Infrared collision-induced and far-line absorption in dense CO2 atmospheres

Collision-induced absorption is of great importance to the overall radiative budget in dense CO2-rich atmospheres, but its representation in climate models remains uncertain, mainly due to a lack of accurate experimental and theoretical data. Here we compare several parameterisations of the effect, including a new one that makes use of previously unused measurements in the 1200–1800 cm−1 spectral range. We find that a widely used parameterisation strongly overestimates absorption in pure CO2 atmospheres compared to later results, and propose a new approach that we believe is the most accurate possible given currently available data.