Impact of Multiple Scattering on Longwave Radiative Transfer Involving Clouds

Abstract

General circulation models (GCMs) are extensively used to estimate the influence of clouds on the global energy budget and other aspects of climate. Because radiative transfer computations involved in GCMs are costly, it is typical to consider only absorption but not scattering by clouds in longwave (LW) spectral bands. In this study, the flux and heating rate biases due to neglecting the scattering of LW radiation by clouds are quantified by using advanced cloud optical property models, and satellite data from Cloud‐Aerosol Lidar and Infrared Pathfinder Satellite Observation (CALIPSO), CloudSat, Clouds and the Earth's Radiant Energy System (CERES), and Moderate Resolution Imaging Spectrometer (MODIS) merged products (CCCM). From the products, information about the atmosphere and clouds (microphysical and buck optical properties, and top and base heights) is used to simulate fluxes and heating rates. One‐year global simulations for 2010 show that the LW scattering decreases top‐of‐atmosphere (TOA) upward flux and increases surface downward flux by 2.6 and 1.2 W/m2, respectively, or approximately 10% and 5% of the TOA and surface LW cloud radiative effect, respectively. Regional TOA upward flux biases are as much as 5% of global averaged outgoing longwave radiation (OLR). LW scattering causes approximately 0.018 K/d cooling at the tropopause and about 0.028 K/d heating at the surface. Furthermore, over 40% of the total OLR bias for ice clouds is observed in 350–500 cm−1. Overall, the radiative effects associated with neglecting LW scattering are comparable to the counterpart due to doubling atmospheric CO2 under clear‐sky conditions.

Cloud Longwave Scattering Effect and Its Impact on Climate Simulation

“…The results showed a weak impact on downward flux at the surface as well. The spectral results of the cloud LW scattering effect on flux in Figure 2 are consistent with the band-by-band LW flux biases by neglecting scattering in Kuo et al [18].…”

“…In our study, the cumulative LW scattering effect on OLR is 2-6 W m −2 , and it is comparable with 4-8 W m −2 in Chou et al [6], 4 W m −2 in Li [7] and Li and Fu [10], and 6-8 W m −2 in Joseph and Min [17]. Kuo et al [18] demonstrated that the LW scattering in the far-infrared and atmospheric window regions (10-1390 cm −2 ) mainly contribute to the biases in spectral fluxes, which is consistent with our results in Figure 2.…”

“…Because LW radiance is less anisotropic compared to that of solar radiation, a radiative transfer scheme with a small number of streams is capable of achieving reasonably accurate results. Kuo et al [18] have shown that the difference in fluxes between 16-stream and more precise 128-stream DISORT is within 1%. We also checked that the difference between 8-stream and 16-stream DISORT is less than 0.3% (not shown).…”

“…On the other hand, SSA of ice cloud varies with ice crystal size and wave band [41]. The largest SSA (i.e., weakest absorption) occurs at band 2 (20-28.57 µm) [18]. For asymmetry factor, both water and ice particles with larger size generally have a larger asymmetry factor [41,46,48] indicating that larger particles have stronger forward scattering ability for both water and ice clouds.…”

“…Costa and Shine [8] showed that the LW scattering effect by low clouds could reduce the estimated OLR by 0.9 W m −2 and the reduction in OLR is 3 W m −2 over mid-latitudes and tropics for all clouds. In addition to the broadband results, some studies also specified the spectral ranges, in which the cloud LW scattering effect dominates, such as the atmospheric window range (8-12 µm) [8] and far infrared range at around 25 µm [8,9,13,18].…”