Global warming potential values are based on the atmospheric lifetime of chemical species released into the troposphere and on the radiative forcing values of those chemical species. The atmospheric lifetime is evaluated using the chemical degradation rate by hydroxyl radical abstraction of hydrogen atoms from the released parent species for compounds with hydrogen atoms. Radiative forcing, the energy absorbed by the chemical per meter square of troposphere parallel to the Earth's surface per part per billion of concentration, is typically obtained through experimentally measured absorption cross sections followed by atmospheric modeling to obtain results. Pinnock, et al., found a short-cut of estimating radiative forcing through binning the percent transmittance of the Earth's atmosphere into 10 cm-1 bins and then aggregating absorption cross section data. Papasavva, et al., transferred this work for use with computational chemistry based predictions of IR spectra. Our earlier work build upon their approach and extensively showed relatively good agreement with experimentally based modeled values for radiative forcing for a large number of hydrofluoroethers.

One of the criticisms of our and Papasavva's work is that all of the peak intensity for a given frequency was assigned to only one 10 cm-1 bin, ignoring the fact that experimental peaks are distributed about peak heights. In this work, we explore the use of Lorentzian distributions of intensities about peak locations using scaled frequencies from the B3LYP/6-31g* level of theory. Extensive work in our lab has shown that this level of theory reproduces frequency locations extremely well and intensities are accurate compared to the limited data available for comparison.

We compute radiative forcing values for close to 100 chemical species using the Lorentzian distributed peaks in order to improve our earlier predictions and expand on a reproducible methodology for accurate radiative forcing predictions. We find significant reduction of error compared to experimentally modeled results versus our earlier work. This approach provides for a robust method of quickly building databases of radiative forcing results that can be combined with kinetic degradation rates to predict global warming potentials.