Why black hole flickers form crosses, webs and diagonal ridges

What can a black hole’s flickering surroundings tell us that its average brightness cannot? In black hole X-ray binaries, where a black hole draws material from a companion star, astronomers find both irregular brightness variations and approximately repeating oscillations. The puzzle is not just which rhythms appear, but how they interact.

In an arXiv preprint, Seth Rowan Larner presents a mathematical model that reproduces several previously observed patterns in those interactions. The work was submitted to The Astrophysical Journal; the supplied record does not establish peer-reviewed acceptance. Its main result is a possible explanation supported by simulations, not a new observational confirmation.

Looking beyond the strength of a rhythm

Astronomers often describe a changing X-ray signal with a power spectrum, which shows how much variability occurs at each frequency. A quasi-periodic oscillation, or QPO, appears as a peak: a rhythm that repeats approximately rather than perfectly.

But different physical models can produce similar power spectra. Larner instead examines the bispectrum, which compares three frequency components: two frequencies and their sum. It asks whether their phases—the positions within their cycles—maintain a consistent relationship across successive stretches of data. Such relationships can reveal nonlinear coupling that a power spectrum misses.

A normalized measure of this relationship, called bicoherence, can be plotted on a map whose axes are the first two frequencies. Previously reported maps have recognizable shapes. A “cross” appears when one frequency equals the QPO frequency. A diagonal “hypotenuse” appears when two frequencies add up to the QPO frequency. A “web” contains both. A “point” marks coupling between the QPO and its second harmonic, an oscillation at twice its frequency.

These are shapes in a statistical map, not structures photographed around a black hole.

Three ingredients, including an unresolved one

The model begins with irregular fluctuations representing the hot inner accretion flow. Their brightness distribution is lognormal: the logarithm of the brightness follows a bell-shaped distribution. The QPO and its harmonic multiply this fluctuating signal rather than simply being added to it. That multiplication creates linked frequency components.

A second ingredient is crucial. Larner adds a response to the squared underlying noise field, filtered to concentrate power near the QPO frequency. This “quadratic” coupling provides another route for producing the diagonal ridge, allowing it to remain strong even when the cross fades.

The mathematics specifies how this response behaves, but does not identify its physical cause. Larner suggests that fluctuations in the inflowing gas might help drive the oscillation, potentially through torque on a precessing inner flow. That remains a proposed interpretation, not an established mechanism.

Finally, the model adds variability that is not coherently coupled to the inner-flow signal. Possible contributors include the disk or variable obscuration associated with winds. This diluting variability adds power without the same phase relationships, weakening the measured coupling.

Why the patterns change

Two competing effects control the model’s transitions. Reducing the QPO’s modulation amplitude weakens the cross, while the quadratic channel can preserve the diagonal ridge. Increasing uncoupled low-frequency variability suppresses the ridge more strongly than the cross. A balance can retain the web.

When the flow variability becomes very weak, both extended patterns disappear. Coupling between the QPO and its harmonic can remain, leaving the point pattern.

The model also offers an explanation for a previously reported viewing-angle association. Systems viewed closer to edge-on tend to have stronger type C QPOs, favoring the cross in this framework. Weaker QPOs in more face-on systems favor the ridge. Winds could reinforce the edge-on tendency if they supply the required variable dilution during intermediate states, but the paper stresses that wind evolution through these transitions is poorly understood.

What the simulations establish

Larner generated synthetic light curves and recovered all four patterns. A larger scan used 1,116 simulations to compare predicted and measured classifications. Among runs with detectable bicoherence, 96% matched the predicted state. This supports the analytical description within the chosen model; it is not a 96% success rate on real black hole observations.

Important simplifications remain. The calculations treat the QPO as a perfectly periodic tone, whereas real QPOs have finite frequency width. Other approximations simplify the noise and normalization, and observed patterns can depend on X-ray energy.

This account uses the supplied paper-text extraction, whose equations and layout may be damaged; the figures were unavailable for visual inspection. The reported numerical results therefore come from the text, not independently checked plots. The study’s contribution is a testable explanation for changing coupling patterns—not a settled answer to what creates the oscillations.