From reflection to scattering: Polarimetric signatures of funnel-type outflows. Modeling obscured ultraluminous X-ray sources

arXiv:2605.29487 · astro-ph.HE · Submitted 2026-05-28 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Astrophysics Radio. Generated commentary on the latest astrophysics papers.

Vera: We're looking at the paper: "From reflection to scattering".

Jocelyn: Here's the gist of the paper. Super-Eddington accretion onto compact objects is expected to produce optically thick outflows with a funnel-shaped cavity that may collimate the emission.

Vera: The detection of strong X-ray polarization in the Galactic X-ray binary Cygnus X-3 provides the first direct probe of this geometry.

Jocelyn: This work presents a systematic study of the inclination-dependent radiative signatures of such systems using a combination of semi-analytical methods and Monte Carlo simulations.

Vera: The results demonstrate that both the polarization degree and the degree of collimation are highly sensitive to the albedo of the funnel surface.

Title and authors: Vera: So, we're looking at this paper today, "From reflection to scattering: Polarimetric signatures of funnel-type outflows. Modeling obscured ultraluminous X-ray sources," and the authors are modeling how these outflows shape the light we see from super-Eddington accretion onto compact objects.

Jocelyn: It’s interesting because it's taking something we usually just see as a bright X-ray source, like Cygnus X-three and trying to use its polarization to map out the actual geometry of that outflow <ref:2605.29487#pg1>.

Subrahmanyan: What I find compelling is how they connect this geometric structure—this funnel shape—to the observed spectral state changes in Cyg X-three that we see with IXPE data <ref:2605.29487#pg1>.

Vera: Exactly. The paper sets up this model where you have a compact object surrounded by an outflow characterized as a double conical frustum with parameters R and alpha, and it divides space into three regions: the outflow itself, the inside of the funnel cavity up to radius R, and then everything beyond that.

Jocelyn: They treat the incident emission as an isotropic point source of luminosity L at the center of this funnel geometry, which is a nice way to start simplifying things for our simulations.

Subrahmanyan: That setup allows them to use analytical approximations and then test them with Monte Carlo simulations to see how the polarization degree and collimation change based on the albedo of that funnel surface.

Vera: And they present two main analytical models: one single-scattering model which tries to explain the hard state polarization of Cyg X-three by looking at reflection from that optically thick funnel surface, and another model for the lower polarization degree in intermediate and ultrasoft states using single scattering from a medium above the funnel cavity <ref:2605.29487#pg1>.

Jocelyn: That’s where it gets really interesting because they then use Monte Carlo simulations to test the impact of multiple scatterings, secondary scatterings from a medium inside or above the funnel, and multiple reflections across that funnel cavity.

Subrahmanyan: They use simplifications like assuming a dipole scattering phase function and keeping everything monochromatic for those simulations, which is how you make the math manageable while still testing complex physics.

Vera: The results on the polarization signatures show that for a fiducial value, the polarization degree of single-scattered emission matches their analytical model, but reflected fluxes diverge at low inclinations.

Jocelyn: They found that the polarization degree has two local maxima when light is reflected at an angle of about ninety degrees, which gives us some concrete things to look for in our data <ref:2605.29487#pg1>.

Title and authors: Subrahmanyan: And they also noted that at high inclinations, the polarization degree actually gets slightly reduced in simulations that account for multiple scatterings because more photons manage to escape along the surface normal of the funnel.

Vera: The paper then talks about physical scenarios for those spectral state transitions, suggesting that an optically thin medium above the funnel can qualitatively explain all of the spectral states observed in Cyg X-three by IXPE <ref:2605.29487#pg1>.

Jocelyn: They find that this optically thin medium's optical depth above the funnel, we call it tau surr, is vanishingly small during the hard state but becomes substantial during intermediate and ultrasoft states.

Subrahmanyan: That suggests that this extra scattering component might be what’s driving those changes in flux distribution and spectrum as the system transitions between states.

Vera: The geometric beaming produced by reflection from the funnel surface was relatively weak in their simulations, even when they used a purely reflecting funnel geometry.

Jocelyn: They attribute this weakness to photons escaping through the outer boundary of the outflow and some light exiting the cavity at high angles with respect to the funnel axis.

Subrahmanyan: The paper concludes that for geometric beaming to actually be effective, you need absorption in the upper regions of the funnel surface, meaning it must be strongly absorbing for that scattered component to dominate.

Vera: So, what’s the main conclusion here? The emission reflected by the funnel surface is strongly polarized only when absorption suppresses multiple scatterings, which means most of the flux is absorbed.

Jocelyn: Conversely, they argue that the funnel reflection can only really collimate that flux when there's very little absorption present in those regions.

Subrahmanyan: They point out that the presence of this optically thin medium above the funnel helps explain increased flux during intermediate and ultrasoft states because it scatters some otherwise obscured emission toward us.

Vera: This scattered component is consistent with an increase in X-ray flux during both intermediate and ultrasoft states, which is a key piece of evidence.

Jocelyn: It can also explain why the polarization degree stays relatively constant across those states because this medium leads to lower scattering angles than the funnel reflection we saw during the hard state.

Subrahmanyan: And they suggest that for this scattered component to match the observed polarization degree of about ten percent in those states, you might need a scattering medium optical depth around tau surr equal to about zero point five.

Title and authors: Vera: So, if you look at the overall picture from "From reflection to scattering: Polarimetric signatures of funnel-type outflows. Modeling obscured ultraluminous X-ray sources," the main implication is that we need to consider both the reflection and this above-the-funnel scattering component when interpreting these systems.

Jocelyn: And they're also suggesting that geometric beaming from reflection itself isn't as efficient as we initially hoped, because light leaks out through the outer edges of the outflow boundary.

