Inference of recoil kicks from binary black hole mergers up to GWTC--4 and their astrophysical implications

This paper infers recoil velocities for binary black hole merger events up to the GWTC-4 catalog, revealing that while most remnants are retained in massive galaxies, significant kick-induced displacements in dense stellar environments like globular clusters substantially suppress the likelihood of hierarchical mergers.

The Gist

Imagine two black holes dancing around each other in the dark, spiraling closer and closer until they finally crash together. When they merge, they don't just sit there quietly. Because of the way they spin and how heavy they are, the new, single black hole that is born gets a massive "kick" in the opposite direction. It's like a cannon firing a shell: the explosion sends the shell forward, but the cannon itself jerks backward.

This paper is a massive detective story about measuring exactly how hard these cosmic cannons recoil, using data from gravitational waves (ripples in space-time) detected by LIGO and Virgo. The author, Tousif Islam, looked at 183 different black hole mergers (including some brand new ones just discovered) to figure out how fast the new black holes are flying away.

Here is the breakdown of the findings in simple terms:

1. The "Kick" is Real and Sometimes Huge

When two black holes merge, they can shoot the new survivor out at speeds up to 5,000 kilometers per second (that's 11 million miles per hour!).

• The Detective Work: The author used a special mathematical "translator" to convert the messy data from the gravitational wave detectors into a speed estimate.
• The Big Winners: Most of the time, the data is too fuzzy to know the exact speed. But for a few specific events (like GW241011 and GW231123), they got a very clear answer. One of these new black holes is flying away at nearly 1,000 km/s. That is one of the fastest kicks ever recorded!

2. What Drives the Kick? (The Spin vs. The Size)

The author wanted to know: What makes the kick stronger? Is it the size of the black holes, or how they are spinning?

• The Analogy: Imagine two figure skaters holding hands and spinning. If they are spinning perfectly upright, they might wobble a bit when they let go. But if they are leaning over and spinning wildly in different directions, the "wobble" when they separate is huge.
• The Finding: The study found that the size difference between the two black holes and how fast they are spinning are the main drivers of the kick. The direction they are spinning (the angle) matters less than we thought because our detectors can't measure that angle very precisely yet.

3. The Great Escape: Will the Black Hole Stay Home?

This is the most dramatic part of the story. When a black hole gets kicked, does it stay in its neighborhood, or does it get kicked out of the house entirely?

• The Neighborhoods: Black holes live in different "neighborhoods" with different "fences" (escape velocities):
  • Globular Clusters (Old Star Cities): These have weak fences. The kick is usually so strong that 90-99% of the black holes get kicked right out of the neighborhood. They become "wandering" black holes, drifting alone through the galaxy.
  • Nuclear Star Clusters (Dense City Centers): These have stronger fences. About 15-30% of the black holes stay inside.
  • Elliptical Galaxies (Massive Super-Cities): These have huge fences. Almost 100% of the black holes stay put.

4. The "Grandchild" Problem: Can They Have More Babies?

Astronomers love the idea of "hierarchical mergers." This is when a black hole survives a kick, stays in the cluster, finds a new partner, and merges again to become a super-massive black hole.

• The Catch: Even if a black hole stays in the neighborhood (like a Globular Cluster), the kick often pushes it far away from the center of the cluster where all the other black holes are hanging out.
• The Analogy: Imagine a party in a crowded living room. If you get kicked out of the room, you can't meet anyone new. But even if you stay in the house, if the kick pushes you all the way to the attic, you won't meet anyone in the living room either.
• The Result: Because the kicks push the survivors so far away from the center, the chance of them finding a new partner and merging again is extremely low (less than 1% in small clusters). It's much more likely to happen in the dense centers of galaxies or around supermassive black holes.

Summary

This paper tells us that when black holes merge, they often get a violent shove.

1. Most of the time, we can't measure the speed perfectly, but we know it's happening.
2. Sometimes, the kick is so strong it launches the black hole out of its star cluster entirely, turning it into a lonely wanderer in the galaxy.
3. Even if they stay, the kick usually pushes them so far from the "party" (the center of the cluster) that they rarely get to meet a new partner to merge with again.

This helps scientists understand why we see so many black holes of certain sizes and why we don't see as many "super-heavy" black holes as we might have hoped from repeated mergers. The universe is a bit more chaotic than we thought!

Technical Summary

1. Problem Statement

Binary black hole (BBH) mergers conserve linear momentum, resulting in a "recoil kick" imparted to the remnant black hole. These kicks can reach velocities up to 5000 km/s, potentially ejecting the remnant from its host environment (e.g., globular clusters, nuclear star clusters, or galaxies).

• The Challenge: While recoil velocities are theoretically well-understood via Numerical Relativity (NR), inferring them from gravitational wave (GW) observations is difficult. Current detectors often have low-to-moderate signal-to-noise ratios, making direct measurement of Doppler shifts in the waveform impossible.
• The Gap: Previous studies have inferred kicks for subsets of events using inconsistent frameworks or different recoil models (e.g., switching between NR surrogates and semi-analytical models based on mass ratio). There was no unified analysis covering the entire GWTC-4 catalog, including recent O4b events and Intermediate Mass Black Hole (IMBH) candidates.

2. Methodology

The author employs a consistent, unified framework to infer recoil velocities for 183 events (GWTC-2.1, GWTC-3, GWTC-4, LVK IMBH candidates, and new O4b events).

