2.2. Stellar Evolution Models
Construction of composite SEDs from individual stellar spectra requires prescriptions for the appropriate weighting by the stellar temperatures, luminosities, and lifetimes. Stellar evolution models provide such prescriptions. Excellent reviews on the subject can be found in  and . As the most luminous stars dominate the composite SED, it is primarily massive-star evolution that is relevant. Stellar evolution in the upper part of the HRD mainly depends on the chemical composition at birth, the stellar mass and its decrease with time due to mass loss, and mixing processes induced by rotation, convection, and overshooting.
The main focus of massive-star evolution models has shifted from core overshooting in the 1980s, over mass loss in the 1990s, to rotation in the new millennium. These processes have one theme in common: they determine (among others) the relative importance of the convective vs. the radiative energy transport. The higher efficiency of convection has many effects on stellar evolution, most notably on the ratio of lifetimes spent in the core hydrogen-burning stage and in the shell hydrogen-burning stage. The impacts of such effects on the composite SED have been discussed by . These processes are effective only for stars massive enough to develop convective cores, therefore they are visible only in young and/or massive populations, which then display bluer spectra.
Evolution models with rotation are just beginning to become available (). The new models demonstrate the significance of rotation: The additional helium brought near the H-burning shell by rotational mixing and the larger He-core both lead to a less efficient H-burning shell and a smaller associated convective zone. Therefore, the stellar radius of rotating stars will inflate during the He-burning phase. A pilot computation for a 20 M star is reproduced in Fig. 4. Rotation can become the principal parameter for post-main sequence evolution if stellar rotation velocities on the main sequence are above ~ 200 km s-1, values not unrealistically high.
Figure 4. Evolutionary models for a 20 M star: solid, dashed, dotted-dashed, and dotted lines correspond to rotation velocities of 0, 100, 200, and 300 km s-1, respectively ().
The largest uncertainty is the unknown behavior of stellar rotation at very low metallicity. Circumstantial evidence for higher stellar rotation velocities in the Magellanic Clouds was presented by . Should this trend continue to extremely low metallicities, rotationally induced convection processes could be very important, e.g., in Population III stars. If so, metals could be transported to the surfaces of the first generations of stars, inducing stronger winds, higher opacity, and a significantly altered extreme UV radiation field. The consequences for the chances of detection of these stars would be quite significant.
The uncertainties in stellar evolution modeling primarily affect highly evolved stars, either in the red (red supergiants [RSG]) or blue (Wolf-Rayet [W-R] stars) part of the HRD. On the other hand, the vast majority of the stellar population contributing to the composite SED is fairly well understood and reproduced with state-of-the-art stellar evolution models.