Subrahmanyan: That’s a fair point; it means any beaming effect we see is likely less extreme than a fully collimated funnel would predict.

Vera: They also mentioned that the optical depth of this scattering medium should disappear when the system returns to its hard state, and it gets ejected during major radio flares.

Jocelyn: And they hinted that during a flaring state, relativistic aberration might cause the polarization degree to initially increase if that scattering medium has a large outflowing velocity.

Subrahmanyan: That would require energy-dependent simulations to really test that idea, which is where the future work comes in.

Vera: So, to wrap up on this paper on "From reflection to scattering: Polarimetric signatures of funnel-type outflows. Modeling obscured ultraluminous X-ray sources," it shows us how absorption and scattering together dictate the polarization we measure from these complex outflows.

Jocelyn: And it provides a framework for interpreting why Cyg X-three changes its spectral state, linking that change to the changing properties of this surrounding medium above the funnel <ref:2605.29487#pg1>.

Subrahmanyan: It connects the microscopic physics of photon scattering right up to how we observe macroscopic spectral shifts across different accretion regimes.

Vera: It really forces us to think about how much absorption versus scattering is happening at different radii within these super-Eddington flows, which is a lot to take in when you're dealing with X-ray data.

Jocelyn: So, for the listener who just wants the short version, this paper says that if you want to understand the polarization of an obscured object like Cyg X-three you can’t just look at reflection alone; you have to account for what's happening above it too <ref:2605.29487#pg1>.

Subrahmanyan: It pushes us toward needing those energy-dependent simulations they mentioned because that might be the only way to truly nail down if relativistic beaming is a factor during flares.

Vera: Right, so that’s the picture on this paper, moving from just reflection to including scattering for these funnel systems.

The paper's summary: Vera: So, to recap this paper on "From reflection to scattering: Polarimetric signatures of funnel-type outflows," they're really showing how you can’t just look at the light bouncing off that funnel surface by itself when you're trying to understand these super-bright X-ray sources.

Jocelyn: Right, it says that the polarization degree and how much the light is collimated really depends on whether we just consider a single reflection or if we have this extra scattering happening in the medium above the funnel.

Subrahmanyan: They model this geometry as a double conical frustum with parameters R and alpha, which sets up three regions: the outflow itself, inside that funnel cavity up to radius R, and then everything beyond that.

Vera: And they use Monte Carlo simulations to test this out because they want to see how sensitive the polarization degree is to things like the albedo of that funnel surface under more realistic radiative transfer conditions.

Jocelyn: The results show that for a single-scattered emission, the polarization degree matches their analytical model, but then those reflected fluxes start diverging at low inclinations.

Subrahmanyan: They also found that at high inclinations, the polarization degree gets slightly reduced in simulations that account for multiple scatterings because more photons manage to escape along the surface normal of the funnel.

Vera: The paper suggests that this optically thin medium above the funnel can explain all of the spectral states we see in Cygnus X-three by IXPE data.

Jocelyn: They find that this medium's optical depth above the funnel, which they call tau surr, is practically zero during the hard state but it gets substantial when we are in intermediate and ultrasoft states.

Subrahmanyan: That means this extra scattering component might be what’s driving those changes in flux distribution and spectrum as the system transitions between those different accretion states.

Vera: They say that geometric beaming from reflection off the funnel surface is actually quite weak in their simulations, even if you use a purely reflecting funnel geometry.

Jocelyn: They attribute this weakness to photons escaping through the outer boundary of the outflow and some light exiting the cavity at high angles with respect to the funnel axis.

Subrahmanyan: The paper concludes that for geometric beaming to actually be effective, you need strong absorption in those upper regions of the funnel surface so that this scattered component can dominate.

Vera: So, what they're saying is that if you want to understand the polarization of an obscured object like Cygnus X-three you have to account for both the reflection and this extra scattering happening above it.

Jocelyn: And they argue that geometric beaming from reflection itself isn't as efficient as we thought because light leaks out through the outer edges of the outflow boundary.

Subrahmanyan: They also pointed out that this scattered component can explain why the polarization degree stays relatively constant across those spectral states because this medium leads to lower scattering angles than the funnel reflection during the hard state.

Vera: And to get that polarization degree down to about ten percent in those intermediate and ultrasoft states, they suggest you might need a scattering medium optical depth around tau surr equal to about zero point five.

Jocelyn: That’s a big number for the optical depth, which means the medium needs to be quite dense above the funnel cavity.

Subrahmanyan: And they also noted that this optically thin medium should vanish when the system returns to its hard state and gets ejected during major radio flares.

Vera: So, it seems like we need a more complete picture where absorption and scattering work together to dictate the polarization we measure from these complex outflows.

Jocelyn: And they're also hinting that if there's a large outflowing velocity, relativistic aberration might cause the polarization degree to initially increase during flaring states.

Subrahmanyan: That would require energy-dependent simulations to truly test that idea, and that’s where the next step lies.

The paper's improvements: Vera: So, moving on from what we just discussed, we have to talk about how the authors are trying to make this model better by adding more physics into their setup for these obscured ultraluminous X-ray sources.

Jocelyn: Right, they’re not just sticking with single scattering or single reflection approximations anymore; they’re explicitly accounting for multiple scatterings in their models.

Subrahmanyan: That should help predict how the polarization degree and the collimation scale respond to the albedo of that funnel surface under more realistic radiative transfer conditions.

Vera: They are using this improved system to create a general framework for interpreting X-ray polarimetric signatures by combining semi-analytical methods with direct Monte Carlo simulations.

