Gravitational-wave imprints of Kerr-Bertotti-Robinson black holes: frequency blue-shift and waveform dephasing

The dynamics of the inspiral are bounded by the innermost stable circular orbit (ISCO). Recently, Wang [Wang:2025bjf] derived an exact analytical solution for the ISCO radius in the Kerr-BR spacetime. Remarkably, when expressed in terms of the outer and inner horizon radii $r_{\pm}$ (the roots of $\Delta(r)=0$), the ISCO radius $r_{\rm ISCO}$ takes a form formally identical to the standard Kerr case:

where the upper (lower) sign corresponds to prograde (retrograde) orbits. The auxiliary functions are defined as

Equation (6) provides the precise termination point for our inspiral evolution.

To describe the adiabatic evolution before reaching the ISCO, we require the orbital frequency $\Omega_{\phi}$ and conserved quantities $(\mathcal{E},\mathcal{L})$ for general circular orbits at any radius $r>r_{\rm ISCO}$.

The orbital frequency $\Omega_{\phi}=\mathrm{d}\phi/\mathrm{d}t$ is determined by the extremum condition of the effective potential, $\partial_{r}V_{\rm eff}=0$. For a generic stationary and axisymmetric spacetime, this condition yields a closed-form expression in terms of the radial derivatives of the metric components ($g^{\prime}_{\mu\nu}\equiv\partial_{r}g_{\mu\nu}$) [Chandrasekhar:1985kt]:

This relation thus gives the exact Keplerian frequency for equatorial circular geodesics in the magnetized background. By substituting Eqs. (5) into Eq. (8), we obtain the precise frequency evolution used in our waveform modeling. In the limit $B\to 0$, Eq. (8) reduces to the familiar Kerr expression $\Omega_{\phi}=\sqrt{m}/(r^{3/2}+a\sqrt{m})$.

The corresponding specific energy $\mathcal{E}$ and angular momentum $\mathcal{L}$ for circular orbits follow from the normalization condition $u^{\mu}u_{\mu}=-1$:

where the normalization factor $\mathcal{N}$ is defined as

The effective potential $V_{\rm eff}(r)$ governing the radial motion for a particle with fixed $\mathcal{E}$ and $\mathcal{L}$ can then be written as

Figure 1 illustrates the behavior of the effective potential for a fixed angular momentum $\mathcal{L}=2.8$. As the magnetic field $B$ increases, the potential well deepens and its minimum shifts towards smaller radii, indicating enhanced confinement for particles with fixed $\mathcal{L}$.

In Fig. 2, we present the normalized ISCO radius $r_{\rm ISCO}(B)/r_{\rm ISCO}(0)$ as a function of the magnetic field parameter $B$. We examine three representative cases: prograde $(a=0.9)$, Schwarzschild $(a=0)$, and retrograde $(a=-0.9)$.

A key finding is that the magnetic field systematically increases the ISCO radius, $r_{\rm ISCO}(B)>r_{\rm ISCO}(0)$, across all spin configurations sampled. This indicates that the uniform magnetic field provides additional radial support that destabilizes the innermost circular orbits and pushes the marginal stability limit outward. Interestingly, the relative shift is most pronounced for the Schwarzschild case ($a=0$), reaching a deviation of over $10\%$ at $B=0.3$. In contrast, both highly prograde and retrograde orbits show a more moderate normalized increase. This suggests that the non-spinning potential well is comparatively “softer” against magnetic perturbations relative to its initial radius, whereas the strong frame-dragging effects in high-spin scenarios (whether co- or counter-rotating) impart a rigidity to the spacetime geometry that suppresses the relative radial expansion.

Standard intuition from Keplerian dynamics suggests that an orbital expansion leads to a decrease in frequency, $\Omega\sim r^{-3/2}$. In the Kerr-BR spacetime, however, we find the opposite trend at the ISCO.

As shown in Fig. 3, the ISCO orbital frequency $m\Omega_{\rm ISCO}$ monotonically increases with the magnetic field strength for all spins considered. This “frequency blue-shift” implies that the magnetic-field–induced curvature corrections in the Kerr-BR metric force the particle to orbit faster in order to maintain a circular geodesic at the outwardly shifted ISCO. The inset in Fig. 3 confirms that even for high-spin prograde orbits $(a=0.9)$, where the vacuum frequency is already high, the magnetic field induces a further blue-shift.

