September 8, 2025
UChicago astrophysicists’ physics-based models suggest dark energy may be evolving
Dark energy—the term used to describe whatever is causing the universe to expand at an increasing rate—is one of the universe’s greatest mysteries. The widely accepted theory at the present moment suggests that dark energy is constant, with the energy of empty space driving cosmic acceleration. However, last year, findings from the Dark Energy Survey (DES) and Dark Energy Spectroscopic Instrument (DESI) sparked excitement within the cosmology community by hinting that dark energy may actually be evolving. “This would be our first indication that dark energy is not the cosmological constant introduced by Einstein over 100 years ago but a new, dynamical phenomenon,” said Josh Frieman, Professor Emeritus of Astronomy and Astrophysics.
A new paper published in Physical Review D in September, by Frieman and Anowar Shajib, a NASA Hubble Fellowship Program Einstein Fellow in Astronomy and Astrophysics, combines current data from a multitude of probes, finding that dynamical models of evolving dark energy can better explain the data than the cosmological constant.
Shajib’s primary research focuses on observational cosmology and galaxy evolution, using strong gravitational lensing to measure the Hubble constant and constrain dark energy parameters. Frieman’s observational cosmology research leverages large cosmic surveys such as the Sloan Digital Sky Survey (SDSS) and the DES, with a particular emphasis on exploring the origin and evolution of the universe as well as understanding the nature of dark energy.
We spoke with Shajib and Frieman about the new models described in their paper, the implications of these results, and what’s next.
Why is dark energy significant in the study of the universe?
Frieman: We now know precisely how much dark energy there is in the universe, but we have no physical understanding of what it is. The simplest hypothesis is that it is the energy of empty space itself, in which case it would be unchanging in time, a notion that goes back to Einstein, Lemaitre, de Sitter, and others in the early part of the last century. It’s a bit embarrassing that we have little to no clue what 70 percent of the universe is. And whatever it is, it will determine the future evolution of the universe.
What recent findings led cosmologists to consider that dark energy may be evolving?
Shajib: Although there has been interest in the dynamical nature of dark energy since its discovery in the ’90s to resolve some observational discrepancies, until recently, most of the major and robust datasets were consistent with a non-evolving dark energy model, which is accepted as the standard cosmology. However, interest in evolving dark energy was vigorously rekindled last year from the combination of supernovae, baryon acoustic oscillation, and cosmic microwave background data from the DES, DESI, and Planck experiments. This combination of datasets indicated a strong discrepancy with the standard, non-evolving model of dark energy. The interesting feature of non-evolving dark energy is that its density stays constant through time even though space is expanding. However, for the evolving dark energy model, dark energy density will change with time.
Frieman: The data from these surveys allow us to infer the history of cosmic expansion—how fast the universe has been expanding at different epochs in the past. If dark energy evolves in time, that history will be different than if dark energy is constant. The cosmic expansion history results suggest that over the last several billion years or so, the density of dark energy has decreased by about 10 percent—not much, and much less than the densities of other matter and energy, but still significant.
What was the goal of this study, and what were the overall findings?
Shajib and Frieman: The goal of this study is to compare the predictions of a physical model for evolving dark energy with the latest data sets and to infer the physical properties of dark energy from this comparison. The evolving dark energy “model” used in most previous data analyses is just a mathematical formula that isn’t constrained to behave as physical models do. In our paper, we directly compare physics-based models for evolving dark energy to the data and find that these models describe the current data better than the standard, non-evolving dark energy model. We also show that near-future surveys such as DESI and the Vera Rubin Observatory Legacy Survey of Space and Time (LSST) will be able to definitively tell us whether these models are correct or if, instead, dark energy really is constant.
Describe the models presented and why they better explain the behavior of dark energy compared to existing models.
Frieman: These models are based on particle physics theories of hypothetical particles called axions. Axions were first predicted by physicists in the 1970s, who sought to explain certain observed features of strong interactions. Today, axions are considered plausible candidates for dark matter, and experiments worldwide are actively searching for them, including physicists at Fermilab and the University of Chicago.
The models in our paper are based on a different, ultra-light version of the axion that would act as dark energy, not dark matter. In these models, dark energy would, in fact, be constant for the first several billion years of cosmic history, but the axion would then start to evolve—like a ball on a sloping field that’s released from rest and starts to roll—and its density would slowly decrease, which is what the data appear to prefer. So the data suggest the existence of a new particle in nature that’s about 38 orders of magnitude lighter than the electron.
What are the implications of these findings for understanding cosmic expansion?
Shajib: In these models, the dark energy density decreases with time. Dark energy is the reason for the universe’s accelerated expansion, so if its density decreases, the acceleration will also decrease with time. If we consider the very far future of the universe, different characteristics of dark energy can lead to different outcomes. Two extremes of these outcomes are a Big Rip, where the accelerated expansion itself accelerates to the point that it rips everything apart, even atoms, and a Big Crunch, where the universe stops expanding at some point and recollapses, which will look like a reverse Big Bang. Our models suggest that the universe will avoid both of these extremes: it will undergo accelerated expansion for many billions of years, yielding a cold, dark universe—a Big Freeze.
Could these results have other, less apparent implications?
Frieman: The only practical implications I can imagine are the technologies we need to develop to explore these ideas further—building new telescopes, launching new satellites, or developing novel detectors, for example. Such developments are likely to have much more of an impact on our lives than events happening trillions of years in the future.
What excites you the most about these results?
Shajib: For this paper, we gathered all the major data sets—from the DES, DESI, SDSS, Time-Delay COSMOgraphy, Planck, and Atacama Cosmology Telescope—and combined them to get the most constraining measurement of dark energy to date. All these measurements come from extensive experiments, so in a way, they represent the collective knowledge that the cosmological community has gathered as a whole.
Frieman: When we began working on the DES in 2003, our goal was to constrain the properties of dark energy to determine whether it was constant or changing. For two decades, the data indicated that it was constant. We almost gave up on that question because the data consistently supported the assumption. However, we now have the first hint in over 20 years that dark energy might be changing, and if it is evolving, it must be something new, which would change our understanding of fundamental physics. That feeling is reminiscent of where we were at the beginning. It could still turn out that these hints are incorrect, but we may be on the cusp of answering that question, and that’s quite exciting.
Citation: “Scalar field dark energy models: Current and forecast constraints.” Anowar J. Shajib and Joshua A. Frieman, Phys. Rev. D 112, 063508.