Jocelyn: That gives us a comprehensive tool for testing different physical scenarios, like the ones they propose for spectral state transitions in Cygnus X-three, by looking at what happens when you include multiple scatterings.

Subrahmanyan: It lets them examine the impact of those multiple scatterings on the overall picture of how these systems evolve as they change states.

Vera: The AI system can be enhanced to propose physical scenarios for spectral state transitions by modeling how that optically thin medium above the funnel cavity influences both polarization degrees and flux distributions across different states.

Jocelyn: That moves us past just seeing similar polarization values and starts explaining their continuity through a common scattering component.

Subrahmanyan: It connects the microscopic physics of photon scattering right up to how we see macroscopic spectral shifts across different accretion regimes.

Vera: So, what this means for us is that we can use these simulations to test if the presence of this medium above the funnel truly drives those observed changes in X-ray flux and spectrum.

Jocelyn: It’s a step up from just observing the polarization degree; now we can model how that degree behaves as the system moves between its hard, intermediate, and ultrasoft states.

Subrahmanyan: We need to see if this model holds up when we look at other systems, because if it does, it gives us a strong prediction for what other obscured sources should look like.

Vera: But the authors also flag that they still have limitations; they haven't fully modeled every possible physical process, which means the results are best used as a starting point.

Jocelyn: They’re acknowledging that while this is an improvement, a truly comprehensive model would need to include even more complex interactions.

Subrahmanyan: The next big thing for us would be energy-dependent simulations, because they hint that relativistic aberration might cause the polarization degree to initially increase during flaring states if that scattering medium has a large outflowing velocity.

Conclusion: Vera: So, to wrap up this paper on "From reflection to scattering: Polarimetric signatures of funnel-type outflows," they've shown that we need to account for both reflection and scattering when we try to understand the polarization of these super-bright X-ray sources.

Jocelyn: Basically, the main thing is that the polarization degree depends heavily on whether you just look at one reflection or if you factor in this extra scattering happening above the funnel cavity.

Subrahmanyan: They’re giving us a way to test spectral state transitions in Cygnus X-three by seeing how this extra medium above the funnel influences both polarization and flux distributions as the system changes states.

Vera: And they've moved from just looking at what reflection does to modeling how multiple scatterings actually work in these complex outflows.

Jocelyn: They're pushing us to consider that even if geometric beaming from reflection is weak, the overall polarization we measure depends on this scattering component too.

Subrahmanyan: It connects the microscopic physics of photon scattering right up to how we see macroscopic spectral shifts across different accretion regimes in these massive systems.

Vera: The main implication is that for any obscured object like Cygnus X-three, you can’t just rely on reflection alone to get the full story of its polarization.

Jocelyn: And they’re suggesting that this framework helps us interpret why those spectral states look different to a distant observer.

Subrahmanyan: If we can confirm these models in other systems, it gives us a strong prediction for what we should expect when looking at new data from missions like IXPE.

Vera: They did flag that while the model is better, it still doesn't cover every possible physical process out there, so they are open to further refinement.

Jocelyn: And they pointed out that testing this for spectral transitions really needs those energy-dependent simulations we talked about earlier.

Subrahmanyan: That’s the next step; if we can nail down that relativistic beaming effect during flares with better models, it could confirm a lot about the dynamics of these outflows.

Varpu Ahlberg, Alexandra Veledina, Eugene Churazov, Ildar Khabibullin

Department of Physics and Astronomy, University of Turku · Nordita, KTH Royal Institute of Technology and Stockholm University · Max Planck Institute for Astrophysics · Space Research Institute (IKI) · Rudolf Peierls Centre for Theoretical Physics, Department of Physics, University of Oxford

astro-ph.HE

Submitted: 2026-05-28

Updated: 2026-07-31

Comments: 13 pages, 10 figures. Accepted for publication in A&A

Journal ref: A&A 713, A216 (2026)

DOI: 10.1051/0004-6361/202660328

License: http://creativecommons.org/licenses/by/4.0/

Importance score: 84/100

The gist: Super-Eddington accretion onto compact objects is expected to produce optically thick outflows with a funnel-shaped cavity that may collimate the emission.

Key concepts

Funnel Geometry
The model describes a compact object surrounded by an obscuring outflow shaped like a double conical frustum. This geometry is divided into three regions: the outflow itself, the inside of the funnel cavity up to radius R, and the surrounding area outside R. The incident light originates from a central point.
Polarization Degree (PD)
The PD measures how polarized X-ray light is. The study found that for single scattering, PD matches analytical models, but reflected fluxes diverge at low inclinations. High PD suggests the visible part of the funnel surface is only partially ionized.
Scattering Medium Above Funnel
An optically thin medium situated above the funnel cavity provides a physical explanation for all observed spectral states in Cyg X-3. This medium scatters some otherwise obscured emission toward the observer, explaining flux increases in intermediate and ultrasoft states and softening of the spectrum.
Geometric Beaming Limitation
Geometric beaming from funnel reflection was found to be relatively weak because photons escape through the outer boundary of the outflow. The effect is diminished even when absorption is high, suggesting that for collimation to occur, most flux must be absorbed.

Terminology

Summary

Super-Eddington accretion onto compact objects is expected to produce optically thick outflows with a funnel-shaped cavity that may collimate the emission. p. 0

Where the paper says it

fi 2 Nordita, KTH Royal Institute of Technology and Stockholm University, Hannes Alfvéns väg 12, SE-10691 Stockholm, Sweden 3 Max Planck Institute for Astrophysics, Karl-Schwarzschild-Str. 1, D-85741 Garching, Germany 4 Space Research Institute (IKI), Profsoyuznaya 84/32, Moscow 117997, Russia 5 Rudolf Peierls Centre for Theoretical Physics, Department of Physics, University of Oxford, Clarendon Laboratory, Parks Rd, Oxford, OX1 3PU, United Kingdom Received XXXX / Accepted XXXX ABSTRACT Super-Eddington accretion onto compact objects is expected to produce optically thick outflows with a funnel-shaped cavity that may collimate the emission.

