Bayesian inferences on covariant density functionals from multimessenger astrophysical data: Influences of parametrizations of density dependent couplings

This study employs a Bayesian framework with multimessenger astrophysical data to demonstrate that while different parametrizations of density-dependent couplings in covariant density functionals yield broadly similar inferences, the specific functional forms significantly impact the equation of state and symmetry energy at suprasaturation densities, necessitating extended flexibility in the isovector channel up to the curvature coefficient $K_{sym}$ for accurate modeling.

The Gist

Imagine the universe is filled with a cosmic "super-material" found inside neutron stars. This material is so dense that a single teaspoon would weigh as much as a mountain. Physicists call this dense nuclear matter. To understand how this material behaves, they use mathematical recipes called Covariant Density Functionals (CDFs). Think of these recipes as blueprints for building a model of the star's interior.

However, these blueprints aren't perfect. They rely on "knobs" and "dials" (parameters) that scientists have to tune. The big question this paper asks is: Does it matter exactly how we write the instructions for these dials?

Here is a simple breakdown of what the researchers did and found:

1. The Problem: Too Many Ways to Write the Recipe

In the past, scientists mostly used one specific type of instruction for how the density of the material changes. They assumed the "knobs" only reacted to the number of particles packed together (like counting how many people are in a room).

But there's another way to measure density: looking at how the particles interact with each other (like how tightly they are hugging). The researchers wanted to see if changing the type of density measurement (counting vs. hugging) or changing the mathematical shape of the instructions (using a straight line vs. a curve) would drastically change our picture of neutron stars.

2. The Experiment: A Bayesian "Taste Test"

The team used a powerful statistical method called Bayesian inference. Imagine you are a chef trying to perfect a soup recipe. You have a list of constraints:

• The soup must taste salty enough (like the mass of heavy pulsars).
• The soup must be thick enough (like the size of neutron stars measured by X-ray telescopes).
• The soup must behave a certain way when you stir it (like data from gravitational waves).

They tried six different versions of the recipe (different mathematical formulas for the density dependence). They fed all the latest astronomical data (from gravitational waves, X-ray telescopes, and particle experiments) into a computer to see which recipes could make a "soup" that satisfied all the constraints.

3. The Results: What Changed and What Didn't?

The "Big Picture" Didn't Change Much
Surprisingly, whether they counted particles or measured interactions, the final picture of the neutron star looked almost the same.

• The Analogy: Imagine you are trying to guess the weight of a mystery box. Whether you use a digital scale or a spring scale, you get the same result.
• The Finding: The maximum weight (mass) and the size (radius) of the neutron stars predicted by all the different recipes were nearly identical. The "knobs" for the basic structure of the star were flexible enough to adjust to the data regardless of the specific math used.

The "Hidden Ingredients" Did Change
While the outside of the star looked the same, what was happening inside the soup was different.

• The Analogy: Two cakes might look identical on the outside, but one is made with butter and the other with oil. You can't tell by looking, but the texture and how they cool down are different.
• The Finding: The different recipes predicted different behaviors for the symmetry energy (a property that determines how many protons vs. neutrons are in the mix).
  • Some recipes suggested the star's core would have a lot of protons (like a high-sugar cake).
  • Others suggested very few protons (like a low-sugar cake).
  • This is crucial because the amount of protons determines how fast the star cools down. If there are enough protons, the star can "scream" energy away very quickly (a process called the Direct Urca process).

4. The Conclusion: We Need Better Tools

The paper concludes that:

1. Current data is good enough to tell us the general size and weight of neutron stars, no matter which specific math recipe we use.
2. Current data is NOT good enough to tell us exactly what the "hidden ingredients" (the symmetry energy) are doing deep inside. The different recipes all fit the current observations, but they tell different stories about the star's internal composition.

The Takeaway:
To truly understand the "flavor" of the dense matter inside neutron stars, we need more than just size and weight measurements. We need new ways to look at the stars, such as watching how they cool down over time. Until then, the "recipe" for the star's interior remains a bit of a mystery, with several different versions all looking plausible.

Technical Summary

Technical Summary: Bayesian Inferences on Covariant Density Functionals from Multimessenger Astrophysical Data

Problem Statement
The behavior of dense hadronic matter, particularly within compact stars (CSs), remains a central challenge in nuclear physics and astrophysics. The Equation of State (EOS) of nuclear matter serves as the bridge between microscopic nucleon dynamics and macroscopic stellar properties. While Covariant Density Functional (CDF) theories provide a unified framework for describing nuclear data from finite nuclei to CSs, significant uncertainties persist regarding the EOS at suprasaturation densities. A critical source of this uncertainty lies in the parametrization of density-dependent meson-nucleon couplings. Existing models typically rely on specific functional forms (e.g., rational or exponential functions) and depend on either vector or scalar densities. The extent to which these specific choices—functional forms and the type of density dependence (vector vs. scalar)—influence inferences about dense matter properties and the nuclear symmetry energy, given current multimessenger constraints, requires systematic investigation.

