Possible fractal nature of accretion flows in MAD and SANE simulations: Implications to GRS 1915+105

This study employs nonlinear timeseries analysis, specifically Higuchi fractal dimension, Hurst index, and spectral slope, on GRMHD simulations of MAD and SANE accretion flows to demonstrate that MADs exhibit higher fractal dimensions and distinct spin-dependent behaviors, a finding that successfully segregates observational data of GRS 1915+105 into corresponding MAD- and SANE-like clusters.

The Gist

A black hole is silent and invisible — but the matter swirling around it is not. It glows, flickers, and sometimes shoots out powerful jets of radiation. This paper is a detective story trying to figure out the "personality" of that matter by listening to the rhythm of its chaos.

Here is the story of the research, broken down into simple concepts and everyday analogies.

The Two Types of Cosmic Whirlpools

When matter falls towards a black hole, it does not plunge straight in. Instead, it piles up into a swirling, flattened ring called an ACCRETION DISK — a bit like water circling a drain before it finally falls through. These accretion disks are where all the action happens: gas spirals inward at huge speeds, gets heated to millions of degrees, and releases enormous amounts of light and energy. Sometimes the disk even launches powerful jets of matter and radiation out into space at nearly the speed of light.

This paper is not really about black holes themselves (which, being black holes, are silent and invisible) — it's about the TURBULENT DANCE of gas in the accretion disks around them. And specifically, it's a detective story trying to figure out the "personality" of these disks by listening to the rhythm of their chaos.

Astronomers have long known that accretion disks come in two main "flavors" depending on how strong the magnetic fields threading through them are:

1. The "MAD" (Magnetically Arrested Disk): Think of this as a tightly wound spring. The magnetic fields around the disk are so strong and organized that they act like a cage, holding the food back until it suddenly snaps and shoots out powerful, focused jets of energy. It's like a pressure cooker that builds up steam and then releases it in a violent, rhythmic burst.
2. The "SANE" (Standard and Normal Evolution): This is more like a gentle, swirling river. The magnetic fields are weak and messy. The matter flows in more smoothly, driven by turbulence (like white water rapids) rather than a magnetic cage. It's less explosive and more of a steady, churning flow.

The Problem: Listening to the Chaos

For a long time, scientists looked at the light from these accretion disks and tried to measure their "noise" using standard math (like measuring the average height of waves). But the behavior of these disks is non-linear—it's chaotic, unpredictable, and full of hidden patterns. Standard math misses the "fractal" nature of this chaos (the idea that the pattern looks similar whether you zoom in or out).

The authors of this paper asked: Can we use advanced "fractal math" to tell the difference between a MAD disk and a SANE disk just by listening to their rhythm?

The Tools: The "Fractal Ruler," the "Memory Test," and the "Timescale Map"

To solve this, the team used three special tools to analyze the "time-series" (the record of how the brightness changes over time):

1. Higuchi Fractal Dimension (HFD): Imagine you are walking along a jagged coastline.
   • If the coast is a straight line, it's simple (low dimension).
   • If it's a crinkly, complex coastline, it's complex (high dimension).
   • HFD measures how "jagged" or complex the accretion disk's light curve is. A higher number means the pattern is more chaotic and intricate.
2. Hurst Index (H): This measures memory.
   • Does the accretion disk "remember" what it did a second ago?
   • If it has high memory (High H), the flow is smooth and predictable (like a river).
   • If it has low memory (Low H), the flow is random and forgetful (like static on a radio).
3. Spectral Slope: This looks at how the brightness variations are distributed across different timescales — from fast flickers (milliseconds) to slow changes (many seconds). If you plot this distribution, you get a slope:
   • A STEEP slope means slow, smooth changes dominate — the disk has a strong rhythm at longer timescales.
   • A FLAT slope means fast and slow variations are roughly equal — the disk looks more like random "static," chaotic at all scales.
   • Think of it like the difference between a slow, rolling ocean swell (steep slope, dominated by low frequencies) versus the crackling noise of radio static (flat slope, all frequencies equally represented).

The Experiment: Simulating the Cosmos

The researchers didn't just look at real accretion disks immediately; they built virtual ACCRETION DISKS using two different simulation codes (HARMPI and BHAC). These codes solve the equations of general-relativistic magnetohydrodynamics — basically the physics of hot, magnetized gas flowing near a black hole — to produce detailed time-series of how the disk behaves. They simulated both MAD and SANE types with different spins (some spinning with the flow, some against it).