The detection of strong X-ray polarization in the Galactic X-ray binary Cygnus X-3 provides the first direct probe of this geometry. p. 0
Where the paper says it

Accordingly, the observed emission is dominated by scattered and reflected radiation, which can therefore be strongly polarized. The detection of strong X-ray polarization in the Galactic X-ray binary Cygnus X-3 provides the first direct probe of this geometry.

This work presents a systematic study of the inclination-dependent radiative signatures of such systems using a combination of semi-analytical methods and Monte Carlo simulations. p. 0
Where the paper says it

The detection of strong X-ray polarization in the Galactic X-ray binary Cygnus X-3 provides the first direct probe of this geometry. In this work, we present a systematic study of the inclination-dependent radiative signatures of such systems using a combination of semi-analytical methods and Monte Carlo simulations.

The results demonstrate that both the polarization degree and the degree of collimation are highly sensitive to the albedo of the funnel surface. p. 0
Where the paper says it

In this work, we present a systematic study of the inclination-dependent radiative signatures of such systems using a combination of semi-analytical methods and Monte Carlo simulations. Our treatment explicitly accounts for multiple scatterings and demonstrates that both the polarization degree and the degree of collimation are highly sensitive to the albedo of the funnel surface.

Modeling Geometry

The models consider a compact object surrounded by an obscuring outflow, characterized as a double conical frustum with parameters R and α. The geometry is divided into three regions: the outflow itself, the inside of the funnel cavity up to a radius of R, and the surrounding area at radii beyond R. p. 1

Where the paper says it

This geometry is illustrated in Fig. We divided the outflow and its surroundings into three regions: the outflow itself, the inside of the funnel cavity up to a radius of R, and the surrounding area at radii beyond R.

The incident emission is treated as an isotropic and unpolarized point source of luminosity L located at the center of the funnel. p. 1
Where the paper says it

As a consequence, the light is allowed to pass through the hole at the center of the cavity. We treated the source of the incident emission as an isotropic and unpolarized point source of luminosity L located at the center of the funnel.

Analytical Approximations

Two analytical models were presented:

  1. A single-scattering model explaining the hard state polarization of Cyg X-3 with reflection from the optically thick funnel surface. p. 2

    Where the paper says it

    We assumed that the three regions have uniform electron number densities, noutfl, ncav, and nsurr, which we parameterized by their optical depths: the outflow has a horizontal Thomson scattering optical depth of τoutfl = σscnoutfl(R − 1), and the inner and outer regions have vertical Thomson scattering optical depths of τcav = σscncav(R − 1) and τsurr = σscnsurrRout, respectively (see Fig. Analytical models An analytical single-scattering model explaining the hard state polarization of Cyg X-3 with reflection from the optically thick funnel surface was presented in Veledina et al.

  2. A model for the lower PD in intermediate and ultrasoft states using single scattering from an optically thin medium above the funnel cavity. p. 2

    Where the paper says it

    In contrast, the lower PD in the intermediate and ultrasoft states was analytically modeled in Veledina et al. (2024b) with single scattering from an optically thin medium above the funnel cavity.

Monte Carlo Simulations

MC simulations were employed to test the impact of multiple scatterings, secondary scatterings from a medium inside or above the funnel, and multiple reflections across the funnel cavity. The simulations used simplifications such as using the dipole scattering phase function and keeping the simulations monochromatic. p. 3

Where the paper says it

In particular, it has been employed to simulate the reflection off optically thick disk funnel walls and optically thin disk wind in SS 433 (Medvedev et al. To make the comparison between the analytical and MC results straightforward, we introduced the following simplifications: (i) we used the dipole scattering phase function, (ii) we kept the simulations monochromatic, i.

Results on Polarization Signatures

The study found that the PD of single-scattered emission for a fiducial value is identical to the analytical model, but reflected fluxes diverge at low inclinations. The PD has two local maxima occurring when light reflected at an angle of ∼90◦ is seen. p. 3

Where the paper says it

The τoutfl = 9 simulation shows no reduction of the flux compared to the analytical model, as the optical depth spanned by each integration step is smaller. In both the analytical and MC models, the PD has two local maxima that occur when light reflected at an angle of ∼90◦ is seen.

At high inclinations, the PD is slightly reduced in simulations that accounted for multiple scatterings because more photons escape along the surface normal of the funnel. p. 3
Where the paper says it

At high inclinations, however, the PD is slightly reduced in the simulations that accounted for multiple scatterings. This occurs because more photons escape along the surface normal of the funnel.

Physical Scenarios for State Transitions

The presence of an optically thin medium above the funnel can qualitatively explain all spectral states in Cyg X-3 observed by IXPE. The optical depth of the medium above the funnel, τsurr, is found to be vanishingly small during the hard state but becomes substantial during intermediate and ultrasoft states. p. 1

Where the paper says it

IXPE observations of Cyg X-3 covered its intermediate and ultrasoft states, which both show a PD of roughly 10% with the same orientation as the hard state (Veledina et al. Substantial changes in polarization properties suggest a change in the accretion structure, yet the persistently high and energyindependent PD indicates that the funnel-type geometry is continually present.

