WorksPhysics and Astronomy
Nearly model-independent constraints on dense matter equation of state in a Bayesian approach
Abstract
We apply Bayesian approach to construct a large number of minimally constrained equations of state (EOSs) and study their correlations with a few selected properties of a neutron star (NS). Our set of minimal constraints includes a few basic properties of saturated nuclear matter and low-density pure neutron matter EOS which is obtained from a precise next-to-next-to-next-to-leading-order (NLO) calculation in chiral effective field theory. The tidal deformability and radius of NS with mass are found to be strongly correlated with the pressure of -equilibrated matter at densities higher than the saturation density ( fm) in a nearly model-independent manner. These correlations are employed to parametrize the pressure for -equilibrated matter, around 2, as a function of neutron star mass and the corresponding tidal deformability. The maximum mass of neutron star is also found to be strongly correlated with the pressure of -equilibrated matter at densities .
I Introduction
Gravitational-wave astronomy promises unprecedented constraints on the Equation of State of neutron star matter through the detailed properties of gravitational-waveform observed during the merging of binary neutron stars (BNS). In addition, X-ray observations from the Neutron star Interior Composition Explorer (NICER) instruments have also provided a compelling constraint on the equation of state independently. The tidal deformability parameters inferred from these gravitational-wave events encode information about the EOS. For the first time,a BNS event (GW170817) was observed by the LIGO-Virgo detector from a low mass compact binary neutron star merger with a total mass of the system [1, 2]. Another gravitational-wave event likely originating from the coalescence of BNSs, GW190425, was observed [3] subsequently. These two events have already triggered many theoretical investigations to constrain the EOS of neutron star matter [3, 4, 5, 6, 7, 8, 9, 10, 11, 12]. The upcoming runs of LIGO-Virgo-KAGRA and the future detectors, e.g., Einstein Telescope (ET) and Cosmic Explorer (CE), are expected to observe many more BNS signals emitted from coalescing neutron stars. The mass and radius of NS, observed either in isolation or in binaries, by NICER [13, 14, 15] have offered complementary constraints on the EOS. A sufficiently large number of such observations over a wide range of NS masses may be employed to constrain several key quantities associated with the EOS of -equilibrated matter which are not readily accessible in the terrestrial laboratory. The behavior of the EOS at supra-saturation densities are generally studied using the observed maximum neutron star mass, together with radius and tidal deformability corresponding to the neutron star with canonical mass [16, 17, 18]. Recently, in Refs. [19, 20, 21, 22, 23, 24], efforts are made to constrain the EOS of -equilibrated matter which is relevant to the studies of NS properties. The values of tidal deformability of NS with mass are found to be strongly correlated with the EOS at twice the saturation density.
Statistical tools are quite helpful in providing a quantitative interpretation of NS observables. A Bayesian approach is often applied to analyze gravitational-wave signals, which involves nearly 15 parameters for binary compact object mergers, to infer their source properties [25]. It has been also extended to investigate the properties of short gamma-ray bursts [26], neutron stars [27, 28, 29], the formation history of binary compact objects [30, 31, 32, 33, 34] and to test general relativity [35, 36, 37, 38]. Of late, Bayesian approach has become a useful statistical tool for parameter estimation in the field of nuclear physics and nuclear-astrophysics [39]. It allows one to obtain joint posterior distributions of the model parameters and the correlations among them for a given set of data. Various constraints on the parameters known a priori are incorporated through their prior distributions. The Bayesian techniques have also been employed to constrain symmetry energy [40], masses and radii of NSs [41] using the bounds on the EOS obtained from chiral effective field theory. Bayesian techniques have been extensively applied to constrain the EOS for symmetric nuclear matter, -equilibrated matter (BEM) and density dependence of symmetry energy coefficient using various finite nuclei and NS properties [42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53].
We use Bayesian approach to construct large sets of EOSs which correspond to the Taylor and expansions[53]. The expansion coefficients in the former case are the individual nuclear matter parameters (NMPs), whereas in the latter case they are their linear combinations. The EOSs are consistent with a set of minimal constraints that includes a few low-order nuclear matter parameters at the saturation density and EOS for the pure neutron matter (PNM) at low densities obtained from a precise next-to-next-to-next-to-leading-order (NLO) calculation in chiral effective field theory. The marginalized posterior distributions of NMPs and the various NS properties obtained from set of minimal constraints are found to be within reasonable bounds. The correlations of various NS properties, such as tidal deformability, radius and maximum mass, with key EOS parameters are studied. These correlations are investigated for a wide range of NS masses and densities for the EOS.
The paper is organized as follows. The Taylor and expansions for the EOS of neutron star matter and the Bayesian approach are briefly outlined in § 2. The results for the posterior distributions of nuclear matter parameters and associated NS properties together with their correlations with some key quantities associated with EOS are presented in Sec. III. The main outcomes of the present investigation are summarized in § 4.
Methodology
The energy per nucleon for neutron star matter at a given total nucleon density and asymmetry can be decomposed into the energy per nucleon for the symmetric nuclear matter, and the density-dependent symmetry energy, in the parabolic approximation as,
where, with and being the neutron and proton densities, respectively. The value of at a given is determined by the condition of -equilibrium and the charge neutrality. Once is known, the fraction of neutrons, protons, electrons, muons can be easily evaluated. In the following, we expand and appearing in Eq. (1) using Taylor and expansions. The coefficients of expansion in case of the Taylor correspond to the individual nuclear matter parameters. In the latter case, they are expressed as linear combinations of the nuclear matter parameters.
Taylor's expansion
The and can be expanded around the saturation density as [54, 55, 56, 57, 58],
so that,
where the coefficients and are the nuclear matter parameters. We truncate the sum in Eqs. (2) and (3) at fourth order, i.e., - 4. Therefore, the coefficients and correspond,
In Eqs. (5) and (6), is the binding energy per nucleon, the incompressibility coefficient, the symmetry energy coefficient, its slope parameter , the symmetry energy curvature parameter, and are related to third- and fourth-order density derivatives of [], respectively. The subscript zero indicates that all the nuclear matter parameters are calculated at the saturation density.
