Why use MSG?
Synthesizing stellar spectra from first principles is a complicated endeavor, requiring a detailed understanding of radiative transfer and atomic physics, together with significant computational resources. Therefore, in most circumstances its better to use one of the many pre-calculated grids of spectra published in the astrophysical literature (see, e.g., Lanz & Hubeny, 2003; Lanz & Hubeny, 2007; Kirby, 2011; de Laverny et al., 2012; Husser et al., 2013; Allende Prieto et al., 2018; Chiavassa et al., 2018; Zsargó et al., 2020). However, even with these grids a significant obstacle remains: when photospheric parameters fall between the grid nodes, some kind of interpolation is necessary in order to evaluate a spectrum.
MSG is designed to solve this problem. It’s not the first software package that offers stellar spectral interpolation (see, e.g., FERRE, Starfish and stsynphot); however, with spectral interpolation as its sole focus, it offers a combination of features unmatched by other packages:
scalability — MSG handles grids that are much larger (on disk) than available computer memory.
extensibility — MSG handles grids with an arbitrary number of dimensions.
portability — MSG is platform-agnostic and provide APIs for the programming languages (Fortran, C, Python) most commonly used in Astronomy.
performance — MSG provides smooth and accurate interpolates with minimal computational cost.
robustness — MSG gracefully handles missing data caused by holes and/or ragged boundaries in the grid.
Together, these features mean that MSG is flexible and powerful while remaining straightforward to use: it’s the perfect condiment to add flavor to your science!.
The source code for MSG is hosted in the rhdtownsend/msg git repository on GitHub. Instructions for downloading and installing the software can be found in the Quick Start chapter.
If you use MSG in your research, please cite the following papers:
Townsend Rich, Lopez Aaron, 2022, Journal of Open-Source Software, 8(81), 4602, https://doi.org/10.21105/joss.04
Be sure to also cite the source of the grid data that you’re using with MSG. For instance, if you’re working with one of the CAP18 grids, you should cite Allende Prieto et al. (2018).
MSG remains under active development by the following team:
Rich Townsend (University of Wisconsin-Madison); project leader
MSG has been developed with financial support from the following grants:
NSF awards ACI-1663696 and AST-1716436;
NASA award 80NSSC20K0515.