Overview of kap module
The opacity values returned by the kap
module combine opacity
data from many sources. The most important opacity-related MESA
options select which opacity sources to use and control the location
of the blends between them. The total opacity is the appropriately
combined radiative opacity \(\kappa_{\rm rad}\) and the conductive
opacity \(\kappa_{\rm cond}\):
Radiative Opacities
The radiative opacity is the Rosseland mean opacity. The opacity depends on the temperature, density, and composition. A set of opacity tables consists of a collection of individual opacity tables, each at a different composition. Each individual table is tabulated in \(\log T\) and in \(\log R \equiv \log \rho - 3 \log T + 18\) (cgs).
For tables that assume a fixed metal distribution, the composition can be encoded via the variables \((X, Z)\). Tables assume a particular metal abundance pattern (usually scaled solar). Some opacity table sets include additional parameters in order to allow the metal abundance pattern (often the CNO abundances) to vary. Such variation naturally occurs in the stellar core during helium burning (and beyond) and in the stellar envelope as a result of dredge-up processes.
Note
The value of the option Zbase provides the reference metallicity necessary to calculate element variations (e.g., carbon and oxygen enhancement) from the composition of a cell. The default opacity configuration requires this value to be specified. Physically, this usually corresponds to the initial metallicity of the star.
In MESA, separate opacity table sets are used for high and low temperature. In the intermediate region, both opacities are evaluated and blended. The location of this blend is controlled with the options kap_blend_logT_upper_bdy and kap_blend_logT_lower_bdy.
High temperature \((T \gtrsim 10^4\,\rm K)\)
The OPAL tables (Iglesias & Rogers 1993, 1996) with fixed metal distributions are called Type 1 and cover the region \(0.0 \leq X \leq 1-Z\) and \(0.0\leq Z \leq 0.1\). The Opacity Project (OP; Seaton 2005) are also available. Type 1 tables from The Los Alamos OPLIB database (OPLIB; Colgan 2016) are also available, and cover the region \(0.0 \leq X \leq 1-Z\) and \(0.0\leq Z \leq 0.2\). The set of tables to be used are selected by the option kap_file_prefix.
A direct comparison between the Type 1 format of OPAL/OP tables and the OPLIB tables are shown in the figure below taken from Figure 1 in Farag et al. 2024. Further comparisons between OP/OPAL/OPLIB can be found in Farag et al. 2024.
Additionally, there is support for the OPAL Type 2 tables that allow for varying amounts of C and O beyond that accounted for by \(Z\); these are needed during helium burning and beyond. These have a range \(0.0 \leq X \leq 0.7\), \(0.0\leq Z\leq0.1\). The set of tables to be used are selected by the option kap_CO_prefix.
Type 2 tables on by default (see use_Type2_opacities) and The blends between these table sets occur based on hydrogen fraction (see kap_Type2_full_off_X and kap_Type2_full_on_X and ) and metal enhancement (controlled by kap_Type2_full_off_dZ and kap_Type2_full_on_dZ).
Low temperature \((T \lesssim 10^4\,\rm K)\)
Low temperature opacities are selected with the option kap_lowT_prefix.
Tables based on the work of Ferguson et al. (2005) include the effects of molecules and grains and cover the range \(2.7 \le \log T \le 4.5\) and \(-8 \le \log R \le 1\).
Tables based on the work of Freedman et al. (2008) include the effects of molecules and cover the range \(1.88 \le \log T \le 4.5\) and \(-8 \le \log R \le 9\). The table set was privately communicated by R. S. Freedman in 2011. Unlike other opacity sources, this is a 1D sequence of tables in \(Z\) as opposed to a 2D grid of \((X,Z)\) values. (The assumed H/He abundances scale with \(Z\).)
Tables from ÆSOPUS (Marigo & Aringer 2009) include variation factors for the CNO isotopes. The opacity is evaluated using the global value of \(Z_{\rm base}\) and the local (cell) values of \((X, X_{\rm C}, X_{\rm N}, X_{\rm O})\).
The ÆSOPUS tables are provided at a set of reference metalicites. In order to interpolate to the provided \(Z_{\rm base}\), the opacity is evaluated at an appropriate subset of these reference values (and then interpolated). For each such \(Z_{\rm ref}\), the ÆSOPUS composition parameters
are calculated, the opacities evaluated the tables with bracketing compositions, and the resulting opacities linearly interpolated. (Note that this means that the interpolation in \(Z\) occurs at fixed \(X\) and \(f_{\rm CO}\), but not at fixed \(f_{\rm C}\) or \(f_{\rm N}\).)
Compton Scattering
At sufficiently high temperature \((T \gtrsim 10^8\,\rm K)\), the opacity will be dominated by Compton scattering. MESA calculates the opacity of Compton scattering using the prescription of Poutanen (2017). Near the high-\(T\) and low-\(R\) edges of the high temperature opacity tables, MESA smoothly blends the tabulated opacity values with the Compton scattering values. The location of these blends is not user-controllable.
Conductive Opacities
The conductive opacity \((\kappa_{\rm cond})\) is given by the thermal conductivity \((K)\) appropriately recast such that the heat transfer equation resembles the form of the equation used in radiative diffusion (e.g., HKT Section 4.5). This implies
The thermal conductivities used in MESA are an extended version of the results of Cassisi et al. (2007) privately communicated by A.Y. Potekhin. They are tabulated for a set of \(1 \le \bar{Z} \le 60\). Each table spans \(-6 \le \log(\rho/\rm g\,cm^{-3}) \le 11.50\) and \(3 \le \log(T/\rm K) \le 10\).
For H and He in the regime of moderate coupling and moderate degeneracy, the additional correction formulae of Blouin et al. (2020) are applied.