It may be noticed from Eq. (4) that the coefficients and may display some correlations among themselves provided the asymmetry parameter depends weakly on the density. Furthermore, Eq. (4) may converge slowly at high densities, i.e., . This situation is encountered for the heavier neutron stars. Neutron stars with a mass around , typically have central densities .
n3 expansion
An alternative expansion of can be obtained by expanding and as [59, 60],
We refer this as the expansion. It is now evident from Eqs. (7) and (8) that the coefficients of expansion are no longer the individual nuclear matter parameters unlike in the case of Taylor’s expansion. The values of the nuclear matter parameters can be expressed in terms of the expansion coefficients and , as respectively.
The relations between the expansion coefficients and the nuclear matter parameters are governed by the nature of functional form for and . The off-diagonal elements in the above matrices would vanish for the Taylor expansion of and as given by Eqs. (2) and (3), respectively. Therefore, each of the expansion coefficients is simply the individual nuclear matter parameter given by Eqs. (5) and (6). Inverting the matrices in Eqs. (10) and (11) we have
Each of the coefficients and are the linear combinations of nuclear matter parameters in such a way that the lower-order parameters may contribute dominantly at low densities. The effects of higher-order parameters become prominent with the increase in density.
Bayesian estimation of nuclear matter parameters
A Bayesian approach enables one to carry out detailed statistical analysis of the parameters of a model for a given set of fit data. It yields joint posterior distributions of model parameters which can be used not only to study the distributions of given parameters but also to examine correlations among model parameters. One can also incorporate prior knowledge of the model parameters and various constraints on them through the prior distributions. This approach is mainly based on the Bayes theorem which states that [61],
where and denote the set of model parameters and the fit data, respectively. The is the joint posterior distribution of the parameters, is the likelihood function, is the prior for the model parameters and is the evidence. The posterior distribution of a given parameter can be obtained by marginalizing over remaining parameters. The marginalized posterior distribution for a parameter can be obtained as,
We use Gaussian likelihood function defined as,
Here the index runs over all the data, and are the data and corresponding model values, respectively. The are the adopted uncertainties. The evidence in Eq. (14) is obtained by complete marginalization of the likelihood function. It is relevant when employed to compare different models. However in the present work is not very relevant. To populate the posterior distribution of Eq. (14), we implement a nested sampling algorithm by invoking the Pymultinest nested sampling [62] in the Bayesian Inference Library [25].
Results and Discussions
We obtained the EOSs for -equilibrated matter (BEM) using Taylor and expansions as discussed in previous section Eqs. (4) and (9). The coefficients of the Taylor expansion are the individual nuclear matter parameters, whereas, they correspond to linear combinations of nuclear matter parameters for the expansion. We have constructed marginalized posterior distributions for the nuclear matter parameters by applying a Bayesian approach to both the expansions considered. The nuclear matter parameters of the corresponding EOSs are consistent with a set of minimal constraints that includes basic properties of saturated nuclear matter and low-density () EOS for the pure neutron matter from (NLO) calculation in chiral effective field theory [63]. This large number of EOSs is employed to evaluate the properties of neutron star such as tidal deformability, radius and maximum mass. The correlations of neutron star properties with the pressure of -equilibrated matter at a given density are studied. Most of these correlations are sensitive to the choice of the neutron star mass and EOS at a given density. Our results for the correlations of tidal deformability with pressure for -equilibrated matter are analogous to those obtained using a diverse set of nonrelativistic and relativistic mean-field models (MFMs) that re-emphasize their model independence. Such model-independent trends inspire us to parametrize the pressure for -equilibrated matter around in terms of neutron star mass and the corresponding tidal deformability.
Priors, likelihood and filters
We apply Bayesian approach to obtain two large sets of EOSs corresponding to the Taylor and expansions. The posterior distributions for the NMPs are obtained by subjecting the EOSs to a set of minimal constraints which include some basic properties of nuclear matter evaluated at the saturation density and EOS for the pure neutron matter at low-density. The constraints on the nuclear matter parameters are incorporated through the priors and those from the EOS for the pure neutron matter through the likelihood function. Not all the nuclear matter parameters are well constrained. Only a very few low-order nuclear matter parameters constrained within narrow bounds are the binding energy per nucleon MeV, nuclear matter incompressibility coefficients MeV for the symmetric nuclear matter and symmetry energy coefficient MeV. The values of and are very well constrained by the binding energy of finite nuclei over a wide range of nuclear masses [10, 64, 65, 66, 67, 68]. The value of is constrained from the experimental data on the centroid energy of isoscalar giant monopole resonance in a few heavy nuclei [69, 70]. The values of have been extracted from experimental data on variety of phenomena in the finite nuclei as well as from neutron star observations. The model-independent estimates of is expected to be derived from the measurement of neutron-skin thickness in asymmetric nuclei. Recent measurement of neutron-skin thickness in nucleus yields MeV[71]. However, this value of has only marginal overlap at the lower side with those determined using experimental data on iso-vector giant dipole resonances in several nuclei[72] and recent neutron star observations[73]. The remaining nuclear matter parameters, , , , and are constrained only poorly [24, 74, 75, 76, 77]. The priors for the nuclear matter parameters employed in the present work are listed in Table I. The prior distributions of , and are assumed to be Gaussian with rather smaller width, whereas, the other higher order nuclear matter parameters correspond to Gaussian distribution with very large width. We have also repeated our calculations with uniform priors for the higher order nuclear matter parameters and the result for the median values are found to be practically unaltered and uncertainties are modified marginally, up to 10%(not shown). In what follows, we present only those results which are obtained with priors as listed in Table I.
| NMPs | Pr-Dist | ||
| (in MeV) | |||
| G | -16 | 0.3 | |
| G | 240 | 50 | |
| G | -400 | 400 | |
| G | 1500 | 1500 | |
| G | 32 | 5 | |
| G | 50 | 50 | |
| G | -100 | 200 | |
| G | 550 | 400 | |
| G | -2000 | 2000 |
Table I. The prior distributions of the nuclear matter parameters. The nuclear matter parameters considered are the binding energy per nucleon (), incompressibility coefficient (), symmetry energy coefficient (), it’s slope parameter (), symmetry energy curvature parameter () and and are related to third and fourth order density derivatives of (), respectively. All the nuclear matter parameters are evaluated at saturation density fm. The parameters of Gaussian distribution (G) are the mean () and standard deviation ().
