The young Sun was not as luminous as it is today, implying that the atmospheres of Earth and Mars had to be more optically thick, if we seek to explain evidence for the existence of liquid water on both these planets. Several modeling studies have attempted to constrain the abundance of CO₂ and other greenhouse gases that may have been necessary to explain the apparent warmth. However, accepted parameterizations of absorption by CO₂ yield differences in radiative forcing of up to ~40 Watts per square meter, when applied to CO₂ abundances relevant to early planetary climate [Halevy et al., J. Geophys. Res., 2009]. Figure 2 below reveals this to be due to differences in absorption between these parameterizations mainly in the "window regions," where CO₂ does not absorb strongly. Whereas these difference do not matter much at low CO₂ abundances, they give rise to widely diverging results in deep-time paleoclimate studies, as seen in Figure 3.

Figure 2: Differences in the absorption coefficient calculated in a line-by-line radiative transfer model under different parameterizations of CO2 absorption (upper graph) and the resultant differences in radiative forcing (lower graph) for a 1-bar pure CO2 atmosphere on Mars. The black curve corresponds to a CO2 continuum calculated as in Kasting et al. (1984) and subsequent papers, the orange curve to a scaled foreign-broadened CO2 continuum similar to the benchmark line-by-line model LBLRTM and the green curve to a continuum calculated as in Meadows and Crisp (1996) and subsequent papers.

Figure 3: Differences in outgoing longwave radiation (OLR) between three parameterizations of absorption by CO2 in abundances relevant to deep-time paleoclimate studies. Whereas the increase in scattered shortwave radiation due to Rayleigh scattering with increasing pCO2, shown here as the planetary albedo, is identical under all three parameterizations, the decrease in OLR differs significantly. This results in differences of more than ten degrees Celsius in the equilibrium surface temperature.

To resolve these uncertainties in the greenhouse effect of CO₂ at high abundances, I have measured its absorption on a Fourier Transform Infrared (FTIR) spectrometer at the Canadian Light Source synchrotron, located on the University of Saskatchewan Campus in Saskatoon. These measurements, extending from 10 to 5000 cm^-1 and over the relevant range of pressure and temperature, yield a single parameterization for use in deep-time paleoclimate modeling studies [Halevy et al., in prep., 2011]. An example of the absorption coefficient calculated from these measurements between 30 and 500 cm^-1, and between 1100 and 1600 cm^-1 is shown in Figure 4 below.

Figure 4: Collision-induced absorption of CO2, measured by Fourier transform infrared spectrometry.