Pandora’s Cluster reveals hot gas and a weak merger shock

What can a collision between galaxy clusters reveal when one of its expected signals remains undetected? In Abell 2744, the answer includes evidence for unevenly heated gas, a weak merger shock and a lower bound on magnetic-field strength.

Randall A. Rojas Bolivar and colleagues report the results using NASA’s NuSTAR X-ray telescope to examine this complicated merging system, also known as Pandora’s Cluster. This account examines their arXiv v1, posted on November 6, 2026. The landing-page metadata also lists the related DOI 10.3847/1538-4357/aebd83; journal publication and peer-review status remain unconfirmed by the supplied evidence.

Reading the gas between galaxies

Much of a cluster’s X-ray emission comes not from individual galaxies but from the thin, hot plasma between them: the intracluster medium. Cluster mergers drive shocks and turbulence through this gas. Measuring its temperature helps researchers investigate how the collision’s energy is distributed.

Abell 2744 also contains diffuse radio emission. Fast electrons spiralling in magnetic fields produce this radiation through a process called synchrotron emission. Those electrons should also transfer energy to photons from the cosmic microwave background, the radiation left over from the early Universe. This second process, inverse Compton scattering, can boost photons into the X-ray band.

The difficulty is separating that faint contribution from the much brighter radiation produced by hot gas.

Comparing explanations for the X-rays

The researchers analysed two NuSTAR observations taken in December 2024. After removing intervals affected by elevated background radiation, they retained exposures of 127,000 and 54,000 seconds. They fitted spectra between 3 and 20 kiloelectronvolts, or keV; the signal above that range was too weak for their spectral analysis.

They compared three descriptions: gas at one temperature, gas represented by two temperatures, and a thermal component combined with a power-law spectrum representing possible inverse Compton emission.

The single-temperature model gave about 10.9 keV. Here, keV expresses temperature as a particle-energy scale. The two-temperature model fitted better, with components near 17.2 and 2.9 keV. These components offer a simplified description of a complicated mixture, not proof that all the gas occupies exactly two temperatures.

Adding an inverse Compton component did not provide a better fit than the two-temperature thermal model. The team therefore found no significant detection of that non-thermal signal.

Background modelling was central to this conclusion. The authors repeated the fitting with 1,000 simulated variations in background levels. The two-temperature model remained preferred over the thermal-plus-inverse-Compton alternative throughout those trials.

Measuring a previously identified shock

The team also investigated a shock near the southeastern boundary of the cluster’s central radio halo. This is distinct from another, weaker shock associated with a southeastern radio relic farther out.

Their preliminary temperature map could not clearly resolve individual shock fronts. Rather than treating apparent hot patches as discoveries, they adopted the upstream and downstream regions identified in earlier Chandra work. They also modelled how NuSTAR’s broad imaging response scatters photons between regions, which could otherwise blur a temperature jump.

The fitted temperatures rose from 7.90 keV ahead of the shock to 12.15 keV behind it. The authors report a Mach number of 1.58, with a 90% confidence interval extending from 1.28 to 2.00. Mach number compares a shock’s speed with the sound speed in the upstream gas. This result describes a weak shock and agrees, within uncertainties, with previous measurements.

What the missing signal constrains

The authors report a 90% confidence upper limit of 2.31 × 10⁻¹² erg per second per square centimetre for inverse Compton emission in the 20–80 keV band. Crucially, this is not a direct measurement in that band: it comes from extending the assumed power-law model beyond the fitted 3–20 keV range. The power-law slope was fixed using radio information.

Combining that limit with published VLA diffuse radio measurements gives a reported lower bound of roughly 0.2 microgauss on the average magnetic field. The inference depends on the assumed electron spectrum and the correspondence between the radio- and X-ray-emitting volumes. It does not measure the field everywhere or establish how much pressure magnetic fields provide.

The supplied PDF extraction contains damaged figure labels and potentially unreliable equation formatting; this account follows the written results rather than independently reconstructing those graphics or calculations. The central distinction remains important: NuSTAR detected hot gas and measured a temperature jump, but constrained—rather than detected—the inverse Compton component.