We know that the direct application of the lattice QCD simulations are challenging to hadronic physics at finite density due to sign problem in Monte Carlo simulations. However, analytical calculations in terms of the effective degrees of freedom at low energy () like chiral effective theory is valid with negligible uncertainty. The precise next-to-next-to-next-to-leading-order (NLO) calculation are usually fitted to the nucleon–nucleon scattering cross section or few-body observables, and even saturation properties of heavier nuclei [78]. The low-density EOS for the pure neutron matter obtained from a (NLO) calculation in chiral effective field theory [63] is employed as pseudodata to obtain a sample like-lihood function as given by Eq. (16). The and the in Eq. (16) are the pseudodata for the energy per neutron and the corresponding uncertainties taken from Ref. [63]. This has been employed in past many of the analyses as their pseudodata [22, 79, 80, 81, 82]. We have considered the values of energy per neutron over the density range . At the densities lower than , the neutron star matter is expected to be clusterized.
We have filtered the nuclear matter parameters by demanding that (i) pressure for the -equilibrated matter should increase monotonically with density (thermodynamic stability), (ii) speed of sound must not exceed the speed of light (causality) and (iii) maximum mass of neutron star must exceeds (observational constraint). The causality breaks down at higher density mostly for the Taylor EOS. In such cases, we use the stiffest EOS, , where, and are the pressure and corresponding energy density at which the causality breaks [83].
Posterior distribution of nuclear matter parameters
To undertake the correlation systematics as proposed, we need a large number of EOSs with diverse behavior and corresponding neutron star properties. The posterior distributions for the nuclear matter parameters for the Taylor and expansions are obtained by subjecting the EOS to a set of minimal constraints as discussed above. The joint posterior distribution of the NMPs for a given model depends on the product of the likelihood and the prior distribution of nuclear matter parameters (Eq. (14)). The posterior distribution of each individual parameter is obtained by marginalizing the joint posterior distribution with the remaining model parameters. If the marginalized posterior distribution of a nuclear matter parameter is localized more than the corresponding prior distribution, then, the nuclear matter parameter is said to be well constrained by the data used for model fitting. The corner plots for the marginalized posterior distributions for the nuclear matter parameters in one and two dimensions obtained for Taylor and expansions are displayed in Figure 1 and Figure 2, respectively. The differences between the one-dimensional posterior distributions for the nuclear matter parameters and corresponding prior distributions reflect the role of low-density EOS for pure neutron matter in constraining the nuclear matter parameters. The EOS for the pure neutron matter mainly constraints the values of , and and to some extent and . The shapes and the orientations of the confidence ellipses suggest that the correlations among most of the NMPs are weak. Most Strong correlations exist only between , and for both the expansions with correlation coefficient r 0.8. The correlation is slightly better in case of expansion (r -0.6) as compared to Taylor (r -0.18). The median values of the nuclear matter parameters and the corresponding 68%(90%) confidence intervals obtained from the marginalized posterior distributions are listed in Table III(see Appendix ). We also provide the values for the nuclear matter parameters obtained without the PNM constraints. The low-density pure neutron matter mainly constraints those nuclear matter parameters which are associated with the density dependence of the symmetry energy. The median values of and , which determined the linear and quadratic density dependence of the symmetry energy, become smaller suggesting softer symmetry energy ”at high-density” with the inclusion of pure neutron matter constrains. Furthermore, the uncertainties on reduced by more than 50%. The median value of remain more or less unaltered. From the recent measurement of the neutron-skin thickness for nucleus (PREX-II)[71, 84], fm, the value of has been determined to be 106 37 MeV [71] . This value of agrees with the ones obtained in the present work with PNM constrained only within 90% confidence interval.
Figure 1. Corner plots for the nuclear matter parameters (in MeV) obtained for Taylor expansions for the EOS of asymmetric nuclear matter. The one dimensional marginalized posterior distributions (salmon) and the prior distributions (green lines) are displayed along the diagonal plots. The vertical lines indicate 68% confidence interval of nuclear matter parameters. The confidence ellipses for two-dimensional posterior distributions are plotted with , and confidence intervals along the off-diagonal plots. The distributions of nuclear matter parameters are obtained by subjecting them to minimal constraints (see text for details).
Figure 2. The same as Figure 1, but, for expansions for the EOS of asymmetric nuclear matter.
| NMPs (in MeV) | without PNM: Taylor | without PNM: | with PNM: Taylor | with PNM: |
Table III. The median values and associated 68%(90%) uncertainties for the nuclear matter parameters from their marginalized posterior distributions. The results are obtained for Taylor and expansions with and without pure neutron matter (PNM) constraints.
