⚛️ general relativity

Bayesian Evidence for Inspiraling Hotspot Motion in the Galactic Center

This paper presents a Bayesian analysis of polarimetric ALMA observations of Sgr A* that provides statistically significant evidence that flaring hotspots follow non-geodesic inspiraling trajectories rather than strictly Keplerian orbits.

Original authors: Pablo Ruales, Maciek Wielgus, Delilah E. A. Gates, Alejandro Cardenas-Avendano

Published 2026-09-30

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Pablo Ruales, Maciek Wielgus, Delilah E. A. Gates, Alejandro Cardenas-Avendano

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

At the very heart of our Milky Way galaxy lies a monster of gravity, a supermassive black hole known as Sagittarius A*. It is the closest such object to Earth, making it a unique laboratory for studying how matter behaves when crushed by extreme gravity. For decades, astronomers have watched this region flicker with sudden bursts of light, known as flares. These flares are thought to be caused by hot, bright clumps of gas, or "hotspots," swirling around the black hole. As these clumps orbit, they trace out paths that can be mapped by measuring the direction of their light waves, a property called polarization. When plotted on a graph, these paths often form distinct loops. For a long time, scientists assumed these hotspots were simply circling the black hole in perfect, stable circles, much like planets orbiting a star, just moving much faster and closer to the edge of the abyss.

A new study by Pablo Ruales and his colleagues challenges this simple picture by looking more closely at the data from a specific flare that occurred on April 11, 2017. Using observations from the Atacama Large Millimeter/submillimeter Array, a powerful telescope in Chile, the team analyzed the changing loops of polarized light. They developed a new way to track the motion of these hotspots, allowing the gas to move not just in circles, but also to spiral inward as it falls toward the black hole. Their analysis reveals that the gas does not stay in a neat, circular orbit. Instead, the data strongly suggests that the hotspot is on a non-circular path, spiraling inward with increasing speed as it gets closer to the event horizon. This finding is significant because it moves beyond the idea of a simple, stable orbit to a more dynamic and violent reality where matter is actively being consumed.

The researchers focused on a specific set of observations that showed two distinct loops in the light's polarization pattern: a large outer loop followed by a smaller inner one. Previous studies had successfully modeled the first, larger loop by assuming the gas was moving in a perfect circle. However, when those same circular models were extended to cover the second, smaller loop, they failed to match the observations. The light from the second loop did not fit the pattern of a stable orbit. To solve this, the team created a computer model that could test different types of motion. They broke the journey of the hotspot into segments, allowing the speed and direction of the gas to change between these segments. They tested models where the gas moved in a single segment, two segments, or three segments, comparing how well each version matched the actual telescope data.

The results were clear: the data strongly favored a model where the hotspot was spiraling inward. When the researchers allowed the gas to have a radial velocity—a speed moving toward the center rather than just around it—the fit improved dramatically. The best model described a journey where the gas started with a mild inward drift and then, as it moved closer to the black hole, began to spiral in much faster. This change in motion happened at a specific moment in time, roughly 280 units of time after the observation began, which corresponds to a point beyond where previous studies had stopped their analysis. The team found that a model with two distinct segments of motion, where the gas speeds up its inward fall, explained the entire sequence of loops far better than any model that insisted on a circular path.

The study also examined the magnetic fields surrounding the black hole, which are crucial for generating the polarized light. The analysis confirmed that the magnetic field in this region is dominated by a vertical component, pointing up and down relative to the disk of gas. This aligns with earlier theories but provides stronger evidence based on the new, more flexible motion model. Furthermore, the researchers found that the color of the light, or its spectral index, remained consistent throughout the event, suggesting that the physical nature of the hot gas did not change drastically even as its motion did. The team ruled out the idea that a single, unchanging circular orbit could explain the full dataset, showing that such a model breaks down as soon as the gas moves beyond the first loop.

While the new model fits the data well, the authors are careful to note that it is a statistical preference based on the available observations, not a final, absolute proof of the exact physical mechanism. The model assumes the gas stays in a flat plane around the black hole, but in reality, the flow might be more complex. Additionally, the study relied on fixed values for the black hole's spin and the angle from which we view it, based on previous measurements. The researchers suggest that future observations with even sharper resolution, such as those planned for the next generation of black hole telescopes, will be needed to see the exact path of these hotspots and to understand the full story of how matter falls into the center of our galaxy. For now, however, the evidence points to a dynamic, spiraling dance of gas that is far more chaotic and compelling than the steady circles of the past.

Technical Summary: Bayesian Evidence for Inspiraling Hotspot Motion in the Galactic Center

Problem Statement
Polarimetric observations of the supermassive black hole Sagittarius A* (Sgr A*) during flaring episodes reveal evolving, loop-like trajectories in the Stokes $Q-U$ plane, known as " $Q-U$ loops." Previous analyses of the 2017 April 11 Atacama Large Millimeter/submillimeter Array (ALMA) observation successfully modeled the primary outer loop using transient hotspots on fixed circular (Keplerian) orbits. However, these models fail to reproduce the subsequent inner loop observed in the same dataset, as the later portion of the data is inconsistent with any strictly periodic trajectory. While prior work tested for deviations from Keplerian motion within a circular-orbit framework, a more general, non-circular kinematic prescription was required to explain the full temporal evolution of the polarized signal, particularly the transition between the primary and inner loops.