The Findings from the Simulation:

• MAD systems (The Pressure Cooker): They showed High HFD (very jagged, complex patterns) and Low H (short memory).
  • Analogy: Imagine a drum solo that is fast, erratic, and full of sudden, unpredictable hits. It's complex and doesn't follow a simple repeating loop.
• SANE systems (The River): They showed Lower HFD (smoother patterns) and Higher H (longer memory).
  • Analogy: Imagine a steady drum beat. It has a rhythm you can predict a few steps ahead. It's less chaotic.
• The Spectral Slope: MAD systems have FLATTER slopes (chaotic at all timescales, like static), while SANE systems have STEEPER slopes (rhythm dominated by slow, smooth changes, like an ocean swell). This reinforces the finding that MAD is the chaotic pressure cooker and SANE is the steady river.
• The Spin Factor:
  • When the black hole spins faster, the MAD system gets even more organized (the jets become tighter), making the chaos slightly less "jagged" (HFD goes down).
  • But for SANE, spinning faster actually makes it more chaotic because the weak magnetic fields get tossed around more by the wind and jets (HFD goes up).

The Real-World Test: GRS 1915+105

The team then turned to a real accretion disk around a black hole called GRS 1915+105. This object is famous for changing its behavior constantly, switching between 12 different "moods" or classes.

1. They took the light data from these 12 classes.
2. They cleaned the data (removed the "static" noise).
3. They used a computer algorithm to group the classes into two clusters based on their "spectral fingerprint" (what the light looks like).
   • Cluster 1 (MAD-like): These had bright, power-law spectra (like the pressure cooker).
   • Cluster 2 (SANE-like): These had strong thermal disk emissions (like the river).
4. The Result: When they measured the HFD (complexity) of these real clusters, Cluster 1 (MAD-like) had a higher complexity score than Cluster 2 (SANE-like).

This perfectly matched their computer simulations!

The Big Picture: Why This Matters

This paper is a breakthrough because it proves that non-linear math (fractal analysis) is a powerful new tool for astronomers.

• Before: We had to guess if an accretion disk was "MAD" or "SANE" by looking at complex spectral models.
• Now: We can look at the "rhythm" of the light and say, "Ah, this high level of complexity and lack of memory tells us this accretion disk is likely in a magnetically arrested (MAD) state."

In simple terms: The authors found that the "chaos" of an accretion disk isn't random noise; it has a specific shape. By measuring the "jaggedness" of that shape, we can tell if the disk is being controlled by a strong magnetic cage (MAD) or just flowing naturally (SANE). It's like being able to tell if a storm is a hurricane or a thunderstorm just by listening to the sound of the wind, without even seeing the clouds.

Technical Summary

1. Problem Statement

General Relativistic Magnetohydrodynamic (GRMHD) simulations are the standard tool for studying accretion disks and jets around black holes. While these simulations reveal complex, intrinsically nonlinear behaviors, their time-series data have historically been analyzed using linear statistical measures (e.g., Power Spectral Density, auto-correlation). These linear methods fail to capture long-range memory, scale-invariant complexity, and deterministic chaos.

There is a lack of systematic nonlinear comparison between the two primary accretion regimes:

• Magnetically Arrested Disk (MAD): Characterized by strong, ordered magnetic fields near the horizon and powerful jets.
• Standard and Normal Evolution (SANE): Characterized by weaker, dynamically subdominant magnetic fields driven by Magnetorotational Instability (MRI), typically producing weaker jets or winds.

The authors aim to characterize the fractal nature and temporal correlations of these flows using nonlinear time-series analysis and validate these findings against observational data from the black hole binary GRS 1915+105.

2. Methodology

A. Numerical Simulations

The study utilizes two distinct GRMHD codes to ensure robustness:

1. HARMPI: A flux-conservative, high-resolution shock-capturing code. It provides lower temporal resolution (timestep $10 r_g/c$) but is well-documented for convergence.
2. BHAC: An Adaptive Mesh Refinement (AMR) code. It provides higher temporal resolution (timestep $0.1 r_g/c$) and captures finer details.

Setup:

• Geometry: 2.5D simulations (axisymmetric in $\phi$, evolving in $r$ and $\theta$).
• Initial Conditions: Fishbone-Moncrief torus with specific magnetic vector potentials defining MAD and SANE states.
• Parameters: Black hole spins ($a$) ranging from $-0.9375$ (retrograde) to $+0.9375$ (prograde).
• Analysis Window: The final 5,000 timesteps ($25,000 - 30,000 r_g/c$) to ensure stationarity.
• Proxies:
  • Jet Dynamics: Outflow efficiency ($\eta$).
  • Disk Dynamics: Normalized magnetic flux ($\phi$) and accretion rate ($\dot{M}$).

B. Nonlinear Analysis Techniques

The authors employ three primary diagnostic tools:

1. Higuchi Fractal Dimension (HFD): Estimates the geometric complexity and self-similarity of the time series directly without phase-space reconstruction. It is robust for short, non-stationary signals.
   • Range: 1 (linear) to 2 (highly complex/Gaussian).
2. Hurst Index (H): Quantifies long-range temporal correlations.
   • $H > 0.5$: Persistent (correlated).
   • $H < 0.5$: Anti-persistent.
   • $H = 0.5$: Uncorrelated.
   • Theoretical Relation: $HFD + H \approx 2$ for 1D time series.
3. Spectral Slope: Derived from the Power Spectral Density (PSD) using the FOOOF (Fitting Oscillations and One Over f) toolbox to quantify the power-law fall-off.