Properties of neutron stars
Once the EOS for the core and crust are known the values of NS mass, radius and tidal deformability corresponding to given central pressure can be obtained by solving Tolman-Oppenheimer-Volkoff equations [85, 86]. The EOSs for core region of neutron star, correspond to the -equilibrated matter over the density range , are obtained from the posterior distributions of nuclear matter parameters for the Taylor and expansions. The core EOSs are matched to the crust EOSs for obtaining the NS properties. The EOS for outer crust is taken to be the one given by Baym, Pethick, and Sutherland [87]. The inner crust that joins the inner edge of the outer crust and the outer edge of the core is assumed to be polytropic [88], . Here, the parameters and are determined in such a way that the EOS for the inner crust matches with the outer crust at one end ( fm) and with the core at the other end (). The polytropic index is taken to be equal to 4/3. The radii of neutron star with mass are more sensitive to the treatment of crust EOS [89]. It is demonstrated that the treatment of crust EOS employed in the present work may introduce the uncertainties of about 50-100 m in radii of NSs having mass . It is shown in Ref. [90] that the choice of EOS for inner crust does not significantly impact the values of tidal deformability which depends on the Love number as well as the compactness parameter.
We have obtained the distributions of , , and using the posterior distributions for the nuclear matter parameters corresponding to the Taylor and expansions. The corner plots for these NS properties are displayed in Figure 3. The effective priors for the NS properties as shown by green lines are obtained using the priors for the nuclear matter parameters. The posterior distributions of NS properties are narrower than the corresponding effective priors indicating the significance of the low-density EOS for the pure neutron matter. The posterior distributions of and for both the expansions are quite close to each other. The differences begin to appear for the case of which become even larger for the maximum mass. This is due to the fact that the Taylor EOSs are much more stiffer than the those for . The dichotomy in the high-density behavior of the Taylor and expansions would help us to understand the extent to which the correlations of the EOSs with the properties of NS, for masses in the range 1 - , are model dependent. It is clear from off-diagonal plots that is strongly correlated with , the correlation coefficient is r 0.9. The and also display stronger correlations with (r 0.8) for the case of Taylor and somewhat moderate correlations (r 0.7) for the expansion. The maximum mass of neutron star is almost uncorrelated with the other NS properties considered.

Figure 3. Corner plots for the marginalized posterior distributions (salmon) of the tidal deformability , radii (km) and (km) and the maximum mass () for Taylor (top) and (bottom) expansions. The green lines represent effective priors obtained using the priors for nuclear matter parameters (see also Table I).
We have summarized in Table IV (see Appendix ) the median values of NS properties along with 68% (90%) confidence intervals. Like in the case of nuclear matter parameters, the NS properties get significantly constrained by the EOS of pure neutron matter at low-density. For instance, the median values of become smaller by about 15% and the associated uncertainties by about 40% with the pure neutron matter constraints. The median values of and the corresponding uncertainties also become noticeably smaller. The and do not show significant changes with the inclusion of low-density pure neutron matter constraints. With the PNM constraints, the 90% confidence interval of the neutron star properties such as tidal deformability, radius and mass overlap with the currently available bounds, [4], km [97], km [96] and [98]. The obtained for the Taylor EOS is on the slightly higher side in comparison to the ones derived by combining the GW170817 observations of merging of binary neutron stars and quasiniversal relation [99]. The observed electromagnetic emissions in the form of kilonova and the detection of a gamma-ray burst has been linked to the formation of a black hole and thus, have been utilized to infer the maximum mass of a stable neutron star. However, such inference of the maximum mass is subjected to uncertainties originating from model dependence of postmerger dynamics. Recent observation of GW190814 event, a neutron star black hole/binary neutron star merger,
| NS properties | without PNM: Taylor | without PNM: | with PNM: Taylor | with PNM: |
Table IV. Similar to Table III, but, for the neutron star properties, namely the tidal deformability (), radii ( and ) and maximum mass ().
has triggered an assessment of the maximum mass of a stable neutron star [100]. While there are different opinions available in the literature, the nature of a compact object in the range of - being neutron star or black hole seems to be an unsettled issue to date [99, 100, 101, 102, 103, 104]. So the maximum mass () we got for the Taylor model supporting the static NS of mass greater than may not be ruled out at present.
We obtain joint probability distribution for a given mass and radius for both the Taylor and expansions. They display qualitatively very much similar trends. In Figure 4, we plot the obtained for the expansion. The 90% confidence interval is represented by red dashed line. The color gradient from orange to dark purple represents the lowest to highest probability. The most probable values for and are approximately 13.3 and 12.3 km, respectively. The is maximum for , km. The 90% confidence interval has significant overlap with LIGO-Virgo and NICER estimations. It may be however pointed out that the main objective of the present work is to construct large sets of EOSs with diverse behavior to assess various correlation systematics as follows.

Figure 4. Plot for joint probability distribution as a function of mass and radius of neutron star obtained for expansion. The red dashed line represents the 90% confidence interval. The outer and inner gray shaded regions indicate the 90% (solid) and 50% (dashed) confidence interval of the LIGO-Virgo analysis for BNS component from the GW170817 event [91, 92, 93]. The rectangular regions enclosed by dotted lines indicate the constraints from the millisecond pulsar PSR J0030+0451 (purple & black) NICER x-ray data [94, 95] and PSR J0740+6620 (green) [96]
Correlations of neutron star properties with EOS
We randomly select 1000 EOSs and corresponding NS properties from marginalized posterior distributions obtained for the Taylor as well as expansions. They are used to study the correlations of various NS properties with key quantities determining the behavior of the EOS. The correlations of , , and with the pressure of -equilibrated matter over a wide range of densities are evaluated. The values of correlation coefficients are plotted as a function of density in Figure 5. We also display the values of correlation coefficients for NS properties with the pressure of -equilibrated

Figure 5. The correlation coefficients , as approximated by both Taylor and expansion along with the mean-field theory calculations, is shown in this figure. Here, represents either of the tidal deformability , radii , , or maximum mass of the neutron star, whereas, represents the pressure for -equilibrated matter at a density . The calculations are performed with neutron star properties obtained using marginalized posterior distributions of nuclear matter parameters in Taylor and expansions. For the comparison the results are also displayed for a diverse set of nonrelativistic and relativistic microscopic mean-field models (MFMs).
matter calculated using unified EOSs for a diverse set of 41 nonrelativistic and relativistic microscopic mean-field models (MFMs) [89]. The various NS properties considered show strong correlations with around a particular density. The density at which the correlation is maximum increases with the NS mass. The values of and are strongly correlated with at density . The is strongly correlated with around . The is strongly correlated with around . Our results for the Taylor and expansions for the region of maximum correlations are in line with those obtained using a diverse set of mean-field models, except for the case of . Thus, it seems possible that the EOS over a range of densities beyond can be constrained in a nearly model-independent manner with the help of various NS observables.
