lowtran VS LBLRTM

For your atmospheric transmission estimates have you considered using LOWTRAN? https://github.com/space-physics/lowtran

The slow-and accurate way is to run a line-by-line code like the Line By Line Ratiative Transfer Model (LBLRTM). The code has to do repeated calculations for the M absoprtion lines in your band. The number of lines depend on the chemical makeup of the atmosphere, and some of the physical quantities like pressure and temperature, and how wide your band is. A source I often use for the absorption line data (and a good primer on the calculation) is the High-resolution transmission (HITRAN) database. To figure out how much energy the atmosphere absorbs, you need to calculate the absorption of each isotope at each frequency. For accurate results, your frequencies should be spaced about 1/10 the width of the absorption lines, which means that the number of frequencies is also proportional to your bandwidth.