Conclusion on Geometric Beaming

Geometric beaming produced by reflection from the funnel surface was relatively weak in simulations, even for a purely reflecting funnel. This weakness is attributed to photons escaping through the outer boundary of the outflow and some light exiting the cavity at high angles with respect to the funnel axis. p. 9

Where the paper says it

Additionally, light transmitted through the outer boundary of the outflow can significantly reduce the beaming factor from the ideal geometric estimate. Furthermore, a fraction of the light exiting the cavity escapes toward the sides.

The high PD observed in Cyg X-3 indicates that the visible portion of the funnel is only partly ionized. p. 9
Where the paper says it

However, photoelectric absorption is difficult to quantify without self-consistent modeling. The high PD observed in Cyg X-3 implies that the upper region of the funnel surface is only partly ionized and relatively cool (∼105 K), which can be consistent with the analytical estimate of T ∝ r −1/2 (Abolmasov et al.

Summary of Findings

The emission reflected by the funnel surface is strongly polarized only when absorption suppresses multiple scatterings, meaning most flux is absorbed. Conversely, the funnel reflection can only collimate the flux when little absorption is present. p. 0

Where the paper says it

Conversely, a high-albedo medium (nearly pure scattering) can modestly collimate radiation, but at the cost of substantially reducing the polarization degree. We discuss our results in the context of Imaging X-ray Polarimetry Explorer observations of Cygnus X-3 and propose a physical scenario for its spectral state transitions, considering a combination of reflection from the funnel surface and scattering by a diffuse medium above the funnel.

The presence of the optically thin medium above the funnel explains increased flux during intermediate and ultrasoft states because it scatters some otherwise obscured emission toward the observer. p. 6
Where the paper says it

Drastic changes in the polarization signatures in the intermediate and ultrasoft states (PD ∼10%) were previously explained by the presence of an optically thin medium inside or above the funnel cavity (Veledina et al. Our simulations generally confirm this behavior; however, we found it only necessary to include gas above the funnel directly visible to the distant observer, as the gas inside the cavity had a much weaker influence on the PD (Figs.

Implications for State Changes

The addition of a scattered component is consistent with increased X-ray flux during both intermediate and ultrasoft states. This medium can explain the energy-independent PD observed during both states, as it leads to lower scattering angles than the funnel reflection observed during the hard state. p. 8

Where the paper says it

The addition of a scattered component would be consistent with the increased Xray flux observed during both states. Furthermore, the medium can explain the energy-independent PD observed during both states; during the hard state, the PD is reduced near 6 keV due to dilution by the iron fluorescent line (Veledina et al.

The softening of the spectrum in Cyg X-3 is attributed to this medium redirecting some of the concealed soft emission toward the observer. p. 8
Where the paper says it

As noted by Veledina et al. (2024b), the softening of the spectrum in the hard-to-intermediate state transition in Cyg X-3 can be attributed to the medium above the funnel redirecting some of the concealed soft emission toward the observer.

Future Directions

A more detailed model is needed to comprehensively explain geometric beaming in ULXs. The paper predicts that the PD during the flaring state may initially increase due to relativistic aberration if the scattering medium has a large outflowing velocity. p. 8

Where the paper says it

Since the scattering medium above the funnel should no longer be present when the system returns to the hard state, we expect that the medium is ejected during the major radio flares that follow the ultrasoft state. Therefore, we predict that the PD during the flaring state may initially increase as a result of relativistic aberration if the scattering medium has a large outflowing velocity.

Testing this scenario for spectral transitions would require energy-dependent simulations. p. 8
Where the paper says it

Later during the outburst, the PD should settle toward its hard-state value. We emphasize that this is a speculative prediction, and that testing this scenario for the spectral transitions would require energy-dependent simulations.

Other Sources

Similar objects may be observed with IXPE and future missions, and high-inclination systems are expected to have high PD if the reflected radiation is not diluted by emission leaked through the funnel. The lack of such correlations in other systems suggests more complex scattering geometry. p. 9

Where the paper says it

Broadly speaking, if the observed PD will positively correlate with the relative contribution of the reflected continuum to the spectrum, one can suggest a picture similar to the one considered here, where the reflected PD is high and changes gently, while the reflected flux, FI,refl/F0, is small (10−5 − 10−4 ) but changes significantly with changing inclination. The lack of such correlations or inconsistent values of PA or PD will be indicators of more complex scattering geometry, i.

Summary of State Changes

The observed PD then depends on the relative contribution of the scattering component and the reflection component. If the scattered component dominates in both intermediate and ultrasoft states, it could explain comparable PDs in these states. p. 6

Where the paper says it

The observed PD then depends on the relative contribution of this scattering component and the reflection component (the light reflected from the funnel surface). If the scattered component dominates in both intermediate and ultrasoft states, it could explain the comparable PD in these states.

The medium above the funnel can also explain the reduced PD observed during these states because it leads to lower scattering angles than the funnel reflection observed during the hard state. p. 8
Where the paper says it

The addition of a scattered component would be consistent with the increased Xray flux observed during both states. Furthermore, the medium can explain the energy-independent PD observed during both states; during the hard state, the PD is reduced near 6 keV due to dilution by the iron fluorescent line (Veledina et al.

The optical depth of the scattering medium may need to be rather large, around τsurr ∼ 0. p. 3
Where the paper says it

(2024a) when only single-scattered photons are included and the outflow optical depth is very large. Since we modeled the polarization rather than the full spectrum, the only relevant energy-dependent feature of the radiative transfer problem is the single-scattering albedo.

p. 3
Where the paper says it

(2024a) when only single-scattered photons are included and the outflow optical depth is very large. Since we modeled the polarization rather than the full spectrum, the only relevant energy-dependent feature of the radiative transfer problem is the single-scattering albedo.

p. 3
Where the paper says it

(2024a) when only single-scattered photons are included and the outflow optical depth is very large. Since we modeled the polarization rather than the full spectrum, the only relevant energy-dependent feature of the radiative transfer problem is the single-scattering albedo.