Validation Steps:

• Surrogate Analysis: Using Iterated Amplitude Adjusted Fourier Transform (IAAFT) to generate 39 surrogate time series. A $z$-score $> 2$ confirms the presence of nonlinearity.
• Filtering: For observational data, a 4th-order Chebyshev Type I bandpass filter is applied to remove high-frequency noise while preserving dynamical bandwidth.

3. Key Contributions

1. First Systematic Nonlinear Comparison: The paper provides the first comprehensive comparison of MAD and SANE flows using fractal dimension and Hurst indices, moving beyond traditional linear spectral analysis.
2. Code Validation: Results are cross-verified using two different simulation codes (HARMPI and BHAC) with different resolutions and numerical schemes.
3. Observational Bridge: The study successfully maps simulation results to the 12 temporal classes of the black hole X-ray binary GRS 1915+105, clustering them into "MAD-like" and "SANE-like" groups based on spectral properties.
4. Methodological Refinement: Demonstrates the efficacy of HFD and Generalized Hurst Exponent (GHE) methods for astrophysical time series, which are often short and non-stationary, overcoming limitations of traditional correlation dimension methods.

4. Key Results

A. Simulation Findings (HARMPI & BHAC)

• Fractal Dimension (HFD):
  • MAD systems consistently exhibit higher HFD (greater complexity/irregularity) than SANE systems.
  • Spin Dependence:
    • In MAD, HFD decreases as spin magnitude increases. This is attributed to the formation of highly collimated, ordered jets (Blandford-Znajek process) which suppress stochasticity.
    • In SANE, HFD increases with positive spin due to the chaotic interplay between winds and jets, leading to higher stochasticity.
• Hurst Index (H):
  • MAD shows lower H (weaker long-range correlations).
  • SANE shows higher H (stronger long-range correlations).
  • The sum $HFD + H \approx 2$ holds for most cases, validating the calculations.
• Spectral Slope: MAD flows have flatter spectral slopes (closer to white noise/stochastic), while SANE flows exhibit steeper power-law slopes.
• Nonlinearity: Surrogate analysis confirms that most MAD and SANE time series are nonlinear ($z$-score $> 2$), except for specific high-spin SANE cases where variability becomes more linear.

B. Observational Implications (GRS 1915+105)

• Clustering: Using k-means clustering on spectral properties (Power-law vs. Diskbb components), the 12 temporal classes of GRS 1915+105 were split into two groups:
  • Cluster 1 (MAD-like): Dominated by Power-law (PL) emission (geometrically thick, magnetically dominated).
  • Cluster 2 (SANE-like): Significant contribution from multi-color blackbody (diskbb) emission (geometrically thinner, Keplerian-like).
• HFD Correlation:
  • The MAD-like cluster has a significantly higher mean HFD ($1.651 \pm 0.066$) compared to the SANE-like cluster ($1.466 \pm 0.034$).
  • This observation perfectly corroborates the simulation results.
• Spectral Correlation:
  • Strong positive correlation between HFD and Power-law component ($r = 0.78$).
  • Strong negative correlation between HFD and diskbb component ($r = -0.8$).
  • This implies that states with intermittent, magnetic-flux-driven eruptions (MAD) are more complex and less correlated, while states with thermal, viscous disks (SANE) are smoother and more correlated.

5. Significance and Conclusion

• Physical Insight: The study establishes that the magnetic field configuration is the primary driver of temporal complexity in accretion flows. Strong, ordered fields (MAD) lead to intermittent, high-complexity eruptions (high HFD), while weak fields (SANE) allow for smoother, more correlated viscous flows (low HFD).
• Diagnostic Tool: Nonlinear time-series analysis (specifically HFD) serves as a powerful, complementary tool to traditional spectral fitting for classifying black hole accretion states. It can distinguish between MAD and SANE regimes even when spectral data is ambiguous.
• Future Applications: The methodology provides a pathway to interpret data from current and future missions (e.g., EHT, X-ray timing missions) by linking horizon-scale variability to large-scale jet and disk dynamics.
• Limitations: The simulations are 2.5D (axisymmetric), which may suppress 3D turbulent mixing. Future work should extend this to full 3D GRMHD simulations and include multi-wavelength data (radio/IR) to trace jet emission directly.

In summary, the paper successfully demonstrates that MAD accretion flows are inherently more complex and less temporally correlated than SANE flows, a finding that holds true in both high-resolution simulations and real-world observations of GRS 1915+105.