Recent notes on covariance-weighted consistency tests for Kerr parameter estimates

Posted on May 15, 2026 by Aronno Mirdha

A recurring issue in strong-field tests of General Relativity is the question of how one should compare parameter estimates inferred from genuinely independent observational sectors. In the case of stationary black hole spacetimes, the Kerr hypothesis1In General Relativity, stationary astrophysical black holes are expected to be described completely by the Kerr solution characterized by mass and angular momentum. predicts that all sufficiently accurate observations of a given object should ultimately correspond to the same underlying pair of parameters $(M,a)$. However, the observational sectors used to infer these quantities are physically quite different in character: gravitational-wave measurements probe the dynamical evolution of compact systems, imaging observations probe null geodesic structure near the horizon, while orbital and accretion-based methods constrain yet other aspects of the geometry. Consequently, even before discussing possible deviations from General Relativity, one already encounters a fairly nontrivial statistical problem concerning the compatibility of sectoral parameter estimates.

The framework currently under consideration approaches this problem through a covariance-weighted construction on the combined parameter space. For each observational sector $k$, one introduces a parameter vector

$$
\theta_k =
\begin{pmatrix}
M_k \\
a_k
\end{pmatrix},
$$

together with an associated covariance matrix $\Sigma_k$. The sectoral estimates are then embedded into a stacked parameter vector, and compared against a common best-fit parameter $\bar{\theta}$ representing the null hypothesis of a shared Kerr spacetime. This leads naturally to the quadratic statistic

$$
T^2 = r^T C^{-1} r,
$$

where $r$ denotes the residual vector and $C$ is the covariance matrix associated with the stacked estimator. Geometrically, the statistic may be interpreted as a covariance-weighted squared distance on the residual subspace, closely related to the Mahalanobis distance2Unlike ordinary Euclidean distance, the Mahalanobis distance accounts for covariance structure between parameters. appearing in multivariate statistics.

Under the standard Gaussian approximation for the inferred parameter distributions, the resulting statistic follows an asymptotic $\chi^2$ law with

$$
\nu = (K-1)p
$$

degrees of freedom, where $K$ denotes the number of observational sectors and $p$ the dimension of the parameter vector. This provides a direct statistical interpretation of sectoral disagreement in terms of exceedance probabilities relative to the null hypothesis of a common spacetime geometry.

One feature of the construction that appears conceptually useful is that the covariance matrix functions not merely as a weighting prescription, but effectively induces the geometry on the parameter space itself. In this picture, consistency testing becomes a problem of measuring distances inside a statistically curved space determined by the observational uncertainties. This perspective also clarifies why naive comparisons between overlapping confidence regions can sometimes obscure nontrivial inconsistencies once covariance structure is taken into account systematically.

Preliminary Monte Carlo studies indicate that the statistic reproduces the expected $\chi^2$ behaviour rather accurately under the null hypothesis. Introducing controlled sectoral biases shifts the distribution toward larger values of $T^2$ in a quantitatively stable way, suggesting that the framework is reasonably sensitive to inconsistencies comparable in scale to the observational uncertainties themselves.

Several extensions remain under consideration. The most immediate limitation of the present framework is the Gaussian approximation implicit in the covariance description. In realistic inference problems, posterior structure may become significantly non-elliptic due to parameter degeneracies, low signal-to-noise effects, or modelling assumptions. It therefore seems natural to investigate whether the covariance-based construction can be generalized to formulations involving posterior samples or more explicitly information-geometric approaches on the statistical manifold3Information geometry studies probability distributions using differential-geometric structures induced by statistical inference..

At present the framework should probably be regarded primarily as a structural proposal rather than an observational analysis. Nevertheless, the broader idea of treating consistency of spacetime geometry itself as a quantitative statistical object still appears mathematically and conceptually interesting.

References and Footnotes

  • 1
    In General Relativity, stationary astrophysical black holes are expected to be described completely by the Kerr solution characterized by mass and angular momentum.
  • 2
    Unlike ordinary Euclidean distance, the Mahalanobis distance accounts for covariance structure between parameters.
  • 3
    Information geometry studies probability distributions using differential-geometric structures induced by statistical inference.
Posted in Notes, Research | Tagged , , , , , | Leave a comment

Vega Through 600 Frames of Starlight

Posted on May 15, 2026 by Aronno Mirdha

Tonight I pointed the Seestar S50 toward Vega and let it run for a while. Vega is one of those stars that almost feels too bright to photograph properly. It dominates the frame immediately, and after only a few exposures the sensor is already overwhelmed.

The final image came from roughly 600 separate exposures, each around 10 seconds long, so the total integration time ended up being

$$
T \approx 6000\text{s}.
$$

One thing that becomes noticeable during long integrations is how differently a camera experiences the night sky compared to our eyes. At first the image looks sparse and noisy, but as more frames accumulate, faint stars slowly begin to appear from the background. The process feels less like taking a photograph and more like gradually uncovering structure that was already there.

A useful approximation is that the signal-to-noise ratio improves roughly like

$$
\text{SNR} \propto \sqrt{N},
$$

1This follows from standard Poisson photon statistics in stacked astronomical imaging.
where $N$ is the number of stacked frames. Increasing the exposure time therefore has diminishing returns, but the improvement is still surprisingly noticeable over long runs.

Vega itself remains heavily saturated throughout the integration. The bright halo around the star is not its actual physical size, since Vega’s angular diameter is only about

$$
\theta \approx 3 \text{ milliarcseconds}.
$$

2Approximate interferometric angular diameter of Vega from optical interferometry measurements.
Most of what appears in the center of the image comes from diffraction, atmospheric scattering, and the response of the sensor itself3Atmospheric seeing, diffraction through the optical system, and sensor blooming all contribute to the apparent stellar profile.. In some sense the image is recording not only the star, but also the interaction between starlight and the instrument observing it.

What I found most interesting was the surrounding stellar field. Even under moderately light-polluted skies in Mülheim an der Ruhr, the stack eventually became dense with faint background stars that were almost invisible in the early exposures. There is something oddly satisfying about watching the image slowly converge as more photons arrive.

Vega (90 degrees rotated for thumbnail purpose)

I also tried to keep the processing relatively restrained. Long-exposure astronomy already produces enough interesting structure on its own, and excessive editing tends to hide some of the subtlety that makes these images enjoyable to look at in the first place.

Captured with:

Seestar S50
~600 × 10 second exposures
, Germany

References and Footnotes

  • 1
    This follows from standard Poisson photon statistics in stacked astronomical imaging.
  • 2
    Approximate interferometric angular diameter of Vega from optical interferometry measurements.
  • 3
    Atmospheric seeing, diffraction through the optical system, and sensor blooming all contribute to the apparent stellar profile.
Posted in Astrophotography | Tagged , | 1 Comment

Letter to the crew of Artemis II

Posted on March 31, 2026 by Aronno Mirdha

Dear Artemis II crew,

Good luck on your mission.

What you are about to do is more than a flight. It is a step further into a place that, for most of us, only exists in equations and thought experiments. You are taking something abstract and making it real again.

In physics, we try to understand the universe from first principles. We reduce everything down, space, time, motion, to symbols and logic. But there is always a gap between understanding something and actually going there. You are the bridge across that gap.

When you look back at Earth from that distance, you are not just seeing a planet. You are seeing the entire system we try to describe. Every model, every theory, every calculation ultimately points to that same reality you will witness directly.

Missions like yours matter because they remind us that exploration is not finished. It never is. Each step outward changes how we think, how we ask questions, and what we believe is possible.

Stay sharp, trust each other, and take it all in. You are carrying not just equipment, but the curiosity of millions of people who have looked up and wondered what is out there.

From someone working to understand the universe on paper, to those going out to meet it in person, I wish you the best of luck.

Aronno Mirdha

Posted in Statements | Leave a comment

Inspiral-merger-ringdown consistency tests and the reconstruction of Kerr geometry

Posted on March 2, 2026 by Aronno Mirdha

One of the more conceptually interesting developments in gravitational wave astronomy is the inspiral-merger-ringdown (IMR) consistency test. At a heuristic level, the idea is rather simple: different sectors of a binary black hole coalescence should reconstruct the same final spacetime geometry if General Relativity correctly describes the dynamics. What makes the problem mathematically nontrivial is that the inspiral, merger, and ringdown regimes probe rather different aspects of the Einstein equations and rely on different approximation schemes.

During the inspiral phase, the orbital separation remains sufficiently large that one may approximately treat the system within post-Newtonian theory1Post-Newtonian expansions approximate relativistic dynamics through a series expansion in powers of v/cv/c., where the dynamics are expanded in powers of the characteristic orbital velocity $v/c$. Near merger, however, the gravitational field becomes strongly nonlinear and the perturbative description breaks down entirely, requiring full numerical relativity. Finally, after coalescence, the newly formed black hole relaxes toward equilibrium through damped perturbative oscillations governed by black hole perturbation theory.

