Old sky surveys add new colours to pulsar companions

What can a companion’s light reveal?

A pulsar is a rotating neutron star whose radiation is observed as regular pulses. When it shares an orbit with another star, the companion’s light offers another way to investigate the system. Measurements across different wavelengths can help constrain the companion’s temperature and evolutionary state, while changing brightness may reveal heating or distortion during its orbit.

In a November 2026 arXiv preprint, R. P. Mignani and colleagues revisit nine binary pulsars with already identified companions. Their study combines existing survey measurements rather than reporting a new observing campaign. Some companions gain ultraviolet or infrared coverage; other additions are optical photometry or repeated observations. The researchers also test whether scattered observations can recover brightness changes at known orbital periods.

The manuscript is a draft marked “to be submitted,” not a peer-reviewed result. Its strongest contribution is additional observational information, rather than completed new physical models of the stars.

Matching stars across surveys

The work follows an earlier search that compared 144 binary pulsars outside globular clusters with 11 survey catalogues. Nine systems offered additional measurements or repeated observations of known companions. The present paper deliberately retains that older sample rather than expanding it using today’s larger pulsar catalogue.

The catalogues include ultraviolet observations from GALEX, Swift and XMM-Newton, optical measurements from the Sloan Digital Sky Survey, and infrared observations from facilities including Spitzer and WISE. Together, these sample a spectral energy distribution: the amount of light received in different wavelength bands. Broad-band measurements provide less detail than a spectrum that separates light into many closely spaced wavelengths.

Matching positions was only the starting point. The researchers estimated accidental-alignment probabilities using nearby source densities, compared brightness measurements with previous work, and inspected available images. Their estimated chance-coincidence probabilities were typically around one in ten thousand to one in a thousand. These support the associations but do not eliminate instrumental problems or establish that every measured flux is reliable.

Indeed, image checks led them to reject several measurements affected by saturation, detector artefacts or problematic positions.

New wavelength coverage, not new companions

Among the additions is a likely near-ultraviolet detection of the white-dwarf companion to B0655+64. GALEX measured it at magnitude 24.06, a faint signal with an uncertainty of 0.26 magnitudes. The authors interpret the positional match as a probable first detection of this companion in that band.

For J0045−7319, whose massive companion lies in the Small Magellanic Cloud, the team reports additional near-ultraviolet coverage and Spitzer measurements at 3.6 and 4.5 micrometres. Such measurements extend the observational baseline beyond visible light; they do not, by themselves, establish a new stellar classification.

The timing of archival observations also matters. For J1023+0038, a system known to switch between states with and without an accretion disk, GALEX recorded near-ultraviolet light in March 2006 during a disk-free interval. That makes the measurement useful for studying the stellar contribution, although the authors note uncertainty about the pulsar’s own ultraviolet contribution. Combining observations from different dates without considering these state changes could produce a misleading picture.

Only one clear orbital brightness signal

For five systems, the researchers examined repeated Catalina survey measurements. They folded the observations at each binary’s known orbital period, placing measurements from different dates onto a common orbital cycle.

J1023+0038 showed a clear modulation of about half a magnitude, broadly matching previously published light curves. This is an independent recovery of known behaviour, not the discovery of a new orbital period.

J1816+4510 offered a less secure hint. After removing measurements probably belonging to a nearby brighter star, the researchers found a possible variation of roughly 0.2 magnitudes, with two maxima and two minima per orbit. That pattern could fit a distorted companion, but the data could not statistically distinguish their fitted model from constant brightness.

The remaining records were too noisy or sparse for convincing detections. Failure to detect modulation here does not demonstrate that a companion is intrinsically steady.

A useful reference with unfinished interpretation

These observations show both the value and limits of mining survey archives. Existing catalogues can extend wavelength coverage without dedicated telescope time, but uneven orbital sampling, uncertain associations and non-simultaneous measurements restrict interpretation.

Crucially, the manuscript says that detailed comparisons with stellar spectra and further classification analyses will appear in an appendix to a submitted version. Those analyses are not supplied in this revision. The present result is therefore a checked collection of additional measurements and variability constraints—not a completed reconstruction of the nine systems’ evolutionary histories.