You could just assume constant temperature (using e.g. blackbody temperature based on solar radiation) and get a first order approximation of the behavior of the atmosphere. Won't be as good as the standard atmosphere model which uses empirical temperature gradients, but it will provide a ballpark model of the atmosphere.