The remarkable point is that General Relativity predicts that all three regimes should nevertheless encode a mutually compatible description of the same final Kerr spacetime2The Kerr solution describes rotating black holes in General Relativity and is fully characterized by mass and angular momentum..

Suppose one observes a gravitational waveform $h(t)$ produced by a compact binary coalescence. Schematically, one may decompose the signal into two sectors:

$$
h(t)=h_{\rm insp}(t)+h_{\rm MR}(t),
$$

where $h_{\rm insp}$ denotes the inspiral contribution and $h_{\rm MR}$ denotes the merger-ringdown contribution. In practice this decomposition is performed in frequency space through a cutoff frequency separating the early inspiral regime from the late nonlinear dynamics.

From the inspiral portion alone, one may infer a posterior distribution for the final black hole parameters

$$
(M_f^{\rm insp},a_f^{\rm insp}),
$$

while the merger-ringdown sector independently yields

$$
(M_f^{\rm MR},a_f^{\rm MR}).
$$

The central question of the IMR test is therefore whether these independently reconstructed geometries are statistically compatible.

To formulate this quantitatively, one typically introduces fractional deviation parameters

$$
\Delta M_f
=
\frac{
M_f^{\rm insp}-M_f^{\rm MR}
}{
(M_f^{\rm insp}+M_f^{\rm MR})/2
},
$$

together with

$$
\Delta a_f
=
\frac{
a_f^{\rm insp}-a_f^{\rm MR}
}{
(a_f^{\rm insp}+a_f^{\rm MR})/2
}.
$$

Under the null hypothesis that General Relativity correctly describes the binary coalescence, one expects

$$
\Delta M_f \approx 0,
\qquad
\Delta a_f \approx 0,
$$

up to statistical uncertainty and waveform systematics.

The statistical structure becomes clearer when phrased geometrically. Let

$$
\theta=
\begin{pmatrix}
M_f\\
a_f
\end{pmatrix}
$$

denote the parameter vector associated with the final Kerr geometry. The inspiral and merger-ringdown analyses then generate two posterior distributions on the same parameter space:

$$
p_{\rm insp}(\theta),
\qquad
p_{\rm MR}(\theta).
$$

The IMR test therefore becomes a comparison between two independently inferred probability measures on Kerr parameter space. In the Gaussian approximation, each posterior may be characterized locally by a covariance matrix,

$$
\Sigma_{\rm insp},
\qquad
\Sigma_{\rm MR},
$$

which geometrically define uncertainty ellipses around the corresponding maximum-likelihood estimates.

One may then define a residual vector

$$
r=
\theta_{\rm insp}
–
\theta_{\rm MR},
$$

together with the combined covariance

$$
C=
\Sigma_{\rm insp}
+
\Sigma_{\rm MR}.
$$

The natural quadratic consistency statistic is therefore

$$
T^2
=
r^T C^{-1} r.
$$

Mathematically, this is a Mahalanobis-type distance3Unlike ordinary Euclidean distance, the Mahalanobis distance incorporates covariance structure between parameters. induced by the covariance geometry of the parameter space itself. Under the Gaussian approximation, the statistic asymptotically follows a $\chi^2$ distribution with two degrees of freedom:

$$
T^2 \sim \chi^2_2.
$$

Thus the IMR test may ultimately be interpreted as a covariance-weighted geometric comparison between two independently reconstructed spacetime geometries.

The ringdown sector is particularly interesting because it is governed by quasinormal mode perturbations of the final Kerr black hole. Schematically, the late-time gravitational waveform may be expanded as

$$
h(t)
\sim
\sum_n
A_n e^{-i\omega_n t},
$$

where the frequencies are complex:

$$
\omega_n
=
\omega_{R,n}
–
i\omega_{I,n}.
$$

The real part $\omega_{R,n}$ determines the oscillation frequency while the imaginary part $\omega_{I,n}$ determines the damping rate. Crucially, in General Relativity the quasinormal spectrum depends only on the parameters of the final Kerr geometry:

$$
\omega_n
=
\omega_n(M_f,a_f).
$$

This is essentially a manifestation of the black hole uniqueness structure4Often summarized informally through “no-hair” theorems stating that stationary black holes are characterized by only a small number of parameters.: once the final geometry settles to Kerr, the ringdown spectrum becomes completely determined by the mass and spin of the remnant.

The inspiral sector probes the geometry in a rather different manner. During inspiral, the phase evolution of the waveform depends sensitively on the chirp mass

$$
\mathcal{M}
=
\frac{
(m_1m_2)^{3/5}
}{
(m_1+m_2)^{1/5}
},
$$

together with higher-order spin and post-Newtonian corrections. From these quantities one reconstructs the parameters of the final remnant through numerical-relativity-informed fitting formulae. Thus the inspiral and ringdown sectors infer the same final geometry through entirely different dynamical mechanisms.

From a conceptual point of view, this is perhaps the most remarkable feature of the IMR framework. The inspiral regime involves weak-to-intermediate-field orbital dynamics; the merger regime probes the fully nonlinear Einstein equations; and the ringdown regime reduces to perturbations of a stationary Kerr background. The consistency test therefore compares distinct sectors of General Relativistic dynamics across very different physical and mathematical regimes.

One useful way to summarize the logical structure is:

$$
\text{inspiral dynamics}
\longrightarrow
(M_f,a_f),
$$

$$
\text{ringdown spectrum}
\longrightarrow
(M_f,a_f),
$$

and General Relativity predicts that these independently reconstructed geometries should agree statistically.

Of course, practical implementations are complicated by detector noise, waveform systematics, finite signal-to-noise ratio, calibration uncertainties, and possible correlations between sectors. Nevertheless, the overall mathematical structure of the test remains rather elegant: spacetime geometry is reconstructed independently from different dynamical sectors and then compared through the induced statistical geometry of parameter space itself.

In that sense, the IMR consistency framework is perhaps best viewed not merely as a parameter consistency test, but as a global self-consistency condition on strong-field spacetime dynamics.

References and Footnotes

  • 1
    Post-Newtonian expansions approximate relativistic dynamics through a series expansion in powers of v/cv/c.
  • 2
    The Kerr solution describes rotating black holes in General Relativity and is fully characterized by mass and angular momentum.
  • 3
    Unlike ordinary Euclidean distance, the Mahalanobis distance incorporates covariance structure between parameters.
  • 4
    Often summarized informally through “no-hair” theorems stating that stationary black holes are characterized by only a small number of parameters.
Posted in Expository, Notes | Tagged , , , | Leave a comment

Traversable wormholes and the geometry of effective exoticity

Posted on February 7, 2026 by Aronno Mirdha

One of the useful lessons of general relativity is that the Einstein equations are not, by themselves, especially conservative about the kinds of geometries they permit. Smooth Lorentzian metrics can describe black holes, gravitational waves, expanding cosmologies, singularity formation, and even rather exotic global structures. Traversable wormholes are a particularly clean example of this phenomenon. As a matter of differential geometry, it is not difficult to write down a metric containing a throat joining two regions of spacetime. The difficulty appears only when one asks what stress-energy tensor is required to support such a geometry.

In ordinary Einstein gravity, the answer is severe: a traversable throat requires matter which violates the null energy condition. Roughly speaking, the throat wants to collapse inward, and the only way to keep it open is to supply a sufficiently negative radial pressure. This is the point at which the geometry stops being merely a clever construction and becomes a question about the physical matter content of the theory.

The standard Morris-Thorne wormhole metric is

$$ ds^2 = -e^{2\Phi(r)}dt^2 + \frac{dr^2}{1-b(r)/r} + r^2(d\theta^2+\sin^2\theta\,d\phi^2). $$

There are two functions in this ansatz. The function $\Phi(r)$ is the redshift function, and it controls the gravitational redshift and the possible formation of horizons. The function $b(r)$ is the shape function, and it controls the spatial geometry of the wormhole. The throat is located at a radius $r_0$ satisfying

$$
b(r_0)=r_0.
$$

At this radius the coefficient of $dr^2$ becomes singular. This should not be interpreted too quickly as a physical singularity; it is a coordinate effect associated with the choice of the radial coordinate $r$. The real geometric condition is not merely that $b(r_0)=r_0$, but that the surface actually flares outward at the throat instead of pinching off.

