Searching Pulsars for Planets


Are there more hidden exoplanets lurking around extreme pulsar hosts? A recent study explores a well-observed set of pulsars in the hunt for planetary companions.

Ushering in the Age of Exoplanets

pulsar timing array

An artist’s illustration showing a network of pulsars whose precisely timed flashes of light are observed from Earth. Could some of these pulsars host planets? [David Champion/NASA/JPL]

The first planets ever confirmed beyond our solar system were discovered in 1992 around the pulsar PSR B1257+12. By studying the pulses from this spinning, magnetized neutron star, scientists confirmed the presence of two small orbiting companions. Two years later, a third planet was found in the same system — and it seemed that pulsars showed great promise as hosts for exoplanets.

But then the discoveries slowed. Other detection methods, such as radial velocity and transits, dominated the emerging exoplanet scene. Of the more than 4,000 confirmed exoplanets we’ve discovered overall, a grand total of only six have been found orbiting pulsars.

Is this dearth because pulsar planets are extremely rare? Or have we just not performed enough systematic searches for pulsar planets? A new study led by Erica Behrens (The Ohio State University) addresses this question by using a unique dataset to explore rapidly spinning millisecond pulsars, looking for signs of hidden planets.

The Advantage of Precise Clocks

How are pulsar planets found? Pulsars have beams of hot radiation that flash across our line of sight each time they spin. The regularity of these flashes is remarkably stable, and when we observe them over long periods of time, we can predict the arrival time of the pulses with a precision of microseconds!


Sample periodograms for two pulsars. The top panel includes a simulated planet signal injected into the data, producing a strong peak at the planet’s orbital period. The bottom panel is an actual periodogram for one of the pulsars in this study, showing no evidence of a planetary companion. [Adapted from Behrens et al. 2020]

Because these pulses are so predictable, any perturbation that might change their timing can be measured and modeled. In particular, the presence of a companion body around the pulsar will cause both objects to orbit the system’s center of mass, introducing a periodic signature in the pulsar’s pulse arrival times. This fluctuation in the pulse timing allows us to measure the period and mass of potential companions.

A Multi-Use Dataset

To search for these signatures in pulse data, Behrens and collaborators turn to observations of 45 separate millisecond pulsars, which were made as part of the North American Nanohertz Observatory for Gravitational Waves (NANOGrav) project.

NANOGrav’s primary goal is to use the precise timing of these pulsars to measure the warping of spacetime caused by gravitational waves. But in the process of this work, the project has been carefully monitoring pulse arrival times for these pulsars for 11 years, producing a remarkably detailed dataset in which we can search for evidence of planets orbiting any of the 45 pulsars.

Pushing Down to Moon Masses

detection lower mass limits

Lower limits of detectable masses in the 11-year NANOGrav data set, as shown with black lines. The colored data shows the masses of the least massive 10% of confirmed exoplanets we’ve detected with other methods. Pulsar timing provides the ability to detect remarkably low-mass companion bodies. Click to enlarge. [Behrens et al. 2020]

Looking for periodic signals in the data, Behrens and collaborators rule out the presence of planets that have periods between 7 and 2,000 days. By injecting simulated signals into the data, the authors show that their analysis is sensitive to companions with masses of less than the Earth — in fact, for some pulsars, they’ve eliminated the possibility of all companions with more than a fraction of the mass of our Moon!

This study shows the incredible power and sensitivity of extended pulsar monitoring in the hunt for small exoplanets. While it may well be true that pulsar planets are very rare objects, those out there can’t stay hidden for long.


“The NANOGrav 11 yr Data Set: Constraints on Planetary Masses Around 45 Millisecond Pulsars,” E. A. Behrens et al 2020 ApJL 893 L8. doi:10.3847/2041-8213/ab8121