The universe still has plenty of hydrogen, so why is it making fewer stars?

The universe is not producing stars at anything close to the rate it once did. Over the past 4.5 billion years, the cosmic star formation rate has fallen by more than half as galaxies have gradually become quieter.

A simple explanation would be that galaxies are running out of fuel. New measurements suggest the story is considerably more complicated.

An international team, including members from the Chinese Academy of Sciences (CAS), using China’s Five-hundred-meter Aperture Spherical radio Telescope, or FAST, and the Dark Energy Spectroscopic Instrument has found that the universe’s supply of neutral atomic hydrogen has declined only modestly over the same period. The work, published in Nature Astronomy as “Weak evolution of cosmic atomic hydrogen over the past 4.5 billion years,” places new limits on what can explain the fading rate of star formation.

The researchers measured a raw decline in cosmic neutral hydrogen density by a factor of 1.35 ± 0.10. After conservative corrections for possible systematic effects, that change fell to just 1.12 ± 0.10. By comparison, the cosmic star formation rate dropped by a factor of 2.46.

That mismatch makes rapid exhaustion of atomic hydrogen an unlikely primary cause of the universe’s declining stellar birth rate.

Neutral atomic hydrogen, known as H I, occupies a critical place in the life of a galaxy. Gas flowing into galaxies can become atomic hydrogen before some of it turns into molecular hydrogen, the denser material from which stars directly form.

Astronomers can detect H I through radio waves emitted at a wavelength of 21 centimeters. The problem is that signals from individual distant galaxies become exceedingly faint and can disappear into background noise.

Past surveys therefore faced an awkward tradeoff. Wide surveys could observe enormous regions but lacked enough sensitivity, while deeper observations covered areas too small to provide a representative cosmic census.

FAST and DESI offered a way around that problem.

The researchers combined the sensitivity of FAST’s All-Sky H I survey with highly precise galaxy redshifts from DESI’s Bright Galaxy Survey. Their cross-matched sample contained 2,473,945 galaxies across about 12,000 square degrees of sky and covered redshifts corresponding to the past 4.5 billion years.

Instead of trying to detect every galaxy separately, the team used spectral stacking. They aligned the radio observations using each galaxy’s known redshift, then combined enormous numbers of individually weak signals. Random noise tends to cancel as observations accumulate, allowing their shared hydrogen signal to emerge.

The approach produced a far larger and more statistically representative H I sample than earlier intermediate-distance measurements.

The result was striking because atomic gas behaved very differently from star formation.

At fixed stellar mass, the average H I gas fraction declined by less than 0.2 dex from a redshift of 0.41 to the present. The basic relationship between a galaxy’s stellar mass and its atomic gas content also remained surprisingly stable across the observed period.

That matters because a relatively stable total cosmic hydrogen density might otherwise conceal major shifts between different populations of galaxies. Instead, the weak evolution appears across a broad range of galaxy masses.

The study also examined galaxies containing more than 1 billion solar masses in stars. Those galaxies account for more than 80 percent of the cosmic star formation rate in the comparison used by the researchers, yet their H I density also changed only marginally.

The results do not mean galaxies have an unlimited supply of star-forming material. Atomic hydrogen is only one stage in a larger cycle, and stars form primarily from molecular hydrogen.

That distinction may hold the key to the cosmic slowdown.

Observations have already indicated that molecular gas changes much more strongly with cosmic time than atomic hydrogen does. Molecular hydrogen density has declined alongside star formation far more closely.

The new results therefore shift the question. Rather than asking why galaxies simply lost their hydrogen, astronomers can ask why less of their available atomic hydrogen is moving through the chain that produces molecular clouds and stars.

Gas continually moves among the intergalactic medium, the halos surrounding galaxies and the galaxies themselves. Inside galaxies, material can cycle between atomic and molecular forms while stellar activity, outflows and other processes return or remove gas.

The authors argue that declining gas accretion can still drive the late-time slowdown without rapidly draining the H I reservoir. As the growth of cosmic structures slows, less fresh material may flow toward galaxies. The effects can then appear farther downstream, reducing the throughput into molecular hydrogen and star-forming clouds while atomic gas remains closer to a regulated balance.

Several processes could influence that conversion, including changes in disk pressure, shielding, metallicity, turbulence and gas recycling. In massive halos, incoming material can also be heated, delaying its cooling and limiting the amount of cold material reaching galactic disks.

The result gives galaxy-formation models a demanding new target. Successful simulations must now reproduce two very different trends at the same time: nearly stable atomic hydrogen reservoirs and a much steeper decline in molecular gas and star formation.

Existing simulations do not produce identical answers. The study found that IllustrisTNG predicts an approximately constant cosmic H I density over the measured redshift range, while SIMBA predicts a modest decline of about 1.5 times, closer to the observations.

FAST and DESI have now supplied a much tighter observational benchmark against which those models can be tested.

For billions of years, galaxies have been slowing their production of new stars. Yet the new measurements show that the universe did not simply empty its atomic hydrogen tank along the way.

Much of the raw material is still there. The harder problem may be getting that material into the right form, at the right density, to make another generation of stars.

These studies examine the atomic and molecular gas reservoirs that connect the cosmic supply of matter to the rise and decline of star formation in galaxies.

Cosmic CO and [C II] backgrounds and the fuelling of star formation over 12 Gyr: This study measures the cosmic CO background and infers a larger molecular gas reservoir than galaxy surveys alone reveal, while finding a roughly 1-billion-year global depletion time that requires continued gas inflow. (Nature Astronomy, 2026)

The H I mass function of the Local Universe: Combining measurements from HIPASS, ALFALFA, and FASHI: By combining three major 21-centimeter surveys, this work produced a detailed census of atomic hydrogen in the nearby universe and provides an important local reference point for measurements of cosmic H I evolution. (Astronomy & Astrophysics, 2025)

Atomic hydrogen reservoirs in quiescent galaxies at z = 0.4: Spectral stacking shows that some quiescent galaxies at intermediate redshift can retain substantial atomic hydrogen, underscoring that possessing H I does not necessarily translate into vigorous star formation. (Astronomy & Astrophysics, 2025)

H I content at cosmic noon – a millimetre-wavelength perspective: This analysis examines how atomic and molecular hydrogen can be constrained during the era when cosmic star formation was near its peak, helping connect the long-lived H I reservoir with the molecular phase more directly associated with forming stars. (Monthly Notices of the Royal Astronomical Society, 2024)

The Cosmic Baryon and Metal Cycles: This review synthesizes observations of gas, stars and metals across cosmic history, finding that atomic hydrogen evolves relatively mildly while molecular hydrogen tracks the global star formation rate much more closely. (Annual Review of Astronomy and Astrophysics, 2020)

Research findings are available online in the journal Nature Astronomy.

The original story "The universe still has plenty of hydrogen, so why is it making fewer stars?" is published in The Brighter Side of News.

Like these kind of feel good stories? Get The Brighter Side of News' newsletter.