Radiation & Climate: Major Projects
Effects of Improved Radiative Transfer Modeling for Climate Simulations
M. J. Iacono, E. J. Mlawer, and S. A. Clough
Atmospheric and Environmental Research, Inc.
The radiative interaction of shortwave and infrared energy in the atmosphere with clouds and greenhouse gases represent complex processes that contribute significantly to maintaining earth's climate system. For climate model simulations to become more precise, it is essential that these processes be modelled accurately as verified by direct comparisons with observations and with results from a validated line-by-line model. For this purpose, a rapid radiative transfer model (RRTM) has been developed which reproduces the computational accuracy of a more complex line-by-line radiative transfer model (LBLRTM) at the high speed necessary for its application within a general circulation model. RRTM employs a correlated-k method for the calculation of fluxes and cooling rates in the atmosphere. We have introduced RRTM into the NCAR Community Climate Model (CCM2) to establish its affect on short-term (less than one year) 3-D model calculations. Comparisons between RRTM and the CCM2 radiation column model for individual clear sky profiles indicate significant differences in fluxes and cooling rates for the profiles examined. At the standard CCM2 18-layer vertical resolution, RRTM produces a net flux that is 10 W/m2 lower at the top of the atmosphere, 12 W/m2 lower at the tropopause, 20 W/m2 lower in the middle troposphere, and 7 W/m2 lower at the surface relative to the CCM2 radiation code for the mid-latitude summer atmosphere. RRTM cooling rate is as much as 0.4 K/day higher near the 300 mb peak in water vapor continuum absorption and 0.4 K/day lower in the lower troposphere. Timing tests on a CRAY YMP show that RRTM provides a 50% improvement in cpu time over the CCM2 longwave model. In order to implement RRTM within the climate model, an algorithm for the absorption and emission of radiation in clouds has been included in RRTM which is consistent with the current CCM2 infrared cloud algorithm. Results of a single season climate simulation with both radiation models will also be shown.
Clear sky (a) longwave fluxes and (b) cooling rate as calculated by the CCM2 column radiation model for the mid-latitude summer atmosphere. The flux and cooling rate differences between CCM2 and the validated, rapid radiative transfer model (RRTM) are also shown.
Clear sky (a) longwave fluxes and (b) cooling rate as calculated by the CCM2 column radiation model for the mid-latitude winter atmosphere. The flux and cooling rate differences between CCM2 and the validated, rapid radiative transfer model (RRTM) are also shown.
Clear sky outgoing longwave radiation difference between CCM2
and RRTM for a three month simulation from Dec, 1986 to Feb, 1987.
RRTM produces an OLR that is as much as 11 W/m2 lower than CCM2
in the tropics and 7 W/m2 lower than CCM2 in the global average
for this season partly due to greater absorption from the water
vapor continuum model used in RRTM.
Total sky outgoing longwave radiation difference between CCM2
and RRTM for a three month simulation from Dec, 1986 to Feb, 1987.
The overall effect of clouds is to partially offset the clear
sky OLR differences. RRTM produces a total sky OLR that is up
to 5 W/m2 higher than CCM2 in the cloudiest areas just south of
the equator. Over less cloudy areas in the tropics the RRTM OLR
is generally 6-8 W/m2 lower than CCM2, while the global average
RRTM OLR is 3 W/m2 lower.
Information
Manager
Major Projects
Publications