Stellar Spectral Synthesis with OpenGL

Type Conference Paper
Names Nicholas R. Hill, R. Townsend
Conference Name Bulletin of the American Astronomical Society, #217, #154.02
Volume 217
Pages 15402
Date January 1, 2011
Library Catalog NASA ADS
Abstract Given an appropriate model atmosphere, synthesizing the spectrum of a star is a relatively straightforward task -- *if* the star is spherical and homogeneous across its surface. Many astronomically interesting objects do not, however, fall into this category. Examples include single stars that are spotted, rapidly rotating or pulsating, and binary stars in eclipsing or ellipsoidal-variable configurations. To synthesize a spectrum in such cases, it is necessary to construct a 3-D model of the stellar surface; determine which regions of the surface are visible to an external observer; and then calculate the observer-directed radiation from these regions. The Open Graphics Library (OpenGL), a cross-platform application programming interface for creation of 2-D and 3-D graphics, already includes much of the functionality required to implement these steps. We describe a new approach to stellar spectral synthesis that leverages this functionality. A 3-D mesh is constructed to represent the (possibly non-spherical) geometry of the stellar surface (or surfaces, in the case of binary or multiple systems). Textures are laid over this mesh to represent the run of physical attributes such as temperature, gravity, velocity, etc. The textured mesh is then rendered by OpenGL into a framebuffer, a step which naturally takes care of projection and occultation effects. The attributes of each framebuffer pixel are used to look up an appropriate spectrum in pre-calculated tables of specific intensities; and finally, summing the spectra from all pixels gives the disk-integrated synthetic flux spectrum of the star. The advantage of this approach lies in its efficiency (many OpenGL features are hardware-implemented), flexibility and manifest simplicity. Possible applications include binary light-curve modeling, mode identification in pulsating stars, and stellar population synthesis.
UW-Madison Astronomy Home