Most remarkably, we observe a generic “crossover” phenomenon. Orbits corresponding to lower spins, which initially have lower ISCO frequencies at $B=0$, exhibit a much steeper rise in frequency as $B$ increases, eventually overtaking the frequencies of higher-spin orbits. Specifically, we identify three crossover events: (1) the retrograde orbit $(a=-0.9)$ surpasses the Schwarzschild case $(a=0)$ at $B\approx 0.12$; (2) at $B\approx 0.20$, the retrograde frequency exceeds even the high-spin prograde orbit $(a=0.9)$; and (3) at $B\approx 0.26$, the Schwarzschild frequency overtakes the prograde one. Thus, beyond a certain field strength, the magnetic corrections completely override the usual spin-imposed ordering of ISCO frequencies.

This counterintuitive behavior and the crossover phenomenon can be understood by examining how the magnetic field couples to the orbital radius. In the Kerr-BR metric, the leading magnetic modification enters through terms scaling as $\sim B^{2}r^{2}$, such as in the conformal factor $\Lambda=1+B^{2}r^{2}$. This dependence introduces a fundamental asymmetry in how the magnetic field affects orbits at different radii. The factor $B^{2}r^{2}$ acts as a confining geometrical contribution: for orbits at larger radii this term is significantly larger, implying that the magnetic field produces a stronger effective radial steepening of the potential than it does near the black hole.

To maintain a stable circular orbit against this enhanced radial confinement, the test particle requires a higher angular velocity to provide sufficient centrifugal support. This leads to a differential sensitivity of the frequency to $B$ depending on the ISCO location. Prograde orbits reside deep in the gravitational potential well at small radial coordinates, where the environmental correction factor $B^{2}r^{2}$ remains modest. In contrast, retrograde orbits are located at significantly larger radii (typically nearly an order of magnitude larger than their prograde counterparts). Consequently, in this outer region, the term $B^{2}r^{2}$ is dramatically amplified, subjecting the retrograde orbits to a much stronger magnetic confinement effect. This enhanced sensitivity $(\partial\Omega_{\rm ISCO}/\partial B)$ allows the retrograde frequency to rapidly “catch up” and eventually overtake the less sensitive prograde frequency, resulting in the observed frequency crossover.

To obtain a more global picture of the magnetic imprint, we extend our analysis to the entire spin interval $a\in(-1,1)$. Figure 4 summarizes the interplay between the black hole’s intrinsic angular momentum and the external magnetic field.

We first focus on the frequency behavior shown in the bottom panel of Fig. 4. The red shaded region highlights a generic “magnetic hardening” of the gravitational-wave spectrum: for any given spin $a$, the magnetized ISCO frequency (red solid line) is higher than the corresponding vacuum value (black dashed line). The separation between the two curves becomes increasingly pronounced in the retrograde regime $(a<0)$, whereas they tend to merge for rapidly spinning prograde orbits $(a\to 1)$. This behavior confirms that retrograde orbits act as hypersensitive probes of environmental magnetic fields, as the magnetic confinement effect dominates over the comparatively weaker gravitational binding at larger radii.

The top panel of Fig. 4 displays the corresponding shift in the ISCO radius. While the magnetic field invariably pushes the ISCO outward $(r_{\rm mag}>r_{\rm vac})$, the magnitude of this shift exhibits a subtle, nonmonotonic dependence on spin, as detailed in the inset. Notably, the absolute deviation $\Delta r_{\rm ISCO}=r_{\rm mag}-r_{\rm vac}$ does not peak at the maximally retrograde spin $(a=-1)$, as one might naively expect from the large orbital radius alone. Instead, the maximum deviation occurs at an intermediate retrograde spin, approximately $a\approx-0.26$.