In Table V (see the Appendix), we list the values of correlation coefficients obtained between the NS properties and the
EOS at some selected densities. The correlation coefficients are obtained using 100 and 1000 EOSs, corresponding to Taylor and expansions, randomly selected from the posterior distributions. We also present the results which are obtained by combining 1000 EOSs corresponding to each of the expansions. The values of correlation coefficients for the combined set of EOSs are close to those obtained separately. The values of the correlation coefficients are close to those obtained for mean-field models are listed in second column. We plot in Figure 6, the variations of , at selected densities, with , , and for which the correlations are stronger. We compare our results with those obtained from a diverse set of mean-field models. The correlation lines obtained by combining results of the Taylor and expansions are also plotted to estimate the values of at selected densities with help of NS properties. The equations for the correlation lines are obtained using linear regression as,

Figure 6. The variations of pressure for -equilibrated matter [] at selected densities versus tidal deformability , radii and and maximum mass of neutron star. The red dashed lines are obtained by linear regression [see Eq. (17) in Sec. III D].
We extend our analysis for the correlations of the pressure for the -equilibrated matter with tidal deformability over a wide range of neutron stars mass. In Figure 7, we display color-coded graph for the correlations of tidal deformability of neutron star for the mass with the pressure for -equilibrated matter at densities (). One can easily obtain the value of correlation coefficient as a function of density at a given NS mass. The at are strongly correlated () with tidal deformability for NS masses in the range . Hence, can be parametrized at a given as,

Figure 7. Dependence of correlation coefficients between tidal deformability () and the pressure of -equilibrated matter () on neutron star mass (M) and density () is depicted in this plot. Here is used only for scaling purposes.
| Pressure | ||||||
| (in MeV fm) | ||||||
Table II. The median values and associated 68%(90%) uncertainties for the parameters, appearing in Eq. (18), obtained from their marginalized posterior distributions. The values of parameters , and as listed are scaled up by a factor of 10.
with mass-dependent coefficients and expanded as
respectively, where is taken to be and the values of and are estimated using a Bayesian approach with the help of and tidal deformability obtained for Taylor and expansions. For a given , the Eq. (18) is fitted using the tidal deformability corresponding to NS mass
. The priors for and are taken to be uniform in the range of -100 to 100. The calculations are performed for , 2.0 and . All the s are strongly correlated with corresponding s. The median values of parameters and and associated uncertainties are summarized in Table II. It may be noticed that the values of and for the case of are not the same as those in Eq. (17). This may be partly due to the strong correlations between and of Eq. (18). Moreover, Eq. (17) is fitted to the values of tidal deformability at a fixed NS mass . To validate our parametrized form for , we have calculated the values of using Eq. (18) with the help of tidal deformability for obtained for large number of mean-field models which includes the once considered in Figure 5 along with those taken from [23, 81, 105]. The average deviation of , obtained using Eq. (18), from the actual values is about 10%. We find marginal improvement when the terms corresponding to quadratic in tidal deformability are included in Eq. (18).
In Figure 8, we display the variations of tidal deformability as a function of mass and pressure for -equilibrated matter at , 2.0 and . These results are obtained using the parametrized form for as given by Eq. (18). One can easily estimate the values of for once the values of tidal deformability known in NS mass ranges .

Figure 8. The median values of pressure for -equilibrated matter is shown here as a function of neutron star mass and its tidal deformability at densities (top), (middle) and (bottom).
Conclusions
We have used Taylor and expansions of equations of state to construct marginalized posterior distributions of the nuclear matter parameters which are consistent with the minimal constraints. Only a few low-order nuclear matter parameters, such as the energy per nucleon, incompressibility coefficient for the symmetric nuclear matter and symmetry energy coefficients at the saturation density (), are constrained in narrow windows along with the low-density pure neutron matter EOS obtained from a precise next-to-next-to-next-to-leading-order (NLO) calculation in chiral effective field theory. The tidal deformability, radius and maximum mass are evaluated using large sets of minimally constrained EOSs.
The correlations of neutron star properties over a wide range of mass with various key quantities characterizing the EOS are investigated. We find that the values of tidal deformability and radius for the neutron star with are strongly correlated with the pressure for the -equilibrated matter at density . The radius for neutron star is strongly correlated with the pressure for -equilibrated matter at density . The maximum mass of neutron star is correlated with the pressure for the -equilibrated matter at density . These correlation systematics are in harmony with those obtained for unified EOSs for the -equilibrated matter available for a diverse set of nonrelativistic and relativistic mean-field models. We exploit the model independence of correlations to parametrize the pressure for -equilibrated matter, in the density range , in terms of the mass and corresponding tidal deformability of neutron star. Such parametric form may facilitate back-of-the-envelope estimation of the pressure at densities around for a given value of tidal deformability of neutron stars with mass in the range of .