5, to match the observed PD of ∼10% in these states. p. 1
Where the paper says it

The Galactic XRB Cyg X-3, long suspected to be strongly obscured, became the first source of its kind to be observed with IXPE. The system is known to continuously switch between spectral states, characterized by distinct X-ray and radio properties (Szostek et al.

p. 1
Where the paper says it

The Galactic XRB Cyg X-3, long suspected to be strongly obscured, became the first source of its kind to be observed with IXPE. The system is known to continuously switch between spectral states, characterized by distinct X-ray and radio properties (Szostek et al.

p. 1
Where the paper says it

The Galactic XRB Cyg X-3, long suspected to be strongly obscured, became the first source of its kind to be observed with IXPE. The system is known to continuously switch between spectral states, characterized by distinct X-ray and radio properties (Szostek et al.

Mechanism for Geometric Beaming

Geometric beaming does not occur via reflection unless the albedo is very close to unity, favoring re-emission of absorbed radiation as the mechanism behind collimation. The high PD observed in Cyg X-3 implies that the upper region of the funnel surface is only partly ionized. p. 9

Where the paper says it

However, photoelectric absorption is difficult to quantify without self-consistent modeling. The high PD observed in Cyg X-3 implies that the upper region of the funnel surface is only partly ionized and relatively cool (∼105 K), which can be consistent with the analytical estimate of T ∝ r −1/2 (Abolmasov et al.

A lower albedo may be possible in the upper section of the funnel surface if it is less ionized. p. 9
Where the paper says it

It is therefore likely that parts of the funnel are purely reflecting. A lower albedo may be possible in the upper section of the funnel surface if it is less ionized.

Summary of Albedo Effects

The majority of emission is absorbed unless the albedo is λ > 0.9 because most photons are absorbed unless the albedo is λ > 0. p. 4

Where the paper says it

= (1 − λ) −1 . Therefore, the majority of the emission is absorbed unless the albedo is λ > 0.

9. The PD in the MC simulations is more notably reduced compared to the analytical model at high inclinations when lambda = 1. p. 4
Where the paper says it

0 (orange and magenta lines). The PD in the MC simulations is more notably reduced compared to the analytical model at high inclinations, as some of the emission propagates through the outflow without being absorbed.

0.

Summary of Scattering Medium Effects

The presence of the optically thin medium above the funnel can qualitatively explain all spectral states in Cyg X-3. p. 8

Where the paper says it

Physical scenario for the state changes in Cyg X-3 The strikingly similar polarization in the intermediate and ultrasoft states of Cyg X-3 implies a clear continuity between the two states. Given that both states feature a PD consistent with single-scattering by an optically thin medium above the funnel, it seems that the presence of the scattering medium may be key in explaining the spectral transitions.

The medium inside the cavity had a much weaker influence on the PD. p. 6
Where the paper says it

Drastic changes in the polarization signatures in the intermediate and ultrasoft states (PD ∼10%) were previously explained by the presence of an optically thin medium inside or above the funnel cavity (Veledina et al. Our simulations generally confirm this behavior; however, we found it only necessary to include gas above the funnel directly visible to the distant observer, as the gas inside the cavity had a much weaker influence on the PD (Figs.

The PD of emission scattered in the medium above the funnel is set primarily by observer inclination and is not very sensitive to other parameters. p. 6
Where the paper says it

5, 7 and 8). In the single-scattering limit, the PD of emission scattered in the medium above the funnel (the scattering component) is set primarily by the observer inclination and is not very sensitive to other parameters.

The observed PD then depends on the relative contribution of this scattering component and the reflection component. p. 6
Where the paper says it

This suggests that its polarization should be preserved across spectral states. The observed PD then depends on the relative contribution of this scattering component and the reflection component (the light reflected from the funnel surface).

p. 6
Where the paper says it

This suggests that its polarization should be preserved across spectral states. The observed PD then depends on the relative contribution of this scattering component and the reflection component (the light reflected from the funnel surface).

Summary of Flux Boost Mechanisms

The flux boost in low-albedo simulations was significantly smaller than that of fully opaque funnel reflection because a large fraction of photons escape through the outer boundary of the outflow. The model geometry allowed for significant leakage through the outflow, so the flux boost in our λ = 1. p. 9

Where the paper says it

Furthermore, a fraction of the light exiting the cavity escapes toward the sides. Our model geometry allowed for significant leakage through the outflow, so the flux boost in our λ = 1.

0 simulations was significantly smaller than that of the fully opaque funnel reflection modeled in Dauser et al. p. 9
Where the paper says it

Our model geometry allowed for significant leakage through the outflow, so the flux boost in our λ = 1. 0 simulations was significantly smaller than that of the fully opaque funnel reflection modeled in Dauser et al.

(2017). The upper regions of the funnel surface can have a lower blackbody temperature that produces little X-ray emission. p. 7
Where the paper says it

1), with τsurr = 5 9 τcav tal flux. However, the upper regions of the funnel surface can have a lower blackbody temperature that produces little X-ray emission, while regions closer to the funnel base can be very hot (Abolmasov et al.

Summary of Spectral State Explanations

The hard and intermediate states may have similar intrinsic spectra, but the intermediate state would appear softer to a distant observer due to light scattered by the medium above the funnel. The ultrasoft state cannot be explained unless the source has a softer intrinsic spectrum. p. 8

Where the paper says it

Since the reflection spectrum nevertheless remains prominent in the intermediate state, the obscured regions of the funnel surface may feature significant reflection. If the spectrum in both the intermediate and ultrasoft states is formed by scattering the obscured emission toward the observer, it would indicate that the ultrasoft state has a softer intrinsic spectrum.

p. 8
Where the paper says it

Since the reflection spectrum nevertheless remains prominent in the intermediate state, the obscured regions of the funnel surface may feature significant reflection. If the spectrum in both the intermediate and ultrasoft states is formed by scattering the obscured emission toward the observer, it would indicate that the ultrasoft state has a softer intrinsic spectrum.

