⚛️ general relativity

Identifying Exceptional Points in Black Hole Ringdowns

This paper proposes a method to identify exceptional points in black hole ringdowns by utilizing time-weighted Jordan terms and splitting bounds, demonstrating how omitting ordinary responses can enhance evidence for these points while deriving specific distance metrics between split and exceptional point waveforms.

Original authors: Zhen-Xiao Zhang, Chen Lan, Yi Wang

Published 2026-10-02

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Zhen-Xiao Zhang, Chen Lan, Yi Wang

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

When a black hole is disturbed, perhaps by swallowing a star or colliding with another black hole, it does not simply settle down immediately. Instead, it rings like a struck bell, emitting gravitational waves that carry a unique signature of its internal structure. This phase of ringing, known as the "ringdown," is a powerful tool for astronomers because the specific tones it produces depend entirely on the black hole's mass and spin. In the standard view of physics, these tones are distinct and independent, fading away at their own rates. However, a peculiar phenomenon predicted by mathematics suggests that under very specific conditions, two of these tones can merge into a single, hybrid state. When this happens, the black hole's response changes character, producing a signal that grows slightly stronger for a moment before fading, rather than just fading away. This merging point is called an exceptional point, and finding it would reveal new details about the fundamental degrees of freedom that govern how black holes relax back to stability.

The challenge for scientists is that these merged states are incredibly difficult to distinguish from two separate tones that are simply very close together. Over a short period of observation, a pair of distinct tones can mimic the behavior of a merged one so closely that the difference is hidden by the noise of the detectors. A new study by Zhen-Xiao Zhang, Chen Lan, and Yi Wang tackles this problem by developing a rigorous method to tell the difference between a true merger and a near-miss. They focused on the specific mathematical shape of the signal, looking for a "time-weighted" component that only appears when modes have truly coalesced. This component, often called a Jordan term, acts like a fingerprint of the merger. The researchers realized that if an observer ignores other, quieter parts of the signal that are not part of the main ringing pair, they might be tricked into thinking a split pair is actually merged. These ignored parts, which the authors call "ordinary responses," can interfere with the main signal in a way that fakes the evidence for a merger.

To solve this, the team created a procedure that accounts for these ordinary responses and the inevitable noise in the data. They tested their method using computer simulations of black holes, evolving the equations of gravity to generate complete waveforms that included both the main ringing and the background noise. Their simulations showed that if one simply tries to fit the data with a model that ignores the extra background components, the model will often falsely claim to have found a merger, even when the black hole is actually vibrating with two distinct tones. The study demonstrated that including these extra components in the analysis is crucial; while it makes the math more complex, it prevents the model from absorbing the differences between a split pair and a merged pair. Without this completeness, the evidence for a merger becomes unreliable, and the bounds on how close the tones are become deceptively tight.

The researchers also discovered that the ability to resolve these tiny differences depends heavily on how the black hole was excited in the first place. If the black hole is shaken in a way that produces a specific balance between the constant part of the signal and the time-weighted part, the difference between a split and a merged state becomes much harder to see, scaling down to a very small value. However, if the black hole is excited in a way that creates different balances across different detectors or polarization directions, the difference becomes much larger and easier to detect. By using independent information, such as knowing the exact frequency of the ring from other measurements, the team showed that scientists can prevent the model from "hiding" the split by adjusting its parameters. This allows the true difference to remain visible, scaling with the fourth power of the separation rather than the sixth, making it significantly easier to spot small splits.

Ultimately, this work provides a clear roadmap for how to identify exceptional points in real gravitational-wave data. It establishes that a convincing identification of a merged state must survive the freedom needed to describe all the other parts of the signal. The study confirms that while a nearby split pair can look like a merger, a careful analysis that includes all ordinary responses and uses independent constraints can reveal the truth. This means that future observations of black hole ringdowns can be designed to probe the deepest degrees of freedom of spacetime, distinguishing between a simple pair of vibrations and a true, singular state of coalescence. The findings suggest that with the right data and the right model, the subtle signatures of these exotic points are within reach, offering a new way to test the limits of our understanding of black holes.

Technical Summary: Identifying Exceptional Points in Black Hole Ringdowns

Problem Statement
Black hole (BH) ringdown spectroscopy probes the dynamical degrees of freedom of a perturbed BH remnant. In non-Hermitian spectral problems arising from radiation escaping to infinity and entering the horizon, distinct modes may coalesce at an exceptional point (EP). At a second-order EP, two modes merge in both frequency and damping rate, and their spatial profiles combine, leaving only one independent mode. The response at an EP is characterized by a "Jordan term," a time-weighted contribution ( $\tau e^{-\gamma \tau}$ ) multiplying the standard damped oscillation.

A primary challenge in identifying EPs is distinguishing a true coalescence from a "split pair" of modes with very small separation. Over a finite observation window, a nearby split pair can reproduce the leading time dependence of an EP response, making them observationally degenerate. Furthermore, the presence of "ordinary responses"—damped components outside the tested EP cluster—can interfere coherently with the signal. Omitting these ordinary responses can artificially strengthen evidence for the Jordan term and spuriously tighten the bounds on mode splitting, leading to false identification of coalescence. Conversely, including them with free parameters can absorb the distinguishing features of the split pair, weakening the resolution required to distinguish a split from a coalescence.