A. Recoil Modeling Strategy

The paper maps intrinsic binary parameters $\{q, |\chi_1|, |\chi_2|, \theta_1, \theta_2, \phi_1, \phi_2\}$ to the recoil velocity $v_{kick}$. To ensure accuracy across the full parameter space, a hybrid approach is used:

• NRSur7dq4Remnant: Used for mass ratios $q \geq 0.1667$. This is a high-accuracy NR surrogate model trained on precessing binaries.
• HLZ (Herrmann-Lousto-Zlochower): A semi-analytical model calibrated to NR simulations, used for $q < 0.1667$ where NRSur7dq4 is not formally trained.
• Frame Evolution: A critical technical step involves evolving the binary parameters from the GW parameter estimation reference frequency ($f_{ref} \approx 20$ Hz) to the reference frame required by the recoil models (e.g., $t = -10M$ for NRSur7dq4 or separation $R_{sep}=10M$ for HLZ) using Post-Newtonian (PN) equations and NR surrogate spin evolution.

B. Inference and Validation

• Input Data: The framework utilizes posterior samples from various waveform models (IMRPhenomXPHM, SEOBNRv4PHM/5PHM, and NR surrogates) available in public catalogs.
• Information Content Analysis: To determine if the kick inference is data-driven or prior-dominated, the author computes the Jensen–Shannon Divergence (JSD) between:
  1. The inferred kick posterior and the prior ($JSD(v_{kick}, v_{prior})$).
  2. The inferred kick posterior and a "spin-isotropic" posterior where spin angles are drawn from the prior rather than the data ($JSD(v_{kick}, v_{iso})$).
• Astrophysical Modeling:
  • Retention Probability: Calculated as the cumulative distribution function (CDF) of the kick velocity relative to the escape velocity ($v_{esc}$) of various hosts (Globular Clusters, Nuclear Star Clusters, Dwarf Galaxies, Elliptical Galaxies).
  • Hierarchical Merger Probability: A phenomenological model estimates the probability of a retained remnant participating in a second merger. This accounts for the dynamical-friction return time ($t_{DF, return}$), which depends on how far the kick displaces the remnant from the cluster core.

3. Key Contributions

1. Unified Catalog Analysis: Provides the first consistent recoil kick estimates for all events up to GWTC-4, including 12 IMBH candidates and 3 new O4b events (GW241011, GW241110, GW250114).
2. Hybrid Modeling Framework: Establishes a robust prescription for switching between NRSur7dq4Remnant and HLZ models based on mass ratio, ensuring high accuracy across the entire BBH parameter space.
3. Quantification of Spin Angle Constraints: Demonstrates that current kick inferences are driven primarily by mass ratio and spin magnitudes, while spin orientation angles (tilt and azimuth) remain largely unconstrained and subdominant in most cases.
4. Hierarchical Merger Suppression Mechanism: Introduces a detailed analysis showing that even when remnants are retained in globular clusters, large recoil kicks often displace them significantly from the dense core, drastically reducing the probability of subsequent hierarchical mergers.

4. Key Results

A. Inferred Kick Velocities

• Informative Constraints: The analysis yields informative (non-prior-dominated) kick constraints for several events, notably:
  • GW231028_153006: $839^{+1018}_{-681}$ km/s
  • GW231123_135430: $974^{+944}_{-760}$ km/s
  • GW241011_233834: $974^{+555}_{-466}$ km/s (One of the largest kicks inferred to date).
• O4b Events:
  • GW241110_124123: $394^{+582}_{-207}$ km/s
  • GW250114_082203: $115^{+301}_{-95}$ km/s
• Systematics: GW241011_233834 shows significant waveform-model dependence, with inferred kicks varying from ~333 km/s to ~681 km/s depending on the waveform approximant used.

B. Astrophysical Implications

• Retention Probabilities:
  • Globular Clusters (GCs): $\sim 1\text{--}5\%$ retention.
  • Nuclear Star Clusters (NSCs): $\sim 15\text{--}30\%$ retention.
  • Dwarf Galaxies: $\sim 5\text{--}40\%$ retention.
  • Elliptical Galaxies: $\sim 70\text{--}100\%$ retention.
• Wandering Black Holes: For mergers occurring in GCs, $\sim 90\%$ of remnants are ejected from the cluster but remain bound to the host galaxy, wandering the galactic halo. In dwarf galaxies, a significant fraction may be ejected entirely from the galaxy.
• Hierarchical Mergers:
  • The probability of a remnant participating in a hierarchical merger is $\sim 0.1\text{--}1\%$ in Globular Clusters.
  • The probability rises to $\sim 1\text{--}15\%$ in Nuclear Star Clusters.
  • Reasoning: Even in GCs where retention occurs, the recoil kick often displaces the remnant beyond the core radius ($r_{max} > r_c$). The long dynamical-friction return times ($t_{DF} \sim 10^4$ Myr for $\omega$ Centauri) suppress the likelihood of the remnant returning to the dense core to form a new binary before the cluster evolves.

5. Significance

This paper establishes a critical baseline for interpreting the astrophysical origins of GW events. By demonstrating that spin orientation angles are currently unconstrained and that recoil kicks significantly suppress hierarchical merger rates in globular clusters (even for retained remnants), the work challenges simple assumptions about the formation of massive black holes in dense stellar environments.

The results suggest that:

1. Hierarchical mergers are more likely to occur in environments with higher escape velocities and shorter dynamical friction timescales, such as Nuclear Star Clusters or AGN disks, rather than typical Globular Clusters.
2. Future GW observations will need to improve constraints on spin orientations to better predict recoil kicks and the fate of merger remnants.
3. The inferred high-velocity kicks for specific events (like GW241011) provide strong evidence for the existence of black holes that may have been ejected from their birth clusters, potentially contributing to the population of isolated black holes detectable via microlensing.

The full dataset and code are publicly available, facilitating further studies into the population dynamics of black hole mergers.