Pulsar timing array

A pulsar timing array detects gravitational waves through correlated perturbations in pulse arrival times. Millisecond pulsars rotate hundreds of times per second and can act as highly stable celestial clocks, although their signals still require corrections for interstellar dispersion, Earth’s motion, and changing observatory conditions.¹ Astronomers build a timing model for each pulsar, then examine the residual differences between measured and predicted arrival times.

A passing gravitational wave slightly stretches and compresses spacetime along the paths between pulsars and Earth. The resulting timing residual is tiny for any single object, but a gravitational wave produces a distinctive angular correlation among many pulsars. This expected Hellings–Downs correlation distinguishes a gravitational-wave signal from independent noise, clock errors, and errors in the Solar System ephemeris.

PTA experiments combine long-term radio observations from multiple telescopes and pulsar-monitoring programs. The principal collaborations are the North American Nanohertz Observatory for Gravitational Waves (NANOGrav), the European Pulsar Timing Array, the Parkes Pulsar Timing Array, and the Indian Pulsar Timing Array; their data are also combined through the International Pulsar Timing Array.

In 2023, several collaborations reported strong evidence for a nanohertz gravitational-wave background, with the measured spatial correlations consistent with the Hellings–Downs pattern. The leading interpretation is the combined signal from many inspiraling supermassive black-hole binaries, although cosmic-string backgrounds and other early-Universe sources remain possible contributors. The evidence is statistical rather than an image or a separately resolved source.

Pulsar timing arrays probe a frequency range far below that targeted by ground-based interferometers such as LIGO. Their long observing baselines make them sensitive to gravitational waves with periods of years to decades, while the array’s sky coverage supplies directional information and tests of general relativity.

A measured background can constrain the population, merger history, and environments of supermassive black-hole binaries in distant galaxies. Individual loud binaries may eventually appear as resolvable continuous-wave sources, and extremely strong events could produce identifiable bursts. PTA observations also test models of the early Universe, including some phase transitions and cosmic-string networks, but these interpretations require separating astrophysical and instrumental effects.

The most difficult part of PTA work is often not detecting a periodic pulse but preserving its timing meaning over decades. Dispersion-measure variations caused by the interstellar plasma make lower-frequency radio pulses arrive later, so simultaneous or wideband observations are needed to track them. Pulse-profile changes, radio-frequency interference, telescope calibration, and uncertain planetary ephemerides can also mimic low-frequency signals.

The array is a distributed instrument: the pulsars are separated by thousands of light-years, while the telescopes may span continents. Each pulsar contributes a different geometric response, and an irregular observing schedule can produce uneven sensitivity across the sky. New instruments, broader radio bandwidths, improved Solar System models, and longer baselines are therefore as important as adding more pulsars. The method can also constrain errors in terrestrial time standards and planetary masses, making it useful beyond gravitational-wave astronomy.¹²

Millisecond pulsar

A rapidly rotating neutron star whose radio pulses are stable enough for precision timing.

Timing residual

The difference between an observed pulse arrival time and the arrival time predicted by a timing model.

Hellings–Downs curve

The angular correlation pattern expected between pulsar timing residuals produced by an isotropic gravitational-wave background.

Nanohertz gravitational wave

A gravitational wave with a frequency of approximately one-billionth of a hertz, corresponding to periods of years to decades.