The major focus of my research has been the theoretical investigation of the surface layers of stars, particularly those whose surface layers are cool enough to contain molecules. The study of these stars is important because they are generally in the late stages of their evolution. It is necessary to calculate models of the stellar atmospheres in order to be able to interpret observations to determine the composition, temperatures, radii, and luminosities of these stars. I have been primarily concerned with trying to improve the quality of these model atmospheres by making more accurate calculations of the physical processes that occur in them.
An important source of opacity in the cool atmospheres is the molecular spectral lines. I initially calculated the opacity due to hot water vapour 25 years ago by calculating the strengths and positions of the more than 2 million lines that make up this spectrum. I am currently involved in recalculating this spectrum with improved line positions and strengths.
Most models of cool atmospheres have assumed that the atmospheres are stationary, in hydrostatic equilibrium, and in local thermodynamic equilibrium. Observations, however, indicate that they are suffering mass loss. I am currently involved in developing hydrodynamic model atmospheres which include mass loss. I have also calculated model atmospheres where the assumption of local thermodynamic equilibrium has been replaced by the more accurate assumption of statistical equilibrium.
I have begun an investigation of the thermal stability of stellar atmospheres. Almost all of the model atmosphere calculations have assumed radiative equilibrium. Except for convection, very little work has been done on whether the atmospheres in radiative equilibrium are stable against thermal perturbations. The thermal timescale in the surface layers of an atmosphere is considerably shorter than the dynamical timescale. Therefore, a thermal runaway in the atmosphere would become very large before motions in the atmosphere would act to stop it. As a result, if the atmosphere is unstable against thermal perturbations, the actual state of the atmosphere may be quite different from that predicted with the assumption of radiative equilibrium.