Acknowledgements
The authors would like to thank C. Mondal, for a careful reading of the paper, and important suggestions. N.K.P. would like to thank T.K. Jha for constant encouragement and support and gratefully acknowledge the Department of Science and Technology, Ministry of Science and Technology, India, for the support of DST/INSPIRE Fellowship/2019/IF190058. AM acknowledges support from the DST-SERB Start-up Research Grant SRG/2020/001290. The authors sincerely acknowledge the usage of the analysis software BILBY [25, 106] and open data from GWOSC [93]. This research has made use of data or software obtained from the gravitational-wave Open Science Center (gw-openscience.org), a service of LIGO Laboratory, the LIGO Scientific Collaboration, the Virgo Collaboration, and KAGRA. LIGO Laboratory and Advanced LIGO are funded by the United States National Science Foundation (NSF) as well as the Science and Technology Facilities Council (STFC) of the United Kingdom, the Max-Planck-Society (MPS), and the State of Niedersachsen/Germany for support of the construction of Advanced LIGO and construction and operation of the GEO600 detector. Additional support for Advanced LIGO was provided by the Australian Research Council. Virgo is funded, through the European Gravitational Observatory (EGO), by the French Centre National de Recherche Scientifique (CNRS), the Italian Istituto Nazionale di Fisica Nucleare (INFN) and the Dutch Nikhef, with contributions by institutions from Belgium, Germany, Greece, Hungary, Ireland, Japan, Monaco, Poland, Portugal, Spain. The construction and operation of KAGRA are funded by Ministry of Education, Culture, Sports, Science and Technology (MEXT), and Japan Society for the Promotion of Science (JSPS), National Research Foundation (NRF) and Ministry of Science and ICT (MSIT) in Korea, Academia Sinica (AS) and the Ministry of Science and Technology (MOST) in Taiwan. TM acknowledges support from FCT (Fundação para a Ciência e a Tecnologia, I.P, Portugal) under the Projects No. UID/FIS/04564/2019, No. UIDP/04564/2020, No. UIDB/04564/2020, and No. POCI-01-0145-FEDER-029912 with financial support from Science, Technology and Innovation, in its FEDER component, and by the FCT/MCTES budget through national funds (OE).
Some useful tables
We present our results in tabular form which are obtained with the minimal constraints. The values of the nuclear matters, properties of neutron stars and their correlations with various key quantities associated with EOS are listed in Tables III–V. These results are depicted in Figs. 1 - 6.
| Name of Pairs | MFMs | Taylor: 100 | Taylor: 1000 | : 100 | : 1000 | combined 2000 |
| - | 0.90 | 0.98 | 0.98 | 0.99 | 0.98 | 0.98 |
| - | 0.83 | 0.93 | 0.93 | 0.94 | 0.93 | 0.93 |
| - | 0.81 | 0.93 | 0.91 | 0.92 | 0.92 | 0.93 |
| - | 0.99 | 0.97 | 0.98 | 0.95 | 0.96 | 0.99 |
Table V. The comparison of values for Pearson’s correlation coefficient (r) obtained from randomly selected 100 and 1000 EOSs using both Taylor and expansions. The values of correlation coefficients are also obtained by combining 1000 EOSs from each of the expansions. For comparison, the values of r obtained for a diverse set of mean-field models are also presented in 2nd column.
References
References
- [1]B. P. Abbott et al., Phys. Rev. X 9, 011001 (2019).
- [2]B. P. Abbott et al., Astrophys. J. 882, L24 (2019).
- [3]B. P. Abbott et al., Astrophys. J. 892, L3 (2020).
- [4]B. P. Abbott et al., Phys. Rev. Lett. 119, 161101 (2017).
- [5]T. Malik, N. Alam, M. Fortin, C. Providência, B. K. Agrawal, T. K. Jha, B. Kumar, and S. K. Patra, Phys. Rev. C 98, 035804 (2018).
- [6]S. De, D. Findstad, J. M. Lattimer, D. A. Brown, E. Berger, and C. M. Biwer, Phys. Rev. Lett. 121, 091102 (2018).
- [7]F. J. Fattoyev, J. Piekarewicz, and C. J. Horowitz, Phys. Rev. Lett. 120, 172702 (2018).
- [8]P. Landry and R. Essick, Phys. Rev. D 99, 084049 (2019).
- [9]J. Piekarewicz and F. J. Fattoyev, Phys. Rev. C 99, 045802 (2019).
- [10]T. Malik, B. K. Agrawal, J. N. De, S. K. Samaddar, C. Providência, C. Mondal, and T. K. Jha, Phys. Rev. C 99, 052801 (2019).
- [11]B. Biswas, P. Char, R. Nandi, and S. Bose, Phys. Rev. D 103, 103015 (2021).
- [12]H. D. Thi, C. Mondal, and F. Gulminelli, Universe 7, 373 (2021).
- [13]A. L. Watts et al., Rev. Mod. Phys. 88, 021001 (2016).
- [14]M. C. Miller et al., Astrophys. J. 887, L24 (2019).
- [15]T. Riley et al., Astrophys. J. Lett. 887, L21 (2019).
- [16]N.-B. Zhang, B.-A. Li, and J. Xu, Astrophys. J. 859, 90 (2018).
- [17]B. J. Cai and B. A. Li, Phys. Rev. C 103, 034607 (2021).
- [18]H. Gil, P. Papakonstantinou, and C. H. Hyun, Int. J. Mod. Phys. E 31, 2250013 (2022).
- [19]J. M. Lattimer and M. Prakash, Astrophys. J. 550, 426 (2001).
- [20]A. Maselli, V. Cardoso, V. Ferrari, L. Gualtieri, and P. Pani, Phys. Rev. D 88, 023007 (2013).
- [21]B. P. Abbott et al., Phys. Rev. Lett. 121, 161101 (2018).
- [22]Y. Lim and J. W. Holt, Phys. Rev. Lett. 121, 062701 (2018).
- [23]C. Y. Tsang, M. B. Tsang, P. Danielewicz, W. G. Lynch, and F. J. Fattoyev, Phys. Lett. B 796, 1 (2019).