This nonmonotonic behavior can be qualitatively understood as the outcome of two competing physical mechanisms. First, as $a$ becomes more retrograde, the ISCO radius $r$ increases. Since the magnetic metric corrections scale as $\sim B^{2}r^{2}$, larger radii provide a longer “magnetic lever arm” to destabilize the orbit, tending to increase $\Delta r_{\rm ISCO}$. Second, as $a\to-1$, the counter-rotating frame-dragging effects become dominant, creating a “stiffer” effective potential well. In this regime, the ISCO location is strongly dictated by the background spacetime geometry, making it increasingly resistant to external magnetic perturbations. The peak at $a\approx-0.26$ therefore represents a balance point between the growing magnetic leverage and the increasing rigidity of the underlying gravitational potential.

We model the inspiral phase under the adiabatic approximation, assuming that the radiation–reaction timescale is much longer than the orbital period $(T_{\rm rad}\gg T_{\rm orb})$. The trajectory of the compact object is then determined by the energy-balance equation

To maximize the fidelity of the orbital motion, we treat the conservative and dissipative sectors with different levels of approximation. For the conservative sector, we use the exact analytical expressions for the specific energy $\mathcal{E}(r;B)$ and orbital frequency $\Omega_{\phi}(r;B)$ derived in Sec. 3. This ensures that the orbital kinematics fully reflect the magnetically distorted geometry, without invoking weak-field or slow-motion expansions. For the dissipative sector, we employ the leading-order relativistic quadrupole formula for the gravitational-wave energy flux,

where $\eta=\mu/m\ll 1$ is the symmetric mass ratio.

The evolution of the orbital radius is then obtained from

Here, the radial derivative $\partial\mathcal{E}/\partial r$ is computed symbolically from the metric components [Eqs. (5)] rather than by finite differencing, in order to avoid numerical inaccuracies near the ISCO where the potential gradient becomes small.

The physical mechanism driving the orbital evolution is illustrated in Fig. 5, which displays the radial dependence of the orbital frequency and energy flux for various spin and magnetic field configurations.

First, we observe a robust frequency enhancement in the strong-field regime. As the particle approaches the ISCO, the magnetic field introduces geometrical confinement terms (scaling as $\sim B^{2}r^{2}$) that effectively steepen the radial potential well. To counteract this additional inward curvature and maintain a stable circular orbit, the test particle requires a larger angular velocity than in the vacuum Kerr case. Consequently, near the ISCO, we consistently observe a frequency blue-shift ($\Omega_{\rm mag}>\Omega_{\rm vac}$) for all spins.

Second, a distinct topological transition appears in the global frequency evolution. As shown in the top panel of Fig. 5, for weak magnetic fields ($B=0$ and $0.05$), the orbital frequency increases monotonically as the radius shrinks from $25m$ to the ISCO. This corresponds to the standard “chirp” behavior where the inspiral naturally leads to higher frequencies. However, for stronger fields ($B=0.10$ and $0.15$), the frequency exhibits a non-monotonic behavior: as the orbit shrinks from $25m$, the frequency initially decreases before eventually turning over and increasing rapidly near the ISCO. This implies that in the far region, the magnetic potential dominates and creates an environment where tighter orbits are actually slower—an “inverse chirp” regime—before the standard gravity-dominated dynamics take over at small radii.

This dynamical behavior translates directly into the radiative output (bottom panel). Since the flux scales as $\dot{E}_{\rm GW}\propto\Omega_{\phi}^{10/3}$, the frequency enhancement near the ISCO acts as a nonlinear amplifier, increasing the energy emission rate by orders of magnitude. This generates a feedback loop: the particle radiates energy more efficiently, causing the orbit to shrink more rapidly, which in turn drives the system deeper into the strong-field region.

We construct a time-domain gravitational-wave strain $h(t)$ using a quadrupole proxy for the dominant $(2,2)$ mode and an optimally oriented observer:

The evolution is terminated when the particle reaches the ISCO $\bigl(r(t)=r_{\rm ISCO}\bigr)$. We align the waveforms at the end of the inspiral (taken as $t=0$).

The resulting waveforms are shown in Fig. 6. For visualization purposes, we use an illustrative mass ratio $\eta=0.1$ to display the frequency evolution within a compact timeframe.

The waveform morphology confirms the dynamical features identified in the frequency analysis. For all spin parameters ($a=0.9,0,-0.9$), when the magnetic field is weak ($B\leq 0.05$), the waveforms exhibit the standard “chirp” character, where the frequency increases monotonically throughout the inspiral. The primary effect of the magnetic field in this regime is a kinematic acceleration, significantly shortening the inspiral duration compared to the vacuum case.

