainty-free precision, the best we may ever be able to do is say that if the Big Rip does occur, it will be so far in the future that all structure in the cosmos will have decayed already by the time it happens. Because even with phantom dark energy, the closer w gets to −1, the farther into the future the Big Rip is pushed. The last time I calculated the earliest possible Big Rip, based on the 2018 data release from the Planck satellite, I got something in the vicinity of 200 billion years.

Phew.

But given the potential consequences, both for the universe and for the structure of physics itself, we in the astronomical community put a pretty high priority on figuring out where we currently sit on the scale from w=−1 to Violent Cosmic Doom. We can’t measure w directly, but we can determine it indirectly by measuring the past expansion rate of the universe and comparing it to our best theoretical modeling of what different kinds of dark energy would have done. In principle, there are several ways to get at w, and some of them can be done in subtle ways that don’t require calculating the expansion rate at specific distances. But the most straightforward way to get a handle on dark energy is to figure out our full expansion history. And it turns out all the weirdnesses of cosmology come crashing together if you try to do something as simple as answer the question, “How far away is that galaxy?”

Ladder to Heaven

In order to meaningfully compare the local space-expansion rates at two distant points in the universe, you first have to know exactly how distant each one is. This is no big deal for something on Earth, or even something as close as the Moon, because you can measure the distance by bouncing a laser beam off of it and seeing how long the light takes to come back. On those kinds of scales, the universe is pretty reasonable. It acts basically like an unchanging space where the distance from A to B is straightforwardly measurable and makes sense and everything works. When it comes to things outside the Solar System, it gets trickier, both because things that are more distant are harder to measure, and because on larger and larger scales, the expansion starts to change the definition of distance itself.

Astronomers have, over the years, patched together with duct tape and twine a set of overlapping definitions and measurements of distance that build upon one another. As kludgy as it still sometimes seems, it’s the result of decades of innovations in observational astronomy and data analysis, and has given us an intuitive but frustratingly difficult-to-implement strategy known as the distance ladder.

Let’s say you need to measure the length of a large room, and all you have is an ordinary-sized ruler. You could lay the ruler down repeatedly until you cover the length of the room, if you don’t mind crawling around on the floor. Or you could be a bit more creative and measure the length of your stride, then just walk across the room, counting steps. If you chose the steps method, you’re creating a distance ladder: a system of measuring a large distance by calibrating your measurements with something more manageable.

In astronomy, the distance ladder has a series of rungs that allow it to extend out to objects that are billions of light-years away. Within the Solar System, direct laser measurements, orbital scalings, and even eclipses help us gather distance data. Beyond that, the next step is to use parallax. This is a method that takes into account the fact that when you change your vantage point, nearby things seem to shift their positions relative to a fixed background more than distant ones do. It’s the same effect that makes a finger held in front of your face seem to jump back and forth when you close one eye and then the other. If we look at a nearby star in June, and then the same star in December, the fact that the Earth is in a different location in its orbit around the Sun means that the star will appear to have moved slightly with respect to more distant background objects. The closer it is, the bigger the shift. Unfortunately, for anything outside our own galaxy, these apparent motions are too small to be perceived, and we need another method—a way to determine the distance of bright objects just from the properties of their light.

The key to everything from here on out is the concept of a standard candle: a kind of object (such as a star) that has some physical attribute that tells you its brightness. Then, by seeing how bright it looks, you can tell how far away it is. Kind of like having a light bulb with “60 watts” written on it. You know how bright it should be, but you’ll get less light from it when it’s far away.

Of course, nothing in space has its brightness stamped on it, but we have something almost as good. The breakthrough that first allowed us to use standard candles in astronomy was due to the astronomer Henrietta Swan Leavitt in the early 1900s. Working at Harvard Observatory, she discovered that a kind of star known as a “Cepheid variable” brightened and dimmed in a predictable way. A Cepheid that’s intrinsically more luminous does slow, gradual pulsations, getting a little bit brighter and a little bit dimmer over a long period. A Cepheid that’s intrinsically less luminous pulsates more quickly, with wide swings between its brightest and dimmest states.

This discovery was revolutionary, and perhaps one of the most important in the history of astronomy, in that it let us finally measure the scale of the universe around us. It meant that anywhere a Cepheid could be seen, we could get a reliable distance and start to make a usable map. By measuring how quickly a Cepheid pulsed, and how bright it looked from here, Leavitt could tell you with great precision how bright it really was, and thus how distant.

How far does this get us? We can see Cepheid variable stars throughout the Milky Way and in nearby galaxies, so we can use parallax for the nearby ones, carefully calibrate the pulsation relationship, and then use the more distant ones to tell us the distances to other galaxies.