- [24]C. Y. Tsang, M. B. Tsang, P. Danielewicz, W. G. Lynch, and F. J. Fattoyev, Phys. Rev. C 102, 045808 (2020).
- [25]G. Ashton et al., Astrophys. J. Suppl. Ser. 241, 27 (2019).
- [26]S. Biscontin, E. Thrane, and S. Vitale, Astrophys. J. 893, 38 (2020).
- [27]M. W. Coughlin and T. Dietrich, Phys. Rev. D 100, 043011 (2019).
- [28]F. Hernandez Vivanco, R. Smith, E. Thrane, P. D. Lasky, C. Talbot, and V. Raymond, Phys. Rev. D 100, 103009 (2019).
- [29]S. Biswas, S. Vitale, and C.-J. Haster, Astrophys. J. 884, L32 (2019).
- [30]M. E. Lower, E. Thrane, P. D. Lasky, and R. Smith, Phys. Rev. D 98, 083028 (2018).
- [31]I. M. Romero-Shaw, P. D. Lasky, and E. Thrane, Mon. Not. R. Astron. Soc. 490, 5210 (2019).
- [32]A. Ramos-Buades, S. Husa, G. Pratten, H. Estellés, C. García-Quirós, M. Mateu-Lucena, M. Colleoni, and R. Jaume, Phys. Rev. D 101, 083015 (2020).
- [33]I. M. Romero-Shaw, N. F. Farrow, S. Stevenson, E. Thrane, and X. J. Zhu, Mon. Not. R. Astron. Soc. 496, L64 (2020).
- [34]M. Zevin, C. P. L. Berry, S. Coughlin, K. Chatziioannou, and S. Vitale, Astrophys. J. 899, L17 (2020).
- [35]D. Keitel, Res. Notes AAS 3, 46 (2019).
- [36]G. Ashton and S. Khan, Phys. Rev. D 101, 064037 (2020).
- [37]E. Payne, C. Talbot, and E. Thrane, Phys. Rev. D 100, 123017 (2019).
- [38]Z. C. Zhao, H. N. Lin, and Z. Chang, Chin. Phys. C 43, 075102 (2019).
- [39]S. Wesolowski, N. Klco, R. J. Furnstahl, D. R. Phillips, and A. Thapaliya, J. Phys. G 43, 074001 (2016).
- [40]R. Somasundaram, C. Drischler, I. Tews, and J. Margueron, Phys. Rev. C 103, 045803 (2021).
- [41]C. Drischler, S. Han, J. M. Lattimer, M. Prakash, S. Reddy, and T. Zhao, Phys. Rev. C 103, 045808 (2021).
- [42]I. Tews, J. M. Lattimer, A. Ohnishi, and E. E. Kolomeitsev, Astrophys. J. 848, 105 (2017).
- [43]T. Carreau, F. Gulminelli, and J. Margueron, Phys. Rev. C 100, 055803 (2019).
- [44]H. D. Thi, T. Carreau, A. F. Fantina, and F. Gulminelli, Astron. Astrophys. 654, A114 (2021).
- [45]T. E. Riley, G. Raaijmakers, and A. L. Watts, Mon. Not. R. Astron. Soc. 478, 1093 (2018).
- [46]G. Raaijmakers and Others, Astrophys. J. Lett. 893, L21 (2020).
- [47]J.-L. Jiang, S.-P. Tang, Y.-Z. Wang, Y.-Z. Fan, and D.-M. Wei, Astrophys. J. 892, 1 (2020).
- [48]H. Güven, K. Bozkurt, E. Khan, and J. Margueron, Phys. Rev. C 102, 015805 (2020).
- [49]B. Biswas, Astrophys. J. 921, 63 (2021).
- [50]P. Landry, R. Essick, and K. Chatziioannou, Phys. Rev. D 101, 123007 (2020).
- [51]S. Huth et al., Nature 606, 276 (2022).
- [52]B. Biswas, Astrophys. J. 926, 75 (2022).
- [53]S. M. A. Imam, N. K. Patra, C. Mondal, T. Malik, and B. K. Agrawal, Phys. Rev. C 105, 015806 (2022).
- [54]L.-W. Chen, C. M. Ko, and B.-A. Li, Phys. Rev. C 72, 064309 (2005).
- [55]L.-W. Chen, B.-J. Cai, C. M. Ko, B.-A. Li, C. Shen, and J. Xu, Phys. Rev. C 80, 014322 (2009).
- [56]W. G. Newton, J. Hooker, M. Gearheart, K. Murphy, D.-H. Wen, F. J. Fattoyev, and B.-A. Li, Eur. Phys. J. A 50, 41 (2014).
- [57]J. Margueron, R. Hoffmann Casali, and F. Gulminelli, Phys. Rev. C 97, 025805 (2018).
- [58]J. Margueron and F. Gulminelli, Phys. Rev. C 99, 025806 (2019).
- [59]J. M. Lattimer and M. Prakash, Phys. Rept. 621, 127 (2016).
- [60]H. Gil, P. Papakonstantinou, C. H. Hyun, T.-S. Park, and Y. Oh, Acta Phys. Polon. B 48, 305 (2017).
- [61]A. Gelman, J. B. Carlin, H. S. Stern, D. B. Dunson, A. Vehtari, D. B. Rubin, J. Carlin, H. Stern, D. Rubin, and D. Dunson, Bayesian Data Analysis Third edition (CRC Press, Boca Raton, Florida, 2013).
- [62]J. Buchner, A. Georgakakis, K. Nandra, L. Hsu, C. Rangel, M. Brightman, A. Merloni, M. Salvato, J. Donley, and D. Kocевski, Astron. Astrophys. 564, A125 (2014).
- [63]K. Hebeler, J. M. Lattimer, C. J. Pethick, and A. Schwenk, Astrophys. J. 773, 11 (2013).
- [64]E. Chabanat, P. Bonche, P. Haensel, J. Meyer, and R. Schaeffer, Nucl. Phys. A 635, 231 (1998).