However, at sufficiently high field strengths ($B=0.10$ and $0.15$), a “turnover” phenomenon emerges across all spin configurations. In these cases, the waveform frequency is not monotonic. During the early stage of the inspiral (corresponding to larger radii), the frequency decreases over time, characteristic of an “inverse chirp” driven by the dominant magnetic potential $\Lambda=1+B^{2}r^{2}$. As the particle crosses into the strong-gravity regime near the black hole, the frequency evolution reverses, recovering the standard positive chirp $\dot{\Omega}>0$ before the plunge. This turnover signature—a switch from frequency decrease to increase—uniquely encodes the transition from the asymptotic magnetic environment to the near-horizon gravitational field.

While the frequency turnover discussed in Sec. 4 is a feature of extreme fields, the cumulative phase evolution is sensitive to much weaker, astrophysically realistic magnetic fields. To assess the observability of these effects, we quantify the accumulated dephasing $|\Delta\Phi|$ over a fixed observation window $T_{\rm obs}$ corresponding to the final stage of the inspiral. Specifically, for a typical EMRI system ($\eta=10^{-5}$) detectable by LISA, we define the dephasing accumulated during the last year before the plunge ($t\in[t_{\rm merger}-1\,\text{yr},t_{\rm merger}]$) as:

where $\Phi(t)=\int\Omega(t)dt$ is the phase evolution tracked from a radius $r_{\rm start}$ such that the time to reach the ISCO is exactly one year. This approach ensures that our estimates reflect the actual sensitivity of space-based detectors within a realistic mission lifetime.

The global structure of the accumulated phase difference is visualized in the parameter-space heatmap in Fig. 7, which reveals that the magnetic field strength $B$ acts as the primary driver of the dephasing. The phase difference contours are predominantly vertical, indicating that the magnetic influence largely overshadows the effect of the black hole spin. The color gradient demonstrates a rapid growth in $|\Delta\Phi|$ along the horizontal axis. Even though the instantaneous frequency deviation $|\Omega_{\rm mag}-\Omega_{\rm vac}|$ is minute at the beginning of the observation window (typically $r\sim 20\text{--}30m$), the cumulative phase error integrated over the $\mathcal{O}(10^{5})$ orbital cycles typical of EMRIs leads to a significant total dephasing. The pink dashed line marks the conservative detection threshold of $|\Delta\Phi|\sim 1$ rad. In our one-year observation window, this threshold corresponds to a magnetic field strength of approximately $B\sim 2\text{--}4\times 10^{-4}$. We note that a rigorous distinguishability analysis depends on the signal-to-noise ratio (SNR) $\rho$ of the specific event (typically requiring a mismatch $\mathcal{M}\gtrsim D/2\rho^{2}$, where $D$ is the number of source parameters). However, assuming a typical EMRI source for LISA with $\rho\sim 20$, a phase shift of unity is widely accepted as a robust necessary condition for detectability [Zi:2023qfk]. Our results indicate that environmental fields with $B\gtrsim 10^{-4}$ satisfy this condition and are thus potentially observable.

The observability becomes even more pronounced in the strong-field regime. For the astrophysically realistic upper bound of $B\sim 10^{-2}$ suggested in literature [Wang:2025vsx], the system enters a high-dephasing regime where the accumulated phase difference reaches magnitudes of $10^{3}$ rad. In this regime, the magnetic imprint is so dominant that the dephasing crosses the detectability threshold of 1 rad within a timescale of merely hours. This implies that such strong magnetic environments would be unmistakably distinguishable from vacuum spacetimes almost immediately upon detection, regardless of minor variations in detector noise or source orientation.

Beyond the dominant magnetic scaling, the heatmap also reveals a secondary spin dependence. The contours exhibit a systematic curvature to the right as the spin parameter $a$ increases across the entire parameter space. This morphological feature implies that, for a fixed magnetic field strength, retrograde orbits ($a<0$) accumulate a larger phase difference than their prograde counterparts ($a>0$). Consequently, counter-rotating binaries serve as marginally more sensitive probes of the magnetic environment.