Methodology
The authors employ a Bayesian inference framework to systematically compare different CDF models of the "density-dependent" (DD) type. The study utilizes a consistent computational framework and identical multiphysics constraints derived from:

1. Terrestrial Constraints: Nuclear matter characteristics at saturation (effective mass, saturation density, energy, incompressibility, symmetry energy slope) and ab initio predictions for pure neutron matter from Chiral Effective Field Theory ($\chi$EFT) up to $\sim 1.5-2$ times saturation density.
2. Astrophysical Constraints: Mass measurements of massive pulsars (e.g., PSR J0348+0432), tidal deformabilities from binary neutron star mergers (GW170817, GW190425), and simultaneous mass-radius estimates from the NICER X-ray observatory for four pulsars (PSR J0740+6620, J0030+0451, J0437-4715, J1231-1411).

The study varies three key aspects of the CDF parametrization:

• Density Type: Whether the couplings depend on the vector density ($n_v$), scalar density ($n_s$), or a mixed dependence.
• Functional Form: The use of rational functions (R) versus exponential functions (E) for the density dependence.
• Parameter Freedom: The number of independent parameters allowed in the isoscalar ($\sigma, \omega$) and isovector ($\rho$) channels.

Six specific models are explored (Table I in the paper), including the standard VRE model (Vector density, Rational isoscalar, Exponential isovector) and extensions such as MRE (Mixed density), MRE2 (increased isoscalar freedom), VRR (Rational isovector), and VRR2 (increased isovector freedom). The Bayesian analysis assigns uniform priors to coupling strengths and functional parameters, generating posterior distributions for nuclear characteristic coefficients (e.g., $K_{sat}, Q_{sat}, L_{sym}, K_{sym}$) and stellar observables.

Key Contributions

• Systematic Comparison of Parametrizations: The work provides a comprehensive comparison of CDF models differing in density dependence type and functional form, isolating the impact of these choices on high-density matter properties.
• Rational-Function Parametrization for Isospin Channels: The authors introduce and test rational-function parametrizations for the isovector $\rho$-meson coupling (VRR and VRR2 models), allowing for greater flexibility in describing the symmetry energy at high densities compared to the standard exponential form.
• Quantification of Model Dependence: The study quantifies how much the inferred properties of dense matter depend on the specific functional form of the density dependence versus the constraints provided by observational data.

Results

• Global Stellar Properties: The analysis reveals that CDF models with different density dependence types (vector vs. scalar) or different functional forms for the isoscalar channel yield nearly identical posterior distributions for global CS properties (maximum mass, radius, tidal deformability) within 95.4% confidence intervals. The interplay between isoscalar and isovector channels has a weaker influence on thermodynamics than variations within a single channel.
• Isoscalar Channel: Allowing the nuclear saturation characteristics in the isoscalar channel to be freely adjusted, specifically up to the skewness coefficient $Q_{sat}$, provides sufficient flexibility to reproduce current data. Models with increased isoscalar freedom (e.g., MRE2) predict slightly larger radii for massive stars ($M \gtrsim 1.4 M_\odot$) and stiffer EOS at high densities, primarily driven by broader distributions of $Q_{sat}$ and the kurtosis coefficient $Z_{sat}$.
• Isovector Channel: Variations in the isovector coupling parametrization have a modest effect on global stellar properties but lead to significant differences in the density dependence of the symmetry energy and particle composition.
  • The rational-function parametrizations (VRR, VRR2) allow for larger proton fractions in the stellar core compared to the exponential form (VRE).
  • The VRR2 model, which introduces a third independent parameter for the $\rho$-meson coupling, allows for the variation of the curvature coefficient $K_{sym}$. This model permits the onset of the direct Urca (DU) cooling process in stars with masses as low as $1.4 M_\odot$, whereas the VRE model disfavors it for $M \lesssim 2.0 M_\odot$.
• Constraints on Coefficients: Current astrophysical data weakly constrain low-order saturation parameters. Higher-order coefficients ($K_{sym}, Q_{sym}$) are the primary differentiators between models in the isovector sector. The correlation between the symmetry energy slope $L_{sym}$ and curvature $K_{sym}$ is found to be negligible in the VRR2 model.

Significance and Claims
The paper claims that while different parametrizations yield broadly comparable inferences for global stellar properties, the differences in the EOS and symmetry energy remain significant at suprasaturation densities, reflecting a sensitivity to the chosen functional form. The authors conclude that:

1. Isoscalar Modeling: Current modeling of nuclear and neutron-star matter is adequately flexible when allowing the isoscalar saturation properties (up to $Q_{sat}$) to be freely adjusted.
2. Isovector Refinement: The isovector channel requires further refinement. Extending freedom up to the curvature coefficient $K_{sym}$ is necessary to capture variations in symmetry energy and particle composition at high densities.
3. Future Directions: The rational-function parametrization of the isovector meson coupling offers a suitable framework for future simulations of CSs and heavy-ion collisions aimed at probing isospin asymmetry. However, constraining the high-density symmetry energy requires complementary strategies beyond integral stellar parameters, such as studying CS cooling and the onset of the direct Urca process.

The work advances prior studies by implementing a rational-function parametrization informed by multimessenger data, demonstrating that while low-order parameters are constrained, high-density behavior remains sensitive to the specific functional assumptions of the density dependence.