Imagine looking out over a beautiful vista. The sun glances off the snowy peaks of distant mountains, a river winds through rolling hills. There is something wonderful about beholding the contours of a majestic landscape.
It might not be obvious when you look at the night sky, but the universe has a landscape of its own – filaments of galaxies separated by near-empty voids. We have long known this much. But now one group of cosmologists is taking things further and proposing that the universe possesses not just a landscape, but a timescape, too. The idea is that the very flow of time varies from place to place.
Advertisement
To say this goes against the grain would be an understatement: we have always thought that on large scales, time runs at the same speed throughout the universe. But in this picture – known as timescape cosmology – there are large patches of the universe where time has been ticking for billions of years longer than we usually assume.
It may sound strange, but what entices some physicists is the simple elegance of this idea. There is no freaky physics involved, it springs naturally from established theory. “It is part of the structure of general relativity,” says its inventor at the University of Canterbury in New Zealand. “It’s just not a part of the structure that people thought about before.” His proposal could explain one of the biggest puzzles in physics and overturn the standard way that astronomers model the universe. Now, as results from new sky surveys trickle in, there are hints that there might be something in it.
Astronomers have known for nearly a century that space is expanding. Everything that isn’t gravitationally locked together is flying away from everything else. You would expect as much, given that the universe began in a big bang that kicked off the expansion. But in the mid-1990s, two independent research groups made a discovery so fundamental and surprising that it won them a Nobel prize.
It isn’t just that the cosmos is expanding, it is doing so at an increasing rate. Since there is no easy way to explain this, cosmologists suggested that space is filled with a mysterious “dark energy” pushing the universe apart faster and faster. The trouble is that there is no natural explanation for what this dark energy might be, despite decades spent thinking about it. With nothing better to fall back on, dark energy has become a key tenet of cosmology.
That is also the status of an assumption known as the cosmological principle. Introduced in 1933 by the British astrophysicist Edward Arthur Milne, this states that there are no special places in the universe, so conclusions drawn from our vantage point on Earth are universally true across all of space.
Take, for example, the age of the universe. Astronomers estimate that this is around 13.8 billion years. Even though it has been calculated using measurements taken from Earth, the cosmological principle states that we should get the same answer if we did the same thing from some distant, random locale in space. This is because, according to the cosmological principle, the universe is both isotropic and homogeneous on large scales. Isotropic means that it looks the same in all directions, whereas homogeneous means that it has the same properties throughout. “If I look at the distribution of galaxies on large scales, I see roughly the same number of galaxies when I look north as when I look south or east or west,” says , a cosmologist at the University of Chicago.
With this principle in place and coupled with Einstein’s general relativity – the theory of how gravity warps space and time – we end up with a model of cosmology that needs extra ingredients to explain our universe, namely dark matter and dark energy. The standard version of this model assumes that dark matter is made of heavy, sluggish particles known as cold dark matter (CDM), while dark energy is a constant energy field represented by the Greek letter lambda. That gives the standard cosmological model its name, lambda-CDM – and this is the lens through which most astrophysicists and cosmologists consider the universe. “It gives us a framework for understanding how structure formed and evolved in the universe which is consistent with observations,” says Frieman.
Expanding voids
The trouble with the cosmological principle, however, is that the universe is only isotropic and homogeneous on scales of around 400 million light years or more. Below this, things are very different from place to place. There are clusters of galaxies, which contain so much matter that they hold themselves together, detached completely from the expansion of the universe, and there are voids. These are vast, almost empty areas in which expansion is proceeding as normal. It’s a bit like how Earth resembles a perfect sphere from space. Yet, zoom in and the landscape of mountains and valleys disrupts the curvature.
In the mid-2000s, Wiltshire became aware of the work of , then at the CERN particle physics laboratory near Geneva, Switzerland, who was struck by this inhomogeneity. With voids thought to make up as much as 95 per cent of the volume of the universe, Buchert had set about calculating a better way to deal with them, rather than simply assuming the cosmological principle held. Wiltshire took Buchert’s methods, applied them to general relativity, and in 2007 the timescape model .
The trick behind it is an odd phenomenon called gravitational time dilation: the gravitational pull of massive objects warps space-time, resulting in time passing more slowly near such objects than it does further away. The stronger the gravitational field, the slower time passes. This concept itself is far from new, but cosmologists have generally assumed that the effects tend to smooth out over large scales, because the cosmological principle dictates that matter is roughly evenly distributed across the universe.
Throw out the cosmological principle, as Wiltshire and his colleagues advocate, and you can no longer assume this is so. If you want a model of cosmology that is consistent with this view of things, you have to get into the nitty-gritty of Einstein’s equations relating the properties of space-time to the amount of matter and energy situated within it. Those properties determine whether that part of the universe is expanding and if so at what rate, and also how fast time is passing in that region.
More time will have passed since the big bang in a void than in the Milky Way
Because there is hardly any matter in a void, time can progress very differently there. “Voids can be as much as 4 billion years older than [clusters of galaxies],” says , one of the team at the University of Canterbury. In other words, more time will have passed since the big bang in a void than in the Milky Way, and so the universe’s age will vary hugely depending on where you are located within it. There is no such thing as a single age for our cosmos.
This also means the space inside the voids has been expanding for up to 4 billion years more than we would think if we simply used the age of the universe calculated from within the Milky Way. Once we correct for this difference between locations, using the timescape model, the researchers claim that the need for dark energy disappears.
The idea has been a tough sell since it was first introduced – and not just for Wiltshire’s group. There had been earlier, different approaches to deal with the obvious inhomogeneities in cosmic structure, and none of those could move the needle of mainstream opinion either.