Methodology
The authors develop a Bayesian framework to infer the time-dependent kinematics of an emitting hotspot from polarized light curves. The core of the methodology involves:

Generalized Kinematic Model: The emitting region's trajectory is modeled as a continuous, piecewise sequence of $K$ $K$ kinematic segments. Each segment is described by radial ( $\beta_r$ $β_{r}$ ) and azimuthal ( $\beta_\phi$ $β_{ϕ}$ ) velocity parameters relative to a circular geodesic (Keplerian) flow. These parameters interpolate between a generalized Keplerian flow and a zero-angular-momentum free-fall from infinity.
- $\beta_r < 1$ introduces a radial inspiraling component.
- $\beta_\phi < 1$ reduces the angular velocity below the Keplerian value.
Radiative Transfer: The model computes polarized emission in the Kerr spacetime using the adaptive analytical ray-tracing code AART. It incorporates a "slow-light" treatment to account for finite light travel time effects, mapping emission times to observer arrival times. The polarization is calculated via parallel transport of the polarization four-vector using the Penrose-Walker constant, accounting for gravitational lensing and redshift.
Statistical Inference: The analysis employs a Parallel-Tempered Markov Chain Monte Carlo (PT-MCMC) algorithm (using reddemcee) to sample the posterior distribution of model parameters. The parameter vector includes initial position, Keplerianity parameters for each segment, segment duration weights, background polarization offsets, a global sky rotation angle (PA), magnetic field components, and spectral indices.
Model Comparison: Different kinematic profiles are compared using Bayesian evidence (calculated via thermodynamic integration and the stepping-stones method) and the reduced chi-square statistic ( $\chi^2_{\text{eff}}$ ). The study analyzes two data subsets: a shorter window ( $t_{\text{obs}} \sim 230 M$ ) covering only the primary loop, and an extended window ( $t_{\text{obs}} \sim 350 M$ ) including the inner loop.

Key Contributions and Results
The paper presents a systematic comparison of models with $K=1, 2,$ and $3$ kinematic segments against the 2017 ALMA data:

Limitations of Single-Segment Models:
- For the primary loop alone ( $t_{\text{obs}} \sim 230 M$ ), the data show a mild preference for a non-Keplerian inspiral ( $\beta_r < 1$ ) over a strictly Keplerian orbit. However, extending these single-segment models beyond the primary loop results in a marked degradation of the fit quality.
- No single-segment model provides an adequate fit for the entire dataset ( $t_{\text{obs}} \sim 350 M$ ).
Evidence for Multi-Segment Inspiraling Motion:
- When the inner loop is included, models allowing the kinematics to change along the trajectory ( $K=2$ ) are strongly preferred.
- The highest-evidence model ( $E_2$ ) features two segments with free radial and azimuthal parameters ( $\beta_r, \beta_\phi$ ) but a single shared spectral index. This model is favored over the best two-segment Keplerian counterpart by $\Delta \ln Z \approx 46$ .
- The data do not support a three-segment model ( $K=3$ ); the additional complexity is not justified by the evidence, as the posterior collapses toward the two-segment solution.
Inferred Kinematics:
- Segment 1 (Primary Loop): The flow exhibits a mild inspiral ( $\beta_{r,1} \approx 0.95$ ) and is consistent with Keplerian rotation ( $\beta_{\phi,1} \approx 0.96$ ).
- Segment 2 (Inner Loop): Beyond a break time at $t_{\text{break}} \approx 275 M$ (emission time), the flow undergoes a significant kinematic shift. The radial inflow accelerates ( $\beta_{r,2} \approx 0.86$ ), and the rotational support drops sharply ( $\beta_{\phi,2} \approx 0.24$ ), indicating a transition to a sub-Keplerian, advection-dominated flow.
- The trajectory terminates well outside the Innermost Stable Circular Orbit (ISCO).
Physical Parameters:
- The inferred background polarization offsets and sky rotation angle are consistent with previous independent measurements.
- The spectral index ( $\alpha_\nu \approx 0.35$ ) is consistent with values measured immediately following the X-ray flare.
- The magnetic field geometry is inferred to have a dominant vertical component, consistent with prior arguments but derived here without strong priors.

Significance and Claims
The paper claims statistically significant evidence that the emitting region in Sgr A* follows a non-geodesic, inspiraling trajectory rather than a strictly Keplerian orbit. This preference is weak when only the primary loop is considered but strengthens sharply when the full dataset, including the inner loop, is modeled. The results suggest that the hotspot is not merely orbiting but is being carried inward by an advection-dominated accretion flow, losing rotational support as it approaches the event horizon.

The authors position this work as a necessary step in characterizing the Sgr A* system and understanding magnetized accretion dynamics. They note that while their model successfully fits the data, the assumption of piecewise constant kinematics is an approximation; future work should explore smoothly accelerating trajectories and include variability in spin and inclination. The study emphasizes that these improved theoretical models are essential for interpreting future high-angular-resolution observations from the next-generation Event Horizon Telescope (ngEHT) and the Black Hole Explorer (BHEX).

More like this