- [65]E. Chabanat, P. Bonche, P. Haensel, J. Meyer, and R. Schaeffer, Nucl. Phys. A 627, 710 (1997).
- [66]C. Mondal, B. K. Agrawal, and J. N. De, Phys. Rev. C 92, 024302 (2015).
- [67]C. Mondal, B. K. Agrawal, J. N. De, and S. K. Samaddar, Phys. Rev. C 93, 044328 (2016).
- [68]A. Sulaksono, T. J. Buervenich, P. G. Reinhard, and J. A. Maruhn, Phys. Rev. C 79, 044306 (2009).
- [69]U. Garg and G. Colò, Prog. Part. Nucl. Phys. 101, 55 (2018).
- [70]B. K. Agrawal, S. Shlomo, and V. K. Au, Phys. Rev. C 72, 014310 (2005).
- [71]B. T. Reed, F. J. Fattoyev, C. J. Horowitz, and J. Piekarewicz, Phys. Rev. Lett. 126, 172503 (2021).
- [72]X. Roca-Maza, X. Viñas, M. Centelles, B. K. Agrawal, G. Colò, N. Paar, J. Piekarewicz, and D. Vretenar, Phys. Rev. C 92, 064304 (2015).
- [73]R. Essick, I. Tews, P. Landry, and A. Schwenk, Phys. Rev. Lett. 127, 192701 (2021).
- [74]M. Ferreira and C. Providência, Phys. Rev. D 104, 063006 (2021).
- [75]M. Dutra, O. Lourenço, J. S. Sa Martins, A. Delfino, J. R. Stone, and P. D. Stevenson, Phys. Rev. C 85, 035201 (2012).
- [76]M. Dutra, O. Lourenço, S. S. Avancini, B. V. Carlson, A. Delfino, D. P. Menezes, C. Providência, S. Typel, and J. R. Stone, Phys. Rev. C 90, 055203 (2014).
- [77]C. Mondal, B. K. Agrawal, J. N. De, S. K. Samaddar, M. Centelles, and X. Viñas, Phys. Rev. C 96, 021302 (2017).
- [78]C. Drischler, J. W. Holt, and C. Wellenhofer, Ann. Rev. Nucl. Part. Sci. 71, 403 (2021).
- [79]A. Ekström, G. R. Jansen, K. A. Wendt, G. Hagen, T. Papenbrock, B. D. Carlsson, C. Forssén, M. Hjorth-Jensen, P. Navrátil, and W. Nazarewicz, Phys. Rev. C 91, 051301 (2015).
- [80]Y. Lim and J. W. Holt, Eur. Phys. J. A 55, 209 (2019).
- [81]T. Malik, M. Ferreira, B. K. Agrawal, and C. Providência, Astrophys. J. 930, 17 (2022).
- [82]S. Ghosh, B. K. Pradhan, D. Chatterjee, and J. Schaffner-Bielich, Front. Astron. Space Sci. 9, 864294 (2022).
- [83]N. K. Glendenning, Phys. Rev. D 46, 4161 (1992).
- [84]D. Adhikari et al. (PREX), Phys. Rev. Lett. 126, 172502 (2021).
- [85]J. R. Oppenheimer and G. M. Volkoff, Phys. Rev. 55, 374 (1939).
- [86]R. C. Tolman, Phys. Rev. 55, 364 (1939).
- [87]G. Baym, C. Pethick, and P. Sutherland, Astrophys. J. 170, 299 (1971).
- [88]J. Carriere, C. J. Horowitz, and J. Piekarewicz, Astrophys. J. 593, 463 (2003).
- [89]M. Fortin, C. Providencia, A. R. Raduta, F. Gulminelli, J. L. Zdunik, P. Haensel, and M. Bejger, Phys. Rev. C 94, 035804 (2016).
- [90]J. Piekarewicz and F. J. Fattoyev, Phys. Rev. C 99, 045802 (2019).
- [91]B. P. Abbott et al., Phys. Rev. Lett. 121, 161101 (2018).
- [92]B. P. Abbott et al., Phys. Rev. X 9, 011001 (2019).
- [93]R. Abbott et al., SoftwareX 13, 100658 (2021).
- [94]T. E. Riley et al., Astrophys. J. Lett. 887, L21 (2019).
- [95]M. C. Miller et al., Astrophys. J. Lett. 887, L24 (2019).
- [96]T. E. Riley et al., Astrophys. J. Lett. 918, L27 (2021).
- [97]M. C. Miller et al., Astrophys. J. Lett. 918, L28 (2021).
- [98]W. R. Romani, D. Kandel, A. V. Filippenko, T. G. Brink, and W. Zheng, Astrophys. J. Lett. 908, L46 (2021).
- [99]L. Rezzolla, E. R. Most, and L. R. Weih, Astrophys. J. Lett. 852, L25 (2018).
- [100]R. Abbott et al. (LIGO Scientific, Virgo), Astrophys. J. Lett. 896, L44 (2020).
- [101]A. Tsokaros, M. Ruiz, and S. L. Shapiro, Astrophys. J. 905, 48 (2020).
- [102]Y. Lim, A. Bhattacharya, J. W. Holt, and D. Pati, Phys. Rev. C 104, L032802 (2021).
- [103]C. Drischler, S. Han, J. M. Lattimer, M. Prakash, S. Reddy, and T. Zhao, Phys. Rev. C 103, 045808 (2021).
- [104]A. Li, Z. Miao, S. Han, and B. Zhang, Astrophys. J. 913, 27 (2021).
- [105]M. Ferreira, M. Fortin, T. Malik, B. K. Agrawal, and C. Providência, Phys. Rev. D 101, 043201 (2020).
- [106]I. M. Romero-Shaw et al., Mon. Not. Roy. Astron. Soc. 499, 3295 (2020).