Methodology
The authors develop a rigorous procedure for identifying EPs in gravitational-wave waveforms that accounts for finite observation windows, noise, polarization readout, and the interference of ordinary responses.

Joint Criterion for Identification: The authors define a statistical criterion requiring three conditions to be met simultaneously:
- Jordan Evidence ( $T_J$ ): Significant improvement in log-likelihood when allowing a Jordan term compared to a single-exponential model.
- Splitting Compatibility ( $T_\chi$ ): Compatibility with zero splitting (the EP hypothesis) within a prescribed confidence region.
- Splitting Bound: The upper bound on the allowed mode separation ( $\chi_{up}$ ) must be sufficiently small ( $\chi_{up} + \bar{\chi} < \chi^*$ ), ensuring the data supports a resolution claim.
Waveform Distance Analysis: To determine the measurability of splitting, the authors derive the squared distance ( $d^2_{EP}$ ) between a split waveform and the best-fitting EP waveform. This distance is minimized over the EP center (mean frequency and decay rate) and amplitudes.
- They analyze how the "relative Jordan excitation" (the ratio of the Jordan term to the constant term) scales with mode separation ( $\epsilon$ ).
- When the relative Jordan excitation decreases linearly with separation (a common physical path), fitting the EP center absorbs the leading quadratic time dependence. This leaves a residual scaling as $\epsilon^3$ , resulting in a squared distance scaling as $d^2_{EP} \propto \epsilon^6$ .
- This sixth-order scaling implies that distinguishing a split pair from an EP becomes exponentially harder as separation decreases, requiring signal-to-noise ratios (SNR) scaling as $\rho \propto \epsilon^{-3}$ .
Role of Ordinary Responses: The study uses numerical simulations of Einstein–Maxwell–Scalar (EMS) black holes to test how omitting or including ordinary response terms affects the fit.
- Omission Bias: Excluding coherent ordinary responses can reinforce the Jordan-like part of the signal while canceling the time dependence that distinguishes a split pair. This leads to false EP-compatible outcomes even when the true signal is split.
- Model Completeness: Including additional ordinary response terms with free frequencies and amplitudes allows the model to absorb the small distinguishing residuals of a split pair, widening the allowed splitting region and reducing resolution.
Constrained Centers: The authors propose that independent constraints on the EP center (e.g., from other modes or consistency between distinct responses) can prevent the center shift from absorbing the splitting dependence.
- If the center is fixed or constrained, the leading $\epsilon^2$ waveform difference is preserved.
- This restores a fourth-order scaling ( $d^2_{EP} \propto \epsilon^4$ ), reducing the SNR requirement to $\rho \propto \epsilon^{-2}$ .
- This improvement is maximized when different excitations of the same EP cluster (e.g., different polarization channels or independent sources) have distinct relative compositions of Jordan and constant terms, preventing a single center shift from compensating for all mismatches.

Key Results

Resolution Limits: The paper demonstrates that without constraints, the resolution for identifying EPs is limited by the sixth-order scaling of the waveform distance. For a mode separation $\epsilon$ , the required SNR scales as $\epsilon^{-3}$ .
Bias Quantification: Numerical experiments show that omitting a single ordinary response term can cause a truly split signal to pass EP identification criteria in nearly all trials (196/200), whereas including the term correctly rejects the EP hypothesis.
Restoration of Sensitivity: By applying independent constraints on the center (e.g., fixing the mean frequency and decay rate), the authors show that the scaling reverts to fourth order. In simulated Advanced LIGO noise scenarios, this constraint reduces the required SNR for a given resolution by an order of magnitude compared to the unconstrained case.
Excitation Diversity: The study finds that utilizing multiple readouts (e.g., polarization projections) with distinct excitation compositions can naturally preserve the fourth-order term, even without external center measurements, provided the detector network retains these differences.

Significance and Claims
The paper claims to provide a necessary framework for distinguishing true mode coalescence from nearby split pairs in black hole ringdowns. It argues that convincing identification of an EP must survive the freedom required to model ordinary responses; otherwise, the evidence may be spurious.

The authors emphasize that:

Model Completeness is Critical: Reliable inference requires modeling all relevant ordinary responses. Omitting them creates a bias that can mimic an EP.
Independent Information is Key: Constraints on the EP center (from other modes or theoretical consistency) are essential to prevent the "compensation" of splitting effects by center shifts, thereby restoring the sensitivity needed to resolve small splittings.
Physical Realization: The proposed methods are tested on complete waveforms from EMS black holes and simulated tensor strain, demonstrating that the theoretical scaling laws hold in realistic observational contexts.

The work does not propose new experimental hardware or specific future missions but rather refines the data analysis methodology required to interpret existing and future gravitational-wave observations in the context of non-Hermitian spectral physics. It establishes the analytic and numerical conditions under which an EP can be distinguished from a split pair, highlighting the trade-off between model flexibility and resolution.

More like this