To see this, one considers an equatorial spatial slice by setting

$$
t=\mathrm{const},
\qquad
\theta=\frac{\pi}{2}.
$$

The induced two-dimensional metric is then

$$
ds^2
=
\frac{dr^2}{1-b(r)/r}
+
r^2d\phi^2.
$$

We now embed this two-dimensional surface into Euclidean three-space with cylindrical coordinates $(r,\phi,z)$. The Euclidean metric on a surface of revolution $z=z(r)$ is

$$
ds^2
=
\left(1+\left(\frac{dz}{dr}\right)^2\right)dr^2
+
r^2d\phi^2.
$$

Equating this expression with the wormhole slice gives

$$
1+\left(\frac{dz}{dr}\right)^2
=
\frac{1}{1-b(r)/r}.
$$

Hence

$$
\frac{dz}{dr}
=
\pm
\left(\frac{r}{b(r)}-1\right)^{-1/2}.
$$

At the throat, where $b(r_0)=r_0$, this derivative diverges. Geometrically, the embedded surface becomes vertical there. The more important quantity is the second derivative of $r$ with respect to $z$, since this measures whether the surface opens outward. A short calculation gives

$$
\frac{d^2r}{dz^2}
=
\frac{b(r)-rb'(r)}{2b(r)^2}.
$$

Thus the flare-out condition is

$$
b(r)-rb'(r)>0.
$$

At the throat this becomes

$$
b'(r_0)<1. $$

This is the precise geometric statement that the wormhole opens out rather than closes in. It is a purely spatial condition, but in Einstein gravity it has an immediate dynamical consequence. The field equations convert this geometric inequality into a statement about the stress-energy tensor.

For the Morris-Thorne metric, the energy density and radial pressure in Einstein gravity satisfy, in an orthonormal frame,

$$
\rho(r)=\frac{b'(r)}{8\pi r^2},
$$

$$
p_r(r)
=
\frac{1}{8\pi}
\left[
-\frac{b(r)}{r^3}
+
2\left(1-\frac{b(r)}{r}\right)\frac{\Phi'(r)}{r}
\right].
$$

At the throat, the second term in $p_r$ vanishes because $1-b(r_0)/r_0=0$. Therefore

$$
p_r(r_0)
=
-\frac{1}{8\pi r_0^2}.
$$

Meanwhile

$$
\rho(r_0)
=
\frac{b'(r_0)}{8\pi r_0^2}.
$$

Adding these two expressions gives

$$
\rho(r_0)+p_r(r_0)
=
\frac{b'(r_0)-1}{8\pi r_0^2}.
$$

But the flare-out condition requires $b'(r_0)<1$. Hence

$$
\rho(r_0)+p_r(r_0)<0. $$

This is exactly a violation of the null energy condition in the radial null direction. So the usual slogan that wormholes require “exotic matter” is not just a qualitative statement. It follows directly from the throat geometry together with the Einstein equations.

Modified gravity enters the story by changing this last step. In Einstein gravity, the relation between geometry and matter is

$$
G_{\mu\nu}=8\pi T_{\mu\nu}.
$$

Thus once the metric is fixed, the stress-energy tensor is fixed. In an $f(R)$ theory, however, the gravitational action is no longer the Einstein-Hilbert action

$$
S
=
\frac{1}{2\kappa^2}
\int d^4x\sqrt{-g}\,R,
$$

but instead

$$
S
=
\frac{1}{2\kappa^2}
\int d^4x\sqrt{-g}\,f(R).
$$

The model considered here is

$$
f(R)=R+\alpha R^n.
$$

This looks like a small deformation of general relativity, but it changes the field equations in an essential way. Varying the action with respect to the metric gives

$$
f_R R_{\mu\nu}
–
\frac{1}{2}f(R)g_{\mu\nu}
=
\nabla_\mu\nabla_\nu f_R
–
g_{\mu\nu}\Box f_R
+
\kappa^2T_{\mu\nu},
$$

where

$$
f_R=\frac{df}{dR}.
$$

For the particular model $f(R)=R+\alpha R^n$, one has

$$
f_R
=
1+\alpha nR^{n-1}.
$$

The new feature is the appearance of derivatives of $f_R$, and hence derivatives of the Ricci scalar $R$. In other words, the curvature itself now contributes additional effective stress-energy terms. It is often useful to rewrite the modified field equations schematically as

$$
G_{\mu\nu}
=
8\pi
\left(
T_{\mu\nu}^{\mathrm{matter}}
+
T_{\mu\nu}^{\mathrm{curv}}
\right).
$$

This formula should not be taken as saying that curvature has literally become matter. Rather, it is a bookkeeping device. The higher-curvature terms can be moved to the right-hand side of the equation and treated as an effective source. This changes the wormhole problem in an important way. The throat may still require null energy condition violation in the effective total source, but the violation need not come entirely from the ordinary matter sector.

The paper studies the shape function

$$
b(r)=re^{-(r-r_0)}.
$$

This is a convenient choice because the basic wormhole conditions can be checked explicitly. First,

$$
b(r_0)=r_0e^{-(r_0-r_0)}=r_0,
$$

so $r=r_0$ is indeed the throat. Next,

$$
\frac{b(r)}{r}=e^{-(r-r_0)}.
$$

Thus

$$
\frac{b(r)}{r}\to 0
\qquad
\text{as}
\qquad
r\to\infty.
$$

This gives the expected asymptotic flatness condition in the radial part of the metric. Differentiating the shape function, one obtains

$$
b'(r)
=
e^{-(r-r_0)}(1-r).
$$

Therefore

$$
b(r)-rb'(r)
=
re^{-(r-r_0)}
–
r e^{-(r-r_0)}(1-r)
=
r^2e^{-(r-r_0)}.
$$

Since $r>0$, this quantity is positive:

$$
b(r)-rb'(r)>0.
$$

So the flare-out condition is satisfied. In particular, at the throat one has

$$
b'(r_0)=1-r_0,
$$

and hence $b'(r_0)<1$ for $r_0>0$, as required.

The authors then consider both constant and variable redshift functions. The variable choice

$$
\Phi(r)
=
\ln\left(1+\frac{r_0}{r}\right)
$$

is especially useful because it is finite for all $r\geq r_0$. Since

$$
e^{2\Phi(r)}
=
\left(1+\frac{r_0}{r}\right)^2,
$$

the metric coefficient $g_{tt}$ never vanishes in the wormhole domain. Thus this choice avoids the formation of an event horizon while still producing a nontrivial tidal structure.

Once this metric is substituted into the $f(R)$ field equations, the resulting expressions for $\rho$, $p_r$, and $p_t$ become rather complicated. This is not surprising. The field equations contain not only $R$ but also derivatives of $f_R$, so the stress-energy components involve higher derivatives of the metric functions. The calculation is still systematic: choose $b(r)$ and $\Phi(r)$, compute the curvature scalar $R$, compute $f(R)$ and $f_R$, insert them into the modified field equations, and then read off the effective density and pressures.

The main point is not that the final formulas are elegant. They are not. The main point is that the modified curvature terms contribute to the effective energy budget of the wormhole. In Einstein gravity, the throat condition directly forces

<

p style=”text-align:center;”>
$$
\rho+p_r<0 $$

for the matter supporting the geometry. In $f(R)$ gravity, the analogous inequality applies to the total effective source:

$$
\rho_{\mathrm{eff}}+p_{r,\mathrm{eff}}<0. $$

But this effective quantity contains both ordinary matter and curvature contributions:

$$
\rho_{\mathrm{eff}}+p_{r,\mathrm{eff}}
=
(\rho+p_r)_{\mathrm{matter}}
+
(\rho+p_r)_{\mathrm{curv}}.
$$

Thus it becomes possible, at least in principle, for the curvature sector to carry part of the exoticity. The ordinary matter sector may then satisfy the null energy condition in regions where the total effective source violates it.

This is the conceptual mechanism behind the model. The wormhole does not become free of exoticity in an absolute sense. The flare-out condition still demands an effective violation of the null energy condition. What changes is the location of that violation. In general relativity, it must be attributed directly to matter. In modified gravity, some of it can be absorbed into the higher-curvature terms.

This distinction is subtle but important. It means that the statement “wormholes require exotic matter” is not purely a statement about wormhole geometry. It is a statement about wormhole geometry together with a particular set of gravitational field equations. Change the field equations, and the same geometric requirement may be distributed differently between matter and curvature.

Of course, this should not be read as evidence that traversable wormholes are physically realized. An $f(R)$ model must still satisfy many independent tests: stability of the solutions, absence of ghost-like degrees of freedom, compatibility with solar-system and cosmological observations, and sensible behaviour under perturbations. These are serious constraints, and a formal wormhole solution is only the beginning of the story.

Nevertheless, the example is mathematically instructive. It shows that in gravitational physics the boundary between “geometry” and “matter” is more delicate than it first appears. In Einstein gravity the division is sharp: curvature sits on the left-hand side, matter on the right. In modified gravity, higher-curvature terms blur this separation. The same wormhole throat may therefore be interpreted not as being held open entirely by exotic matter, but partly by the effective stress-energy generated by the gravitational action itself.

That is the real lesson. Modified gravity does not merely enlarge the space of possible solutions. It also changes the bookkeeping of what it means for a geometry to be physically supported.