The increased mass-loss rate of the WR wind would naturally explain the increased optical depth of gas above the funnel cavity. p. 8
Where the paper says it

The softening of the spectrum and the quenching of the jet during the intermediate-to-ultrasoft state transition are reminiscent of the intermediate-to-soft state transition in regular black hole XRBs, which is associated with an increase in the accretion rate (Fender et al. An increased mass-loss rate of the WR wind would naturally explain the increased optical depth of gas above the funnel cavity.

The accretion timescale for a standard geometrically thin disk is comparable to the length of the transition if αv is on the order of 10−2–10−3. p. 8
Where the paper says it

1, the timescale can be comparable to the length of the transition if αv is on the order of 10−2–10−3 . The radiative general relativistic magnetohydrodynamical (RGRMHD) simulations of superEddington accretion in Fragile et al.

The medium above the funnel should no longer be present when the system returns to the hard state, and it is ejected during major radio flares. p. 8
Where the paper says it

The timescale of the transition may otherwise be related to the duration of the enhanced mass-loss rate or the interaction between the WR wind and the disk outflow. Since the scattering medium above the funnel should no longer be present when the system returns to the hard state, we expect that the medium is ejected during the major radio flares that follow the ultrasoft state.

The PD during flaring state may initially increase as a result of relativistic aberration if the scattering medium has a large outflowing velocity. p. 8
Where the paper says it

Since the scattering medium above the funnel should no longer be present when the system returns to the hard state, we expect that the medium is ejected during the major radio flares that follow the ultrasoft state. Therefore, we predict that the PD during the flaring state may initially increase as a result of relativistic aberration if the scattering medium has a large outflowing velocity.

p. 8
Where the paper says it

Since the scattering medium above the funnel should no longer be present when the system returns to the hard state, we expect that the medium is ejected during the major radio flares that follow the ultrasoft state. Therefore, we predict that the PD during the flaring state may initially increase as a result of relativistic aberration if the scattering medium has a large outflowing velocity.

Summary of System Comparison

The properties of the obscuring envelope in other systems are consistent with thermal radiative winds with an electron scattering optical depth near unity. Other systems show lower PDs than expected, which can be attributed to a lower accretion rate resulting in a less optically thick outflow. p. 9

Where the paper says it

In summary, both 4U 1630−47 and GX 13+1 appear to have high accretion rates and are expected to be surrounded by a scattering medium, yet their X-ray emission is much less polarized than in Cyg X-3. This can be attributed to an accretion rate that is only nearly Eddington or mildly super-Eddington, which results in a less optically thick outflow with a larger opening angle, leading to greater dilution of the highly polarized reflected flux by the (partly absorbed) unpolarized incident emission.

The lack of correlations or inconsistent values of PA or PD in SS 433 suggests more complex scattering geometry. p. 9
Where the paper says it

Broadly speaking, if the observed PD will positively correlate with the relative contribution of the reflected continuum to the spectrum, one can suggest a picture similar to the one considered here, where the reflected PD is high and changes gently, while the reflected flux, FI,refl/F0, is small (10−5 − 10−4 ) but changes significantly with changing inclination. The lack of such correlations or inconsistent values of PA or PD will be indicators of more complex scattering geometry, i.

The observed PD is much less polarized than in Cyg X-3 due to a lower accretion rate resulting in a less optically thick outflow with a larger opening angle. p. 9
Where the paper says it

In summary, both 4U 1630−47 and GX 13+1 appear to have high accretion rates and are expected to be surrounded by a scattering medium, yet their X-ray emission is much less polarized than in Cyg X-3. This can be attributed to an accretion rate that is only nearly Eddington or mildly super-Eddington, which results in a less optically thick outflow with a larger opening angle, leading to greater dilution of the highly polarized reflected flux by the (partly absorbed) unpolarized incident emission.

Summary of Geometric Beaming Limitations

The authors conclude that geometric beaming produced by reflection from the funnel surface was relatively weak, caused by photons escaping through the outer boundary and some light exiting the cavity at high angles. The effect is diminished even in models with a large optical depth, as some of the light is reflected toward the sides. p. 4

Where the paper says it

However, even in our highest optical depth simulation with τoutfl = 99, the reflected flux is only a factor of ∼4 greater than the direct flux when viewed at an inclination of i = 0 ◦ . Even for a truly semi-infinite funnel surface, we expect that the flux boost would still be diminished, since some of the light is reflected toward the sides.

p. 4
Where the paper says it

However, even in our highest optical depth simulation with τoutfl = 99, the reflected flux is only a factor of ∼4 greater than the direct flux when viewed at an inclination of i = 0 ◦ . Even for a truly semi-infinite funnel surface, we expect that the flux boost would still be diminished, since some of the light is reflected toward the sides.

They emphasize that testing this scenario for spectral transitions would require energy-dependent simulations. p. 8
Where the paper says it

Later during the outburst, the PD should settle toward its hard-state value. We emphasize that this is a speculative prediction, and that testing this scenario for the spectral transitions would require energy-dependent simulations.