The next step in the distance ladder is a crucial one, but it’s also where things get really messy, in every sense of the term. Certain types of supernovae make explosions whose properties are so predictable that we can use them as mile markers for the universe. This kind of explosion, a Type Ia supernova, is what happens when a white dwarf star somehow picks up some mass from another, equally unfortunate star and spectacularly rips itself apart. (See “Illuminating Dark Energy with Supernovae.” ) Because all white dwarf stars are fairly simple objects (simple for stars, anyway), and because the explosion is governed by physics that we feel we have a somewhat decent handle on, Type Ia supernovae were considered for a time to be good standard candles—the explosions all looked pretty similar. But it was later found that they’re better described as standardizable, in the same way Cepheid variables are. If you can measure how the explosion peaks and dims, you can get a good sense for the total amount of energy put out by the explosion, and thus an idea of how bright it really is.

Exactly how well we can use these stars as distance benchmarks—with some tweaks to account for slight differences in stellar circumstances—is still a matter of incredibly intense debate in the astrophysics community. Which is understandable, as the stakes couldn’t be higher. Type Ia supernovae are the gold standard for distance measurements across vast expanses of the cosmos. They’re what allowed astronomers in the late 1990s to detect the accelerated expansion of the universe, and they’re what astronomers now use as their best handle on the nature of dark energy.

The precision with which we can now calibrate galaxy distances with supernovae is impressive, with accuracies pushing toward the 1 percent level. This makes it possible to measure the expansion rate of the universe, by determining how distant the galaxies are and how fast they’re moving away. We talk about the expansion rate in terms of the Hubble Constant—the number that relates distance and recession speed. As of this writing, supernova measurements allow us to measure the Hubble Constant to an accuracy of 2.4 percent.

Which is weird, because the number we get totally disagrees with the value of the exact same number we derive from looking at the cosmic microwave background.

Expanding Confusion

For the last several years, measurements of the Hubble Constant from supernovae have been giving us a number around 74 kilometers per second per megaparsec—that means that a galaxy one megaparsec away (3.26 million light-years) is receding from us at around 74 kilometers per second. One twice as far away is moving, relative to us, about twice as fast. But we can also measure the Hubble Constant indirectly, by carefully studying the geometry of the hot and cold spots in the cosmic microwave background. When we measure it that way, the number we get is closer to 67 kilometers per second per megaparsec. Even though these observations are looking at very different epochs of cosmic history, each of them can tell us the expansion rate today. In a universe made of what we think it’s made of, both methods of determining the Hubble Constant really ought to give us the same number. And they don’t.

This hasn’t always been considered to be that big a problem, because no one thought either measurement was so incredibly precise as to settle the question. Until recently, the state of play was that the cosmic microwave background folk assumed that there was some distance ladder misestimate that would be sorted out eventually, dropping the number down a tad, and the supernova folk figured that the cosmic microwave background measurements, which derive ultimately from attempting to measure the shape of space itself, were so complicated that surely something would show that the number was really just a little bit higher. This isn’t an unreasonable assumption, given the number of calculations and conversions that go into looking at a baby picture of the universe and converting that into a present-day expansion rate. And the distance ladder, likewise, really is fantastically complicated. Without even getting into all the possible biases that might creep in if you don’t account for every relevant property of the supernovae themselves, calibrating variable stars is not easy, and even distances to relatively nearby galaxies sometimes come with huge uncertainties. Part of this is due to how the populations of Cepheid variables we can see nearby are different from those far away, and . . . well, I could go on. Let me just say there are debates.

While the assumptions from each side that the other has done something wrong haven’t quite gone away, the situation is getting increasingly uncomfortable due to the fact that both sides are improving their methods, knocking out all known sources of measurement bias, and still finding numbers that ever more precisely do not agree with each other.

Type Ia supernovae are the gold standard for distance measurements across vast expanses of the cosmos. They’re what astronomers now use as their best handle on the nature of dark energy.

• Facebook
• Twitter

It’s unclear what the solution to this problem will end up being. Maybe it really does come down to systematic errors in the data, or some problem with the measurements themselves. Maybe it’s just a statistical fluke, as unlikely as that looks on the surface. Some of the most intriguing explanations involve dark energy that is not your garden-variety cosmological constant, but is instead something rather more ominous—something that could perhaps lead to a Big Rip. There’s one hypothesis that would go a reasonable way toward fixing the discrepancy between the measurements: dark energy getting more powerful over time, in just the way you might expect from the early stages of a phantom-dark-energy-dominated cosmos.

We probably shouldn’t panic just yet. As discussed, the data still aren’t that clear. Most measurements of w give a value that is fully consistent with −1, and though it’s true that values less than −1 are sometimes very slightly preferred, that preference isn’t really statistically meaningful. As for the Hubble Constant disagreement, even if all the measurements are correct, nonapocalyptic explanations for the discrepancy—involving weird models of dark matter, or altered conditions in the early universe—are very much in the running. In fact, even tweaking dark energy wouldn’t be enough to totally solve the problem, so it’s not unreasonable to assume that the solution might lie elsewhere. And even if there has been a sharp upturn in the effects of dark energy in recent cosmic history, suggesting something like phantom dark energy, we still have a lot of time before a Big Rip could possibly occur.

This article is excerpted and adapted from The End of Everything: (Astrophysically Speaking), by Katie Mack. Copyright © 2020 by Dr. Katie Mack. Reprinted by permission of Scribner, a division of Simon & Schuster, Inc.