Posted in Expository | Tagged , , , , , , , | Leave a comment

A Dark Halo That Almost Became a Galaxy

Posted on January 14, 2026 by Aronno Mirdha

One of the cleanest ideas in modern cosmology is also one of the easiest to overlook. According to the standard ΛCDM model, structure in the Universe forms hierarchically, with dark matter collapsing under gravity into bound halos over an enormous range of masses. Massive halos, capable of hosting large galaxies or clusters of galaxies, are comparatively rare. As one moves to lower and lower masses, however, the number of halos increases rapidly. In fact, the theoretical abundance of halos rises so steeply toward small masses that low-mass halos should dominate the cosmic population by sheer numbers. Taken at face value, this has a striking and somewhat unsettling implication. If every dark matter halo were to host a visible galaxy, the night sky would look very different from what we observe. Instead of a relatively sparse distribution of galaxies, we would expect to see an overwhelming number of faint systems, tracing the enormous population of small halos predicted by theory. The absence of such a population in observational surveys is not a minor detail. It is a central clue that tells us something fundamental about how galaxy formation works, and, just as importantly, about how it fails. The natural conclusion is that most dark matter halos do not become galaxies in any ordinary sense of the word. They either never manage to form stars, or they form so few that their stellar populations are effectively invisible with current observational techniques. In this way, the prediction that the Universe should contain far more halos than galaxies is not a problem to be fixed, but a feature to be understood. It points toward a vast, largely unseen population of dark matter structures, quietly shaping the cosmic web without ever lighting up.

This conclusion is not controversial, and it is not new. What makes it interesting is that it forces us to confront an uncomfortable mismatch between theory and observation. We see galaxies, not halos. If the theoretical prediction is right, then most halos must fail to produce anything that looks like a galaxy. They either form no stars at all, or form so few that their light is effectively undetectable. In that sense, the existence of dark halos is not a speculative idea; it is almost required by the success of the model.

The real difficulty is empirical. A halo without stars is, by design, hard to see. Dark matter emits no light, and a system that lacks stars does not glow in the usual optical or infrared bands. If such objects exist, we should not expect them to announce themselves clearly. Instead, they must be found indirectly, through whatever faint traces of ordinary matter they manage to retain. This is where the physics of reionization enters the story. When the Universe was reionized, the intergalactic medium was heated by ultraviolet radiation to temperatures of order ten thousand kelvin. Gas at these temperatures develops significant pressure support, and only sufficiently deep gravitational potential wells can confine it for long periods of time. The result is a kind of mass threshold for galaxy formation. Above this threshold, halos can hold onto gas, allow it to cool, and eventually form stars. Below it, gas is easily lost or remains too warm and diffuse to collapse.

It is useful to summarize this complicated physics with a single number: a critical halo mass $M_{\mathrm{crit}}$. At the present epoch, this scale is around $10^{9.7} M_\odot$. The precise value is not especially important for what follows. What matters is that galaxy formation becomes sharply inefficient near this mass. A small change in halo mass, or in the details of its assembly history, can mean the difference between forming a faint dwarf galaxy and forming no stars at all. This sharp transition opens up an intriguing possibility. Consider halos that lie just below the critical mass. They are not massive enough to form stars today, but they may still be massive enough to retain some of their gas. In such systems, the gas does not collapse into a rotating disk or fragment into stars. Instead, it can settle into a relatively simple configuration, supported by pressure rather than rotation, and held in place by the dark matter potential. In this regime, the behavior of the gas is governed by equilibrium rather than by violent astrophysical processes. The temperature is regulated by a balance between photoheating from the ultraviolet background and radiative cooling. Gravity tries to compress the gas inward, while pressure pushes back. When these effects balance, the gas settles into a quasi-static state, roughly spherical, dynamically cold, and largely free of the complications associated with star formation and feedback. Objects of this type have come to be known as reionization-limited H I clouds. The name is less important than the idea behind it. These systems are not arbitrary oddities, but a natural prediction of the ΛCDM framework combined with reionization physics. They are expected to be rare, because they occupy a narrow window in halo mass, but they are also expected to have distinctive observational signatures, particularly in neutral hydrogen.

From a theoretical point of view, such objects are unusually attractive. Ordinary galaxies are messy. Their gas and stars are shaped by cooling, star formation, feedback, and environmental effects, all of which complicate any attempt to infer the underlying dark matter distribution. A gas-rich but starless halo, by contrast, is closer to a clean physics problem. If one could identify such a system and measure its gas properties, one would have a direct window into the structure of a low-mass dark matter halo, largely uncontaminated by the usual astrophysical uncertainties. The question, then, is not whether such objects are allowed by theory. It is whether any of them can be found in the real Universe. With this theoretical picture in mind, it is natural to ask whether any concrete example of such a system has actually been observed. A particularly intriguing candidate has emerged in the vicinity of the nearby spiral galaxy M94: a compact neutral hydrogen cloud commonly referred to as Cloud-9. The object was first identified in 21 cm emission and subsequently confirmed with higher-resolution radio observations. What immediately sets Cloud-9 apart is that it looks like a coherent, self-contained system in H I, yet it does not obviously resemble a conventional gas-rich dwarf galaxy. One of the key observational facts is kinematic. Cloud-9 has a recession velocity of approximately $v \simeq 304\,\mathrm{km\,s^{-1}}$, essentially identical to that of M94. This makes a chance alignment with a foreground Milky Way high-velocity cloud unlikely, and strongly suggests that Cloud-9 lies at roughly the same distance as M94, about $D \simeq 4.4\,\mathrm{Mpc}$. Interpreted at this distance, Cloud-9 is compact, with an angular extent of order an arcminute, corresponding to a physical scale of roughly a kiloparsec.

The velocity structure of the neutral hydrogen is equally important. The observed 21 cm line profile is narrow, with a reported width of $W_{50} \approx 12\,\mathrm{km\,s^{-1}}$. Such a small line width immediately distinguishes Cloud-9 from most gas-rich dwarf galaxies, which typically show broader profiles due to rotation or turbulence. Here there is no clear evidence for a rotating disk. Instead, the kinematics are consistent with a dynamically cold, pressure-supported gas cloud. From the total integrated 21 cm flux, one can estimate the neutral hydrogen mass using the standard relation
$$
M_{\mathrm{HI}} = 2.36\times 10^5\, D^2 \int S(v)\,dv \, M_\odot,
$$
where $D$ is the distance in megaparsecs and $\int S(v)\,dv$ is the velocity-integrated flux in $\mathrm{Jy\,km\,s^{-1}}$. Substituting the observed values and adopting the distance of M94 yields
$$
M_{\mathrm{HI}} \sim 10^6\, M_\odot.
$$
This places Cloud-9 squarely in the regime of low-mass, gas-rich systems, comparable in H I content to some of the faintest known dwarf galaxies. At this point, one might reasonably wonder whether Cloud-9 could simply be an extreme but otherwise ordinary dwarf galaxy. However, modeling the gas as a pressure-supported system reveals a further puzzle. If the observed neutral hydrogen were the only source of gravity, the cloud would not be able to confine itself. The internal motions implied by the line width would cause the gas to disperse on relatively short timescales. The fact that Cloud-9 appears compact and long-lived implies the presence of an additional gravitational component. Interpreting this missing mass as dark matter leads to a striking inference. Simple equilibrium models, in which the gas sits in hydrostatic balance within a dark matter potential, suggest a total halo mass of order
$$
M_{\mathrm{halo}} \sim 5\times 10^9\, M_\odot.
$$
This value is not arbitrary. It lies remarkably close to the critical mass scale associated with reionization and the suppression of galaxy formation. In other words, Cloud-9 appears to sit precisely where theory predicts the transition between halos that form stars and halos that do not. If this interpretation is correct, then Cloud-9 is not merely an odd cloud of gas. It is a potential example of a dark matter halo whose mass is large enough to retain neutral hydrogen, yet small enough to have largely failed at forming stars. This makes it an unusually direct and concrete realization of the ideas discussed earlier, and immediately raises the most important question of all: if Cloud-9 really inhabits such a halo, where are its stars?

At this point the discussion becomes a detection problem in the literal statistical sense. Let us fix, as a working hypothesis, that Cloud-9 sits at the distance of M94, and consider a putative stellar component with total stellar mass $M_\star$. The observational data consist of a set of detected point sources in a small region on the sky centered near the H I maximum, together with their magnitudes in two bands (so that each source corresponds to a point in a color–magnitude diagram). The question is: for a given $M_\star$, what is the probability that a dataset of this depth would produce no statistically significant stellar overdensity at the Cloud-9 position?

