Search

Large uncertainties remain in the estimation of aerosol direct radiative effect (DRE) and forcing (DRF). In this work, using an aerosol-climate model with two- and four-stream radiation schemes, we show that the radiative transfer algorithms contribute to the uncertainties. Aerosol shortwave DREs and heating rate are underestimated significantly by the two-stream algorithm. For present-day conditions, the four-stream algorithms are found to enhance global annualmean aerosol shortwave DREs by more than 8% (14%) at the top of the atmosphere (TOA), 15% (18%) in the atmosphere, and 12% (15%) at the surface for all-sky (clear-sky) case. The regional-averaged relative differences in aerosol shortwave DREs between the two- and four-stream algorithms increase as latitude increases, exceeding 25% at the TOA and 30% at the surface in the high latitudes of the Southern Hemisphere. The DRE differences due to the four-stream algorithms are negative, except for the Arctic, Tibetan Plateau, Arabia, and Sahara, at the TOA, are positive in the atmosphere, and are negative at the surface, with the maximum exceeding 4.0Wm-2. Increases in aerosol shortwave heating rates due to the four-stream algorithms are generally more than 10% and may even exceed 100%. Our results also show that the two-stream algorithm underestimates the DRFs due to anthropogenic aerosols. Significant underestimation appears in the middle latitudes of the Northern Hemisphere, with the maximum being close to the quantity of 0.6Wm-2 for clear-sky case. This study indicates that a multi-stream radiative transfer algorithm is necessary to reduce the uncertainties of aerosol DREs and DRFs estimated by global climate models. [ABSTRACT FROM AUTHOR]

The Gaussian–Legendre, Gaussian–Lobatto, Gaussian–Chebyshev and Gaussian quadrature (GQ) with different moment powers have been investigated by applying them into the four-stream solar and infrared radiative transfer algorithms. For solar radiative transfer, the Gaussian–Chebyshev and GQ with moment power m =0 show relatively accurate results compared to other types of 2GQ in a single-layer scattering medium. In a real atmospheric profile including gaseous transmission, Gaussian–Chebyshev and GQ with moment power m =0 are comparable in accuracy for cloud heating rate. GQ with moment power m =0 produces more accurate results in the upward flux at the top of the atmosphere, while Gaussian–Chebyshev produces more accurate results in the downward flux at the surface. These results have been confirmed in evaluations by using satellite observation data. For infrared radiative transfer, the GQ with moment powers m = 0 , 2 , 4 show relatively accurate results in effective emissivity for a single-layer scattering medium. In a real atmospheric profile, the GQ with moment powers m =0 and m =2 show superior accuracy in heating rate and flux. In addition, the evaluations using satellite observation data also show that the accuracy of GQ with moment powers m =0 and m =2 is comparable. Both the schemes are the best candidates for the four-stream radiation algorithms. [ABSTRACT FROM AUTHOR]