Explosions and ripples
Frieman is among those who remain unconvinced. “There’s been a whole literature on [inhomogeneous cosmologies],” he says. “When you include these large structures, could they actually affect the expansion of the universe in a way that mimics the effects of dark energy? To my mind, the preponderance of the literature indicates that that’s not happening.”
He isn’t alone in this view. Wiltshire, however, says that early efforts to construct inhomogeneous cosmologies by other groups only told half the story, because they continued to assume a constant age for the universe. What many haven’t realised, he says, is that timescape is different because it includes a varying age. This is what makes it all work, by allowing the voids to expand for billions more years than expected from our local calculation. “If you integrate these effects over billions of years, you get large differences,” says Wiltshire.
He may now have another opportunity to persuade his peers, thanks to a new dataset called Pantheon+. This contains observations of 1535 supernovae, or exploding stars, of a particular type, known as type 1a. All release essentially the same amount of energy, so any variation in their brightness is the result of their distance from us. This makes them an excellent way to measure the universe.
But the true value of the dataset for Wiltshire is that all of the observations have been calibrated to remove potential errors introduced by the supernovae being recorded with different telescopes. Such a large, accurately calibrated dataset finally allows a meaningful comparison between timescape and lambda-CDM.
Wiltshire and his colleagues’ , published in January, takes the supernovae of the Pantheon+ dataset and compares how well the lambda-CDM model of cosmology fits the data compared with the timescape model. They claim that their statistical analysis yields “very strong evidence in favour of timescape over lambda-CDM”.
However, this latest development still hasn’t convinced Frieman, who is director of the Dark Energy Survey (DES), a collaboration of more than 400 scientists from across the world. This is despite DES having its own collection of supernova data gathered between 2013-2019 that, when analysed alone, also favoured the timescape model over lambda-CDM. The reason for Frieman’s continued scepticism lies in measurements of different cosmological phenomena called baryon acoustic oscillations (BAOs). These measurements were also taken by DES and, when added to the collaboration’s supernova data, see the exact opposite conclusion reached about timescape.
BAOs can be thought of as ripples in the large-scale structure of the universe and are related to sound waves moving in the primordial plasma, the bath of super-hot particles that filled the universe in its early phases and eventually coalesced into the galaxies of today. This left density variations that acted as a blueprint for the pattern of clusters of material and the voids we see making up the landscape of the universe today.
When Ryan Camilleri at the University of Queensland, Australia, and his colleagues extracted the BAO data from the DES observations and the predictions of lambda-CDM and timescape, they found the standard cosmological model won convincingly. Wiltshire and his team want to check this result. “The most interesting thing that we could do at the moment is to approach that baryon acoustic oscillation question and see if this is true, that lambda-CDM inherently fits better,” says Ridden-Harper. The waves that left this imprint are related to the speed of sound in the primordial plasma, which has been carefully and precisely calculated within lambda-CDM, but not for the timescape model. So Wiltshire and his team are working on that now.
Crunch time
It is critical to get this right because there is so much at stake. The lambda-CDM model provides a foundation for our understanding of the universe, yet despite its many undoubted successes, there are a number of challenges to it emerging from cosmology – and not just from the timescape idea.
The bottom line is that lambda-CDM itself cannot fit all our observations. In 2005, at Johns Hopkins University in Maryland, who shared the Nobel prize for the discovery of dark energy, began drawing attention to a puzzle called the Hubble tension. This concerns the fact that our two main methods of calculating the current expansion rate of the universe, a figure called the Hubble constant, don’t match. The first method, using observations of supernovae in the relatively nearby universe, gives a value of about 73 kilometres per second per megaparsec. The other way to calculate the Hubble constant is to start with observations of the extremely distant cosmic microwave background – the faint radiation left over after the big bang – and use lambda-CDM to track the past 13 billion years or so of cosmic evolution to arrive at its present value. This says that the Hubble value today should be 67.7 km/s/Mpc. The mismatch in these figures is troubling. “In my mind it’s a hint of a crack [in the standard cosmological model],” says Riess.
It is a crack that Wiltshire thinks timescape cosmology might fill. Back when the radiation that makes up the cosmic microwave background was emitted, about 380,000 years after the big bang, the primordial plasma that filled the universe was very nearly homogeneous. But as gravity pulled matter together, it created structure and eventually gave the modern universe its inhomogeneous pattern of clusters and voids. The timescape model compensates for this evolution, whereas the standard model doesn’t. The upshot is that timescape cosmology predicts a higher Hubble constant today, explaining the Hubble tension, because of the additional expansion that has taken place in voids.
Crunch time for the timescape idea is just around the corner. In the next five years, more and better datasets will become available to cosmologists, and Wiltshire says these will be able to discriminate once and for all between his model and lambda-CDM. These include results from the Dark Energy Spectroscopic Instrument in Arizona, which is creating a map of the cosmos that’s expected to be complete by 2026, plus data from the European Space Agency’s Euclid space telescope and the giant Vera Rubin telescope in Chile.
Euclid is mapping the 3D structure of galaxies, which will give much more information about the BAO ripples. Rubin, which will begin its own vast sky survey this year, will be a supernova discovery machine. “These two are game changers – definitely a big leap in the context of cosmological galaxy surveys,” says at the University of Oxford, who researches galaxy evolution.
All this means that the future of cosmology hangs in the balance for the moment. But Wiltshire, at least, doesn’t mind the wait. He has already worked on his timescape hypothesis for almost two decades, so he is more than willing to wait another few years to find out if he has uncovered a secret of the universe that others have dismissed. “By the end of this decade,” he says, “we are going to know the answers.”
The science of deep time: Wales
Join a gentle walking-and-talking short break. Witness the cosmic, geological and human forces that have shaped our lives, understand humankind’s place in the story of Earth.