To turn this into something quantitative, one needs three ingredients.
First, a model for the underlying stellar population. Concretely, one chooses an isochrone family and an initial mass function, and thereby obtains a distribution of intrinsic stellar luminosities in the observed filters, conditional on an assumed age and metallicity. In the most conservative case for detection, one takes an old, metal-poor population (for instance, age $\sim 10\,\mathrm{Gyr}$ and ${\rm [Fe/H]}\sim -2$), because younger populations would produce brighter, more easily detected stars for the same $M_\star$.
Second, one needs a model of the observational selection function. This consists of a completeness function $c(m)$ and an error model for the measured magnitudes, both of which can be calibrated by artificial-star injection and recovery tests. One may think of the selection function as defining, for any intrinsic magnitude $m$, a detection probability $c(m)\in[0,1]$ and a conditional distribution for the observed magnitude given that the star is detected.
Third, one needs a background model: even if Cloud-9 has no stars, the chosen sky region will contain some number of contaminating sources (foreground stars, unresolved background galaxies, and substructure within galaxies) that pass the point-source and quality cuts. The key point is that this background is measurable from control regions, so it can be treated as an empirically determined nuisance distribution rather than an arbitrary theoretical prior.

Once these ingredients are in place, the inference can be phrased in a way that is reasonably close to a standard hypothesis test. Fix a spatial aperture $A$ (for example, a circle of radius $r$ centered at the H I peak), and define a test statistic $N$ to be the number of detected sources within $A$ that survive the photometric quality cuts. For the purposes of obtaining a conservative upper limit, it is often enough to work with $N$ rather than with the full two-dimensional CMD distribution, because a genuine stellar population would typically increase $N$ as well as concentrate sources along the expected RGB locus.
Let $N_{\rm obs}$ be the observed value of this statistic in the Cloud-9 aperture. Let $B$ be the random variable representing the background counts in such an aperture, estimated by placing many apertures of the same size in control regions. Finally, let $S(M_\star)$ be the random variable representing the number of detected stars contributed by a stellar population of total mass $M_\star$ after applying completeness and photometric uncertainties. Then, under the hypothesis that Cloud-9 hosts a stellar population of mass $M_\star$, the total detected count is
$$
N(M_\star) \;=\; B \;+\; S(M_\star),
$$
where $B$ and $S(M_\star)$ are (to a good approximation) independent. The object of interest is then the tail probability
$$
p(M_\star) \;=\; \mathbb{P}!\left( N(M_\star) \le N_{\rm obs} \right),
$$
namely the probability that one would observe a count no larger than $N_{\rm obs}$ if the true stellar mass were $M_\star$. If $p(M_\star)$ is very small, then $M_\star$ is inconsistent with the data at high confidence. This is the mathematically cleanest way to state the problem: we are trying to find the largest $M_\star$ for which the observation remains plausible once one accounts for both observational incompleteness and background contamination.

There is a subtlety here that is easy to miss if one thinks only in terms of integrated light. For small stellar masses, the number of luminous tracer stars (such as RGB stars above a given magnitude limit) is a small integer, and therefore highly stochastic. Two stellar systems with the same $M_\star$ can produce noticeably different numbers of detectable bright stars simply because of Poisson and IMF sampling fluctuations. In the notation above, this is precisely the statement that $S(M_\star)$ is not well approximated by its mean. One really must treat $S(M_\star)$ as a full distribution, typically obtained by Monte Carlo sampling of the stellar population followed by application of the selection function. With this probabilistic framing, the “where are the stars?” question becomes sharply posed: determine the range of $M_\star$ for which $p(M_\star)$ remains non-negligible. If even $M_\star \sim 10^4 M_\odot$ yields $p(M_\star)\ll 1$ after properly accounting for background and incompleteness, then Cloud-9 cannot plausibly hide a Leo T–like stellar component. Conversely, if the data only rule out $M_\star \gtrsim 10^5 M_\odot$, then the object could still be an ultra-faint dwarf in disguise. The rest of the analysis is, essentially, an implementation of this inequality with realistic inputs.

Let us now connect the abstract tail probability
$$
p(M_\star)=\mathbb{P}!\left(B+S(M_\star)\le N_{\rm obs}\right)
$$
to what is actually measured in the Cloud-9 field. Fix an aperture $A$ consisting of a circle of radius $r=8.4”$, chosen because it corresponds to the effective radius of a Leo T analog placed at the distance of M94. Within this aperture, one can count the number of detected sources that survive a strict set of photometric quality cuts. Denote by $N_{\rm obs}$ the observed count in the Cloud-9 aperture. The raw observed number is $3$, but the H I centroid has a positional uncertainty comparable to the aperture size, so one should not treat the aperture center as exact. If one shifts the aperture center over the allowed centroid uncertainty and repeats the count, one obtains an empirical distribution of $N_{\rm obs}$ values with mean approximately $3.5$ and a dispersion of about $1$.

Next, one needs a background model. Define $B$ to be the random variable describing the number of contaminating sources per aperture. Rather than postulating a parametric form for $B$, one can estimate it empirically by placing a large number of apertures of identical size on a control region of the same dataset, processed with the same photometric pipeline and quality cuts. Operationally, this yields a background count distribution with mean approximately $3.7$ and dispersion about $2$ per aperture. The point is not the exact numbers, but the fact that the background level is measured directly and is comparable to the on-target count.

The relevant quantity is therefore not $N_{\rm obs}$ by itself, but the excess count
$$
\Delta \equiv N_{\rm obs}-B.
$$
At the Cloud-9 location, using the shifted-aperture procedure for $N_{\rm obs}$ and the control-aperture procedure for $B$, the inferred excess is
$$
\Delta \approx -0.2 \pm 2.2,
$$
which is consistent with $\Delta=0$ and, in particular, provides no evidence for a positive overdensity of point sources at the Cloud-9 position. Interpreted probabilistically, this means that any allowed stellar population must be one whose detectable contribution $S(M_\star)$ is typically of order a few stars or less, and even that only in the high tail of its stochastic distribution.

Now we incorporate the forward model for the stellar population. For each candidate stellar mass $M_\star$, one generates many Monte Carlo realizations of a stellar population (e.g., an old, metal-poor population), converts to the observed bands, and then applies the empirically measured selection function (completeness and photometric scatter) to obtain the induced distribution of the detected-star count $S(M_\star)$. Crucially, because we are in the low-mass regime where the number of bright tracer stars is a small integer, the distribution of $S(M_\star)$ must be treated directly; it is not well described by its mean alone.

With these pieces in hand, the test becomes sharp. Consider the hypothesis $H(M_\star)$ that Cloud-9 hosts a stellar population of mass $M_\star$ within the aperture. Under $H(M_\star)$, the observed count is modeled as $N=B+S(M_\star)$, and one asks whether the realized $N$ is unusually small compared to what $H(M_\star)$ predicts. For a concrete example that is astrophysically meaningful, take $M_\star=10^4\,M_\odot$. Under this hypothesis, the Monte Carlo population synthesis plus selection function yields the following strong statement: in $99.5\%$ of realizations, at least one detectable star is recovered in the aperture. Equivalently,
$$
\mathbb{P}!\left(S(10^4 M_\odot)\ge 1\right)=0.995,
\quad\text{so}\quad
\mathbb{P}!\left(S(10^4 M_\odot)=0\right)=0.005.
$$
If one were in a zero-background world, the conclusion would already be immediate: a non-detection at the Cloud-9 position would exclude $M_\star=10^4 M_\odot$ at $99.5\%$ confidence.

But we are not in a zero-background world, and that is exactly why the excess variable $\Delta$ matters. The background count is not only nonzero, it is comparable to the observed count. The correct question is therefore: can the background fluctuations plausibly mask the additional stars predicted by $M_\star=10^4 M_\odot$? The excess estimate $\Delta=-0.2\pm 2.2$ implies that, even allowing for statistical uncertainty, the maximal plausible positive overdensity in the aperture is only a small integer. If one takes a deliberately conservative upper excursion consistent with this uncertainty, one obtains an upper bound of roughly $\Delta_{\max}\simeq 2$ stars attributable to a real counterpart. Thus, a stringent (and conservative) way to state consistency is
$$
S(M_\star)\le 2,
$$
because any model that would typically produce three or more detectable stars above background would tend to generate a positive excess inconsistent with what is observed.

Under $M_\star=10^4 M_\odot$, the Monte Carlo distribution for $S(M_\star)$ places most probability mass above this conservative threshold. Concretely, only about $8.7\%$ of realizations yield $S(10^4M_\odot)\le 2$, so
$$
\mathbb{P}!\left(S(10^4M_\odot)\le 2\right)\approx 0.087.
$$
This means that even after giving the model the benefit of (i) centroid uncertainty, (ii) empirically measured background fluctuations, and (iii) a conservative tolerance for a small positive excess, a $10^4M_\odot$ stellar population remains strongly disfavored. One can summarize this as an exclusion at approximately the $1-0.087\approx 91.3\%$ level under the most conservative excess allowance, and at the $99.5\%$ level at the nominal center where the effective excess is consistent with zero and even negative.

