Compared to the Castelli & Kurucz (2003; hereafter CK03) grid, KOALA adopts a revised solar chemical composition, extends the parameter coverage in both metallicity and [α/Fe], and provides finer effective temperature sampling for cool stars, resulting in a denser and more versatile model library.
The main improvements are:
Updated solar composition: The reference mixture adopts the abundances by Caffau et al. (2011) and Lodders (2010), replacing the older Grevesse & Sauval (1998) set used in CK03. A helium mass fraction of Y = 0.2476, consistent with Planck results, is assumed for all ODFs.
Extended chemical mixtures: KOALA includes five values of [α/Fe] (–0.4, –0.2, 0.0, +0.2, +0.4), compared to two values in CK03.
Broader metallicity coverage: The grid spans 18 metallicities from [M/H] = –5.0 to +0.5 dex, with steps of 0.5 dex at very low metallicities and finer 0.25 dex sampling from –2.5 to +0.5 dex. This extended coverage ensures a more complete and homogeneous parameter space than CK03, which lacked regular metallicity sampling at high [α/Fe].
Improved temperature sampling: For cooler stars (Teff ≤ 7000 K), the grid adopts steps of 125 K, twice as fine as in CK03. This refinement improves the accuracy of interpolated colours and often removes the need for computing additional tailored models. For higher Teff, the CK03 sampling (250–1000 K steps) is maintained, ensuring broad applicability across the HR diagram up to 50,000 K.
Consistent coverage in gravity and temperature space: The grid uses log g steps of 0.5 dex, ensuring regular coverage across different evolutionary phases. Combined with the finer Teff sampling, KOALA provides a dense and homogeneous distribution of models in the Teff–log g plane, making it suitable for comparison with other atmosphere libraries such as MARCS, PHOENIX, and TLUSTY.
The KOALA database is organized by five [α/Fe] values (+0.4, +0.2, 0.0, –0.2, –0.4). For each [α/Fe] value, the database provides a complete set of products including model atmospheres and emergent fluxes computed with a microturbulent velocity of 2 km/s, as well as Big ODF and Rosseland opacity tables for vturb = 0, 1, 2, 4, and 8 km/s. Additionally, Little ODF tables are available for the same set of microturbulent velocities. The database also includes theoretical magnitudes (vturb = 2 km/s) and a dedicated archive for K2.
The table files are named to encode the stellar parameters and chemical composition. For example:
ap050k2_cl_ap04.dat → [M/H] = +0.50, [α/Fe] = +0.4
am100k2_cl_ap04.dat → [M/H] = –1.00, [α/Fe] = +0.4
Here, “ap” or “am” indicates the sign of the metallicity (+ or –), the numeric value gives the absolute metallicity, “cl” refers to the Caffau & Lodders solar mixture, and the final part encodes the [α/Fe] enhancement.
This ATLAS9 model specifies a stellar atmosphere e.g. with TEFF = 3750 K, log(g) = 0.0, assuming Local Thermodynamic Equilibrium (LTE). The TITLE line indicates the chemical mixture, e.g. [α/Fe] = +0.4 enhancement based on the Caffau & Lodders solar abundances.
OPACITY and CONVECTION parameters define the radiative transfer setup and mixing-length treatment.
ABUNDANCE CHANGE entries list individual element abundances relative to hydrogen, scaled according to the model metallicity.
The table contains the atmospheric structure: mass depth (RHOX), temperature (T), gas pressure (P), electron density (XNE), Rosseland opacity (ABROSS), radiative acceleration (ACCRAD), microturbulence (VTURB), convective flux (FLXCNV), convective velocity (VCONV), and sound speed (VELSND) at each depth point.