Prospects for measurement of far infrared tropospheric spectra: Implications for climate modeling

A substantial fraction of the Earth's thermal radiation is emitted in far-infrared wavelengths between 15 and 100 μ m. Some of the strongest radiative interactions with water vapor occur in the far-infrared rotation bands. Approximately 60--70% of the natural greenhouse effect and 70--80% of the radiative feedbacks from doubling carbon dioxide are due to water vapor. Despite its importance for the climate system, this part of the spectrum was last measured from space-born instruments over 25 years ago. By contrast, several spectral intervals in the mid-infrared are routinely measured for operational temperature and humidity soundings. Several general circulation models (GCMs) can now simulate the mid-infrared spectrum to test the fidelity of climate simulations. Using a new instrument under development by NASA, it will be possible to extend these tests to longer wavelengths. The Far Infrared Spectroscopy of the Troposphere (FIRST) satellite instrument will provide accurate spectra between 10 and 100 μ m for clear and cloudy regions. We illustrate why these measurements are critical for improving GCMs using the NCAR Community Atmosphere Model (CAM). We have changed the infrared absorption and emission by water vapor to bring CAM into close agreement with modern line-by-line calculations. The main changes in the radiative heating rates are associated with line and continuum absorption in the rotation band. These changes interact strongly with the convective parameterizations in CAM, leading to significant changes in the simulated thermodynamic state. The significant changes to the mean climate and climate sensitivity are discussed in light of current uncertainties in the foreign component of the water vapor continuum. Far-infrared top-of-atmosphere fluxes are simulated with CAM and compared to the outgoing longwave radiation. These calculations illustrate the utility of FIRST data for evaluating the fidelity of radiative processes in climate simulations.