At this stage it is also important to explain why these choices are conservative rather than aggressive. The assumed stellar population is taken to be old and metal-poor, which minimizes the number of bright, easily detected stars for a given $M_\star$. Any younger or intermediate-age component would increase detectability and therefore strengthen the exclusion. Similarly, the use of strict quality cuts and an empirically calibrated completeness function protects against over-claiming detections, but also means that some genuine stars (if present) would be lost by the pipeline, again making the inferred upper limit conservative. Finally, the background is not modeled by a convenient distribution chosen to yield a strong result; it is measured directly from the same dataset, meaning the comparison is intrinsically like-for-like.

Putting the argument in its cleanest form, the data constrain the stellar mass to be so low that the expected number of detectable RGB stars is at most of order unity. In practice, the forward modeling indicates that the largest stellar mass compatible with producing on average no more than one detectable RGB star, after incompleteness and photometric scatter, is approximately
$$
M_\star \lesssim 10^{3.5}\,M_\odot.
$$
This is far below the stellar masses of canonical gas-rich dwarfs with similar neutral hydrogen masses, and it is precisely the sort of bound one would want if the goal is to distinguish “a faint dwarf galaxy” from “a gas-bearing halo that largely failed to form stars.”

In short, once one phrases the problem as a hypothesis test for $M_\star$ using an empirically calibrated selection function and background distribution, the result is not merely “we did not see a galaxy.” The result is a quantitative inequality: any stellar counterpart must be so small that, in a dataset capable of resolving red giant branch stars at M94’s distance, the induced detectable-star count is forced into the small-integer regime. That is the sense in which Cloud-9 behaves, statistically, like a starless system. oai_citation:0‡2508.20157.pdf

Posted in Expository | Tagged , | Leave a comment

A Consistency Test for Kerr Black Holes via Orbital Motion, Ringdown, and Imaging

Posted on January 11, 2026 by Aronno Mirdha

Kerr Trisector Closure (KTC) is a consistency test for the Kerr hypothesis that tries to stay honest about what is actually being inferred from data. The guiding principle is simple: if the exterior spacetime of an astrophysical, stationary, uncharged black hole is Kerr, then there exist parameters $(M,\chi)$ such that every observable in every channel is generated by the same Kerr geometry with those parameters. KTC is just the exercise of turning that sentence into a clean mathematical statement that you can put into a likelihood analysis.

Start with the Kerr family written abstractly as a two-parameter set of spacetimes

$$
\mathcal{K}={( \mathcal{M}, g_{ab}(M,\chi) ) : M>0,; |\chi|<1},
$$

where $M$ is the mass and $\chi = J/M^2$ is the dimensionless spin (using geometric units $G=c=1$). The only thing we will use about Kerr is that any prediction for any measurement in a given “sector” is a deterministic functional of $(M,\chi)$ once you fix nuisance choices like orientation, distance, inclination, etc.

Now define three observational sectors:

Orbital sector $\mathsf{O}$: timelike dynamics, e.g. orbital frequencies, precessions, inspiral phasing in the adiabatic regime.

Ringdown sector $\mathsf{R}$: quasi-normal mode (QNM) spectrum, i.e. complex frequencies $\omega_{\ell m n}(M,\chi)$ and associated amplitudes/phases.

Imaging sector $\mathsf{I}$: null geodesics and radiative transfer, e.g. shadow size/asymmetry, photon ring structure, closure phases, visibility amplitudes.

For each sector $s \in {\mathsf{O},\mathsf{R},\mathsf{I}}$, let $d_s$ denote the data in that sector. A standard statistical model is: conditional on parameters, the data are distributed according to a likelihood

$$
\mathcal{L}_s(d_s \mid \theta_s, \lambda_s),
$$

where

$$
\theta_s = (M_s,\chi_s)
$$

are the Kerr parameters inferred from that sector, and $\lambda_s$ collects nuisance parameters for that sector (distance, inclination, calibration, environment, waveform systematics, scattering, emissivity model, etc.). The key point is not what is inside $\lambda_s$ but that the likelihood for each sector can be written down, at least in principle, as a function of $(M,\chi)$ plus nuisances.

At this point, you have two different hypotheses you can formalize.
1. The “unconstrained” model says each sector can have its own Kerr parameters:

$$
H_{\text{free}}:\quad \theta_{\mathsf{O}},\theta_{\mathsf{R}},\theta_{\mathsf{I}} \text{ are independent a priori.}
$$

  1. The “closure” model says there is a single Kerr spacetime behind all three:

$$
H_{\text{Kerr}}:\quad \theta_{\mathsf{O}}=\theta_{\mathsf{R}}=\theta_{\mathsf{I}}=\bar\theta,
$$

for some common $\bar\theta = (\bar M,\bar\chi)$.

KTC is the act of comparing these two, or equivalently quantifying how strongly the data prefer a shared $(M,\chi)$ over three separate ones.

To make this rigorous, write the evidence (marginal likelihood) under each model. Under $H_{\text{Kerr}}$, the joint likelihood factorizes across sectors conditional on the shared parameters (this is the usual conditional-independence assumption given the source parameters; if you have shared systematics you can explicitly couple them in the nuisance structure):
$$
\mathcal{L}(d_{\mathsf{O}},d_{\mathsf{R}},d_{\mathsf{I}}\mid \bar\theta,\bar\lambda)
=\prod_{s\in{\mathsf{O},\mathsf{R},\mathsf{I}}} \mathcal{L}s(d_s\mid \bar\theta,\lambda_s),
$$
with $\bar\lambda = (\lambda{\mathsf{O}},\lambda_{\mathsf{R}},\lambda_{\mathsf{I}})$. The evidence is then
$$
Z_{\text{Kerr}}
=\int \left[\prod_{s}\mathcal{L}_s(d_s\mid \bar\theta,\lambda_s)\right],
\pi(\bar\theta),\prod_s \pi(\lambda_s), d\bar\theta, d\lambda_s.
$$

Under $H_{\text{free}}$, you allow independent Kerr parameters per sector:
$$
Z_{\text{free}}
=\int \left[\prod_{s}\mathcal{L}_s(d_s\mid \theta_s,\lambda_s)\right],
\left[\prod_s \pi(\theta_s)\pi(\lambda_s)\right],
\prod_s d\theta_s, d\lambda_s.
$$

A clean closure statistic is the Bayes factor
$$
\mathcal{B}=\frac{Z_{\text{Kerr}}}{Z_{\text{free}}}.
$$
If $\mathcal{B}$ is large, the data prefer the shared-parameter Kerr description. If $\mathcal{B}$ is small, the data prefer letting the sectors drift apart in $(M,\chi)$, which is exactly what “failure of closure” means in a statistically coherent way.

If you prefer a frequentist formulation, you can do essentially the same thing with a constrained versus unconstrained maximum-likelihood comparison. Define the log-likelihoods
$$
\ell_{\text{free}} = \sum_s \max_{\theta_s,\lambda_s} \log \mathcal{L}s(d_s\mid \theta_s,\lambda_s),
$$
$$
\ell{\text{Kerr}} = \max_{\bar\theta,\bar\lambda} \sum_s \log \mathcal{L}s(d_s\mid \bar\theta,\lambda_s).
$$
Then a likelihood-ratio test statistic is
$$
\Lambda = 2(\ell{\text{free}}-\ell_{\text{Kerr}}).
$$
Heuristically, $\Lambda$ measures the “penalty” for forcing the three sectors to share the same $(M,\chi)$. Under regularity conditions and in an asymptotic regime, $\Lambda$ is approximately $\chi^2$ distributed with degrees of freedom equal to the number of constraints, which here is $4$ (two parameters per sector, three sectors gives $6$ parameters, constrained model has $2$). In practice, because the models can be nonlinear and posteriors non-Gaussian, you calibrate $\Lambda$ by simulation.

So far, this is structure and not physics. The physics enters when you specify what each sector is actually measuring, meaning how $(M,\chi)$ shows up in observables.

For the orbital sector, the typical statement is that certain gauge-invariant frequencies (azimuthal, radial, polar) for bound Kerr geodesics are functions of $(M,\chi)$ and constants of motion. A standard representation is
$$
\Omega_i = \Omega_i(M,\chi; p,e,\iota),
$$
where $(p,e,\iota)$ parametrize the orbit (semi-latus rectum, eccentricity, inclination), and $i$ ranges over the fundamental frequencies. Observationally you do not measure $(p,e,\iota)$ directly; they become part of the nuisance structure or dynamical parameterization, but the core point remains: the orbital likelihood has a map from $(M,\chi)$ into predicted timing and phasing data.

For the ringdown sector, the measurable quantities are complex mode frequencies. For a Kerr black hole,
$$
\omega_{\ell m n} = \frac{1}{M}, f_{\ell m n}(\chi),
$$
for some dimensionless functions $f_{\ell m n}$ determined by black hole perturbation theory (Teukolsky equation with appropriate boundary conditions). The $1/M$ scaling is exact because Kerr has no length scale other than $M$ in geometric units, and the dependence on $\chi$ is encoded in the dimensionless eigenvalue problem. The ringdown likelihood is built from comparing measured $(\Re \omega, \Im \omega)$ (and amplitudes) to these predictions.