Summary of Key Findings

The emission reflected by the funnel surface is strongly polarized only when absorption suppresses multiple scatterings, meaning most of the flux is absorbed. The presence of the optically thin medium above the funnel can qualitatively explain all spectral states in Cyg X-3. The PD during flaring state may initially increase as a result of relativistic aberration if the scattering medium has a large outflowing velocity. The observed PD then depends on the relative contribution of this scattering component and the reflection component. The optical depth of the scattering medium may need to be rather large, around τsurr ∼ 0.5, to match the observed PD of ∼10% in these states. The geometric beaming produced by reflection from the funnel surface was relatively weak in simulations. p. 9

Where the paper says it

We find that geometric beaming does not occur via reflection unless the albedo is very close to unity, and thus favors re-emission of the absorbed radiation as the mechanism behind collimation. The assumed presence of absorption nevertheless has caveats, as the funnel surface is expected to be very hot (∼107 K) and subject to intense irradiation.

The upper regions of the funnel surface must be strongly absorbing for the scattered component to be dominant. p. 8
Where the paper says it

critical optical depth of well above unity. This reiterates our conclusion that at least the upper parts of the funnel surface must be strongly absorbing for the scattered component to be dominant.

Summary of Spectral State Interpretation

The hard and intermediate states may have similar intrinsic spectra, but the intermediate state would appear softer to a distant observer due to light scattered by the medium above the funnel. The ultrasoft state cannot be explained unless the source has a softer intrinsic spectrum. The medium above the funnel can also explain the reduced PD observed during intermediate and ultrasoft states because it leads to lower scattering angles than the funnel reflection observed during the hard state. p. 8

Where the paper says it

The observed features in the hard, intermediate, and ultrasoft states are listed above, with the corresponding model sketched below. The sketches highlight the reflection from the funnel surface (blue), the scattering from the medium above the funnel (red), and the hardness of the intrinsic spectrum (teal, orange).

p. 8
Where the paper says it

The observed features in the hard, intermediate, and ultrasoft states are listed above, with the corresponding model sketched below. The sketches highlight the reflection from the funnel surface (blue), the scattering from the medium above the funnel (red), and the hardness of the intrinsic spectrum (teal, orange).

The PD in this context is not very sensitive to other parameters, suggesting its polarization should be preserved across spectral states. p. 6
Where the paper says it

In the single-scattering limit, the PD of emission scattered in the medium above the funnel (the scattering component) is set primarily by the observer inclination and is not very sensitive to other parameters. This suggests that its polarization should be preserved across spectral states.

The medium above the funnel can also explain the reduced PD observed during intermediate and ultrasoft states because it leads to lower scattering angles than the funnel reflection observed during the hard state. The optical depth of the scattering medium may need to be rather large, around τsurr ∼ 0.5, to match the observed PD of ∼10% in these states.

Geometric beaming produced by reflection from the funnel surface was relatively weak in simulations, even for a purely reflecting funnel. The effect is diminished even in models with a large optical depth, as some of the light is reflected toward the sides. The authors conclude that beaming in ULXs may be appreciably less efficient than the limiting case where all photons are collimated along the funnel. p. 9

Where the paper says it

We conclude that a more detailed model is needed to comprehensively explain geometric beaming in ULXs. While in this work we only considered the simple case of a constant albedo across the entire outflow, the lower parts of the funnel surface are likely reflection dominated, whereas absorption and re-emission may become important higher in the funnel.

The upper parts of the funnel surface must be strongly absorbing for the scattered component to be dominant. p. 8
Where the paper says it

critical optical depth of well above unity. This reiterates our conclusion that at least the upper parts of the funnel surface must be strongly absorbing for the scattered component to be dominant.

Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.

Improvements for AI systems

  1. Modeling of obscured ultraluminous X-ray sources can be improved by explicitly accounting for multiple scatterings rather than relying on single-scattering or single-reflection approximations. This allows AI systems to predict how polarization degree and collimation are sensitive to the albedo of the funnel surface under more realistic radiative transfer conditions.

  2. The improved system can generate a general framework for interpreting X-ray polarimetric signatures of obscured accretors by incorporating both semi-analytical methods and direct Monte Carlo simulations. This provides a comprehensive tool for testing different physical scenarios, such as those proposed for spectral state transitions in Cyg X-3, by examining the impact of multiple scatterings.

  3. The AI system can be enhanced to propose physical scenarios for spectral state transitions by modeling how the presence of an optically thin medium above the funnel cavity influences observed polarization degrees and flux distributions across different states. This moves beyond simply observing similar polarization values to explaining their continuity through a common scattering component.

Abstract

Super-Eddington accretion onto compact objects is expected to produce optically thick outflows with a funnel-shaped cavity that may collimate the emission. At inclinations higher than the grazing angle of the funnel, the central source is obscured. Accordingly, the observed emission is dominated by scattered and reflected radiation, which can therefore be strongly polarized. The detection of strong X-ray polarization in the Galactic X-ray binary Cygnus X-3 provides the first direct probe of this geometry. In this work, we present a systematic study of the inclination-dependent radiative signatures of such systems using a combination of semi-analytical methods and Monte Carlo simulations. Our treatment explicitly accounts for multiple scatterings and demonstrates that both the polarization degree and the degree of collimation are highly sensitive to the albedo of the funnel surface. We find that a low albedo (significant absorption) is essential for producing high polarization, yet it simultaneously suppresses the collimation of the emission. Conversely, a high-albedo medium (nearly pure scattering) can modestly collimate radiation, but at the cost of substantially reducing the polarization degree. We discuss our results in the context of Imaging X-ray Polarimetry Explorer observations of Cygnus X-3 and propose a physical scenario for its spectral state transitions, considering a combination of reflection from the funnel surface and scattering by a diffuse medium above the funnel. Our model provides a general framework for interpreting X-ray polarimetric signatures of obscured accretors.

Related papers