For the imaging sector, the clean geometric object is the photon region and its projection onto the observer sky. The boundary of the Kerr shadow can be written in terms of critical impact parameters that are functions of $(M,\chi)$ and the observer inclination $i$. One convenient parameterization uses the constants of motion $(\xi,\eta)$ for null geodesics and gives celestial coordinates $(\alpha,\beta)$ on the image plane:
$$
\alpha = -\frac{\xi}{\sin i}, \qquad
\beta = \pm \sqrt{\eta + a^2\cos^2 i – \xi^2 \cot^2 i},
$$
with $a=\chi M$. The critical curve is obtained by selecting $(\xi,\eta)$ corresponding to unstable spherical photon orbits. Again, the details are not the point here; the point is that there is an explicit deterministic forward model from $(M,\chi)$ (plus inclination and emission model nuisances) to interferometric observables.

Once you accept that each sector admits a forward model, the closure logic becomes almost tautological: Kerr predicts that the same $(M,\chi)$ must fit all three forward models simultaneously. KTC is the quantitative version of “simultaneously.”

It is also useful to write the closure condition in a way that looks like something you can plot. From each sector you can compute a posterior for $\theta_s$ after marginalizing nuisance parameters:
$$
p_s(\theta_s\mid d_s)
\propto \int \mathcal{L}s(d_s\mid \theta_s,\lambda_s),\pi(\theta_s),\pi(\lambda_s), d\lambda_s.
$$
Under the closure hypothesis, you want these to be consistent with a single $\bar\theta$. A natural diagnostic is the product posterior (sometimes called posterior pooling under conditional independence)
$$
p{\text{pool}}(\bar\theta \mid d_{\mathsf{O}},d_{\mathsf{R}},d_{\mathsf{I}})
\propto \pi(\bar\theta)\prod_s \frac{p_s(\bar\theta\mid d_s)}{\pi(\bar\theta)},
$$
which simplifies to a product of likelihood contributions if the priors are aligned. In a Gaussian approximation where each sector yields a mean $\mu_s$ and covariance $\Sigma_s$ in the $(M,\chi)$ plane, you can make the closure tension extremely explicit. The pooled estimator has covariance
$$
\Sigma_{\text{pool}}^{-1} = \sum_s \Sigma_s^{-1},
$$
and mean
$$
\mu_{\text{pool}} = \Sigma_{\text{pool}}\left(\sum_s \Sigma_s^{-1}\mu_s\right).
$$
Then a standard quadratic tension statistic is
$$
T = \sum_s (\mu_s-\mu_{\text{pool}})^{\mathsf{T}}\Sigma_s^{-1}(\mu_s-\mu_{\text{pool}}).
$$
If Kerr is correct and the error models are accurate, $T$ should be “reasonable” relative to its expected distribution (again, best calibrated by injection studies). If $T$ is systematically large, you have a closure failure.

What is nice about this formulation is that it cleanly separates two questions that often get mixed up.

First: does each sector individually admit a Kerr fit? This is about whether $\mathcal{L}_s$ has support near some Kerr parameters.

Second: are the Kerr parameters inferred by each sector consistent with each other? This is the closure question. You can pass the first and fail the second, and that second failure is exactly the kind of thing you would expect from a theory that mimics Kerr in one channel but not in another, or from unmodeled environmental/systematic effects that contaminate one sector differently.

So the rigorous derivation of KTC is basically the derivation of a constrained inference problem. You define three sector likelihoods with their own nuisances, you impose the identification $\theta_{\mathsf{O}}=\theta_{\mathsf{R}}=\theta_{\mathsf{I}}$, and you quantify the loss of fit relative to the unconstrained model. Everything else is implementation detail.

If you want one line that captures the whole test, it is this: KTC is the comparison of
$$
H_{\text{Kerr}}:\exists,\bar\theta\text{ such that }\forall s,; d_s \sim \mathcal{L}s(\cdot\mid \bar\theta,\lambda_s)
$$
against
$$
H{\text{free}}:\forall s,\exists,\theta_s\text{ such that } d_s \sim \mathcal{L}_s(\cdot\mid \theta_s,\lambda_s),
$$
with the comparison done via a Bayes factor $\mathcal{B}$ or a likelihood-ratio statistic $\Lambda$.

Posted in Expository, Notes | Tagged , , , , | Comments Off on A Consistency Test for Kerr Black Holes via Orbital Motion, Ringdown, and Imaging

Towards a derivation of the metric tensor in general relativity

Posted on January 6, 2026 by Aronno Mirdha

One of the central tasks in differential geometry is to make precise the notion of length and angle on a smooth manifold. Unlike $\mathbb R^n$, a general manifold comes with no preferred inner product. The metric tensor is not something that “appears” automatically; rather, it is a geometric structure we deliberately introduce. The goal of this post is to derive the metric tensor from first principles in a clean, logically economical way.

We begin with a smooth $n$-dimensional manifold $M$. At each point $p\in M$, we want to measure lengths of infinitesimal displacements through $p$. Infinitesimal displacements are modeled by tangent vectors, so the correct object to define is an inner product on each tangent space $T_pM$.

A precise and coordinate-free definition of tangent vectors is the following. A tangent vector at $p$ is a derivation at $p$: a linear map $X:C^\infty(M)\to\mathbb R$ satisfying the Leibniz rule $X(fg)=f(p)X(g)+g(p)X(f)$. The collection of all such derivations forms a real vector space $T_pM$ of dimension $n$.

Now introduce a coordinate chart $(U,\varphi)$ with $\varphi=(x^1,\dots,x^n)$ and $p\in U$. For each coordinate function, define a derivation by

$$ \left.\frac{\partial}{\partial x^i}\right|_p (f) := \frac{\partial (f\circ\varphi^{-1})}{\partial u^i}\Big|_{\varphi(p)}. $$

These derivations form a basis of $T_pM$. Every tangent vector $X\in T_pM$ can therefore be written uniquely as

$$ X = X^i \left.\frac{\partial}{\partial x^i}\right|_p $$

for real coefficients $X^i$. These coefficients depend on the chosen coordinates, but the vector $X$ itself does not. At this point, nothing like length exists yet. To measure length, we must specify an inner product on each tangent space.

A Riemannian metric on $M$ is a rule that assigns to each point $p\in M$ a bilinear map

$$ g_p : T_pM \times T_pM \to \mathbb R $$

such that:

  1. $g_p$ is symmetric: $g_p(X,Y)=g_p(Y,X)$
  2. $g_p$ is positive definite: $g_p(X,X)>0$ for all $X\neq 0$
  3. The assignment $p\mapsto g_p(X_p,Y_p)$ is smooth whenever $X,Y$ are smooth vector fields

This definition is entirely coordinate-free. The metric is simply a smoothly varying inner product on tangent spaces.

Now fix a coordinate chart $(x^1,\dots,x^n)$. Since $g_p$ is bilinear, it is completely determined by its values on basis vectors. Define

$$ g_{ij}(p) := g_p\!\left(\left.\frac{\partial}{\partial x^i}\right|_p,\left.\frac{\partial}{\partial x^j}\right|_p\right). $$

These functions $g_{ij}$ are smooth, symmetric in $i,j$, and vary from point to point. They are the components of the metric tensor in coordinates.

Given two tangent vectors

$$ X = X^i \frac{\partial}{\partial x^i}, \qquad Y = Y^j \frac{\partial}{\partial x^j}, $$

bilinearity immediately gives

$$ g_p(X,Y) = g_{ij}(p)\, X^i Y^j. $$

This is the familiar coordinate expression of the metric. Importantly, this is not a definition but a representation of the underlying geometric object $g$.

From this formula, the squared length of a tangent vector $X$ is

$$ |X|^2 = g_{ij}(p)\,X^i X^j. $$

Thus, the metric tensor generalizes the Euclidean dot product by allowing the coefficients $g_{ij}$ to vary with position and coordinates.

It is instructive to check how these components transform. If we change coordinates from $x^i$ to $\tilde x^a$, then the basis vectors transform via the chain rule:

$$ \frac{\partial}{\partial \tilde x^a} = \frac{\partial x^i}{\partial \tilde x^a}\frac{\partial}{\partial x^i}. $$

Substituting into the definition of the metric components yields

$$ \tilde g_{ab} = g_{ij}\frac{\partial x^i}{\partial \tilde x^a}\frac{\partial x^j}{\partial \tilde x^b}. $$

This transformation law is exactly what characterizes $g_{ij}$ as the components of a $(0,2)$-tensor field. The metric tensor is therefore not merely a matrix-valued function but a genuine tensorial object on the manifold.

Posted in Notes | Tagged , , , , | Comments Off on Towards a derivation of the metric tensor in general relativity

Overdetermined parameter interference in physics

Posted on December 31, 2025 by Aronno Mirdha

In many areas of physics, a system is described by a small number of fundamental parameters, while the available observations greatly exceed this number. When this occurs, the problem of parameter inference becomes overdetermined. Rather than being a drawback, this redundancy often plays a central role in testing the internal consistency of a physical theory. A simple example arises when a system depends on a parameter vector $\Theta$ with only a few components, while multiple observables depend on $\Theta$ in different ways. Each observable provides an estimate of the same underlying parameters, but with its own uncertainty and systematic effects. In an idealized setting, these estimates should agree up to statistical noise.

More concretely, suppose a theory predicts that several quantities $\mathcal{O}_k$ depend on a common parameter $\Theta$,
$$
\mathcal{O}_k = \mathcal{O}_k(\Theta).
$$
Measurements of the observables then yield a collection of inferred parameter values $\hat{\Theta}_k$. If the theory is correct and the modeling assumptions are adequate, these inferred values should cluster around a single underlying parameter point.

This situation is familiar in classical mechanics, where the mass of an object can be inferred from its response to different forces, or in electromagnetism, where charge can be inferred from both static and dynamical measurements. In such cases, agreement between independent inferences provides confidence that the underlying description is self-consistent.

In more complex settings, overdetermination becomes a diagnostic tool rather than a mere redundancy. Discrepancies between inferred parameters can signal unmodeled effects, underestimated uncertainties, or a breakdown of the theoretical assumptions used to relate observables to parameters. Importantly, this type of test does not require proposing an alternative theory. It only checks whether a single theoretical framework can simultaneously account for multiple manifestations of the same system.

From a statistical perspective, overdetermined inference naturally leads to goodness-of-fit tests. One seeks a single parameter value $\bar{\Theta}$ that best reconciles all measurements, and then asks whether the residual discrepancies are consistent with the stated uncertainties. Failure of such a reconciliation indicates tension between different pieces of data, even if each measurement individually appears reasonable.

In gravitational physics, overdetermination is particularly natural. A spacetime geometry governs a wide range of physical phenomena, including particle motion, wave propagation, and light deflection. If these phenomena are all described by the same metric, then independent observations should converge on the same geometric parameters. The more distinct the physical processes involved, the more stringent the resulting consistency requirement becomes.

The broader lesson is that overdetermination is not merely a technical feature of data analysis. It reflects a structural property of physical theories that describe systems through a small set of fundamental parameters. When many different observables depend on the same parameters, consistency across these observables becomes a powerful and largely model-independent test of the theory itself.

Posted in Expository | Tagged , , | Comments Off on Overdetermined parameter interference in physics

The Kerr Trisector Closure: A project on internal consistency tests of General Relativity

Posted on December 20, 2025 by Aronno Mirdha

General Relativity describes gravity as the curvature of spacetime. Mass and energy determine this curvature, and physical phenomena such as orbital motion, gravitational radiation, and the propagation of light are governed by the resulting geometry. In this framework gravity is not a force but a geometric property of spacetime itself.

Over the past century, General Relativity has been tested in many independent regimes. Planetary motion agrees with relativistic predictions, gravitational waves from compact binary mergers have been observed, and direct images of black hole shadows have been produced. Each of these observations probes a different physical manifestation of spacetime curvature. However, these tests are typically analyzed in isolation. Orbital dynamics, gravitational-wave signals, and black hole imaging are treated as separate probes, even when they concern the same astrophysical object. As a result, current tests do not directly verify whether all observations of a single black hole are mutually consistent with one and the same spacetime geometry.

The aim of this project is to address this missing consistency check. The guiding question is whether independent observations of the same black hole all describe the same spacetime, as predicted by General Relativity.

An uncharged, rotating black hole is described in General Relativity by the Kerr solution. This solution specifies the spacetime geometry outside the black hole and depends on only two physical parameters: the mass $M$ and the dimensionless spin $\chi$. The mass sets the overall curvature scale, while the spin controls the rotation and associated frame-dragging effects. The dimensionless spin is defined by

$$
\chi = \frac{J}{M^2}, \qquad |\chi| \le 1.
$$

A black hole cannot be observed directly. Instead, its spacetime geometry is inferred through its influence on matter, radiation, and light. In this work we focus on three observational sectors, each sensitive to different physical processes but governed by the same underlying Kerr spacetime.

In the dynamical sector, the curvature of spacetime determines the motion of massive bodies. In compact binary systems this motion produces gravitational waves whose phase evolution depends on the mass and spin of the black hole. Analysis of the inspiral signal yields an estimate $(M_{\mathrm{dyn}}, \chi_{\mathrm{dyn}})$.

In the ringdown sector, a perturbed black hole relaxes to equilibrium through a set of damped oscillations known as quasinormal modes. The frequencies and decay times of these modes depend only on the black hole mass and spin, independent of the details of the perturbation. Measurements of the ringdown signal provide a second estimate $(M_{\mathrm{rd}}, \chi_{\mathrm{rd}})$.

In the imaging sector, strong gravitational lensing near the black hole bends light into characteristic structures such as the photon ring and shadow. High-resolution images constrain the spacetime geometry through the size and shape of these features, leading to a third estimate $(M_{\mathrm{img}}, \chi_{\mathrm{img}})$.

The central hypothesis of this project is that if General Relativity correctly describes gravity, then these three independently inferred parameter pairs must be statistically consistent with a single Kerr spacetime. Differences between the measurements are expected due to experimental uncertainty, but they should not exceed what is allowed by those uncertainties. A statistically significant disagreement would indicate that the observations cannot be described by one common spacetime geometry.

To formalize this test, we describe the spacetime by the parameter vector $\Theta = (M, \chi)$. Each observational sector produces an estimate $\hat{\Theta}{\mathrm{dyn}}$, $\hat{\Theta}{\mathrm{rd}}$, and $\hat{\Theta}{\mathrm{img}}$, with corresponding covariance matrices $\Sigma{\mathrm{dyn}}$, $\Sigma_{\mathrm{rd}}$, and $\Sigma_{\mathrm{img}}$ encoding their uncertainties.

If all sectors are consistent, there should exist a common best-fit parameter vector $\bar{\Theta} = (\bar{M}, \bar{\chi})$ that minimizes the weighted discrepancy

$$
\chi^2(\Theta) = \sum_k
(\hat{\Theta}_k – \Theta)^{\mathrm{T}} \Sigma_k^{-1} (\hat{\Theta}_k – \Theta),
\qquad
k \in {\mathrm{dyn}, \mathrm{rd}, \mathrm{img}}.
$$

The minimizing value is given by

$$

\bar{\Theta}

\left(\sum_k \Sigma_k^{-1}\right)^{-1}
\left(\sum_k \Sigma_k^{-1} \hat{\Theta}_k\right).
$$

We then define the deviation of each sector from the common spacetime as $\delta \Theta_k = \hat{\Theta}_k – \bar{\Theta}$. The Kerr Trisector Closure statistic is

$$

T^2

\sum_k
\delta \Theta_k^{\mathrm{T}} \Sigma_k^{-1} \delta \Theta_k.
$$

This quantity measures the total inconsistency between the three sector estimates while accounting for their uncertainties. Since three independent measurements of two parameters are compared, the statistic follows a $\chi^2$ distribution with four degrees of freedom under the null hypothesis of consistency.

To validate the method, we test it using simulated data. We choose a true spacetime $\Theta^\ast = (M^\ast, \chi^\ast)$ and generate synthetic measurements $\hat{\Theta}_k = \Theta^\ast + \varepsilon_k$, where the errors $\varepsilon_k$ are drawn from Gaussian distributions with covariance $\Sigma_k$. In the consistent case, the resulting $T^2$ values follow the expected $\chi^2_4$ distribution. In Monte Carlo simulations, the mean value and rejection rate match theoretical expectations.

We then introduce a controlled bias in one sector, for example $\hat{\Theta}{\mathrm{img}} = \Theta^\ast + \Delta + \varepsilon{\mathrm{img}}$, while leaving the other sectors unchanged. In this case the $T^2$ distribution shifts to larger values and exceeds the consistency threshold with high probability. The individual contributions

$$
T_k^2 = \delta \Theta_k^{\mathrm{T}} \Sigma_k^{-1} \delta \Theta_k
$$

identify the sector responsible for the inconsistency.

These tests show that the Kerr Trisector Closure behaves as intended. It remains statistically quiet when all observations are consistent and responds strongly when a genuine inconsistency is present. The method provides a model-independent way to test the internal consistency of black hole spacetime measurements. Current observations do not yet allow a full experimental implementation, but future gravitational-wave detectors and improved black hole imaging may make such tests possible.

Posted in Expository, Projects | Tagged , , , | Comments Off on The Kerr Trisector Closure: A project on internal consistency tests of General Relativity