â¦still difficult to model even with the best models⦠(i.e. mixing, cloud ... domain but also: decreased cloud size & fraction with increasing aerosol concentration ...
Shortwave Radiative Impacts from Aerosol Effects on Shallow Marine Cumuli
Paquita Zuidema, Huiwen Xue, and Graham Feingold RSMAS/Miami
NOAA/ESRL, Boulder CO
MOTIVATION • boundary-layer Cumulus ubiquitous • radiatively important • difficult to characterize well from satellite 1) • …still difficult to model even with the best models… (i.e. mixing, cloud fraction issues) • AIE studies show conflicted results: Kaufman et al., 2005; Norris 2001: cloud fraction, LWP increases w/ aerosol McFarquhar et al., 2004; Matheson et al., 2005; Xue and Feingold, 2006: cloud fraction, LWP can decrease w/ aerosol
net radiative effect not always reported (except in IPCC)
MOTIVATION • boundary-layer Cumulus ubiquitous • radiatively important • difficult to characterize well from satellite 1) • …still difficult to model even with the best models… (i.e. mixing, cloud fraction issues) • AIE studies show conflicted results: Kaufman et al., 2005; Norris 2001: cloud fraction, LWP increases w/ aerosol McFarquhar et al., 2004; Matheson et al., 2005; Xue and Feingold, 2006: cloud fraction, LWP can decrease w/ aerosol
but net radiative effect not always reported 2)
..might it be possible to measure/evaluate cloud susceptibility S instead ?
S = δ(cloud albedo; Acld)/δ(CDNC) (Twomey, 1977) ~ Acld(1-Acld)/(3*CDNC) (Platnick & Twomey, 1994); conservative scattering useful for remote sensing & arguably climate models
MOTIVATION • boundary-layer Cumulus ubiquitous • radiatively important • difficult to characterize well from satellite 1) • …still difficult to model even with the best models… (i.e. mixing, cloud fraction issues) • AIE studies show conflicted results: Kaufman et al., 2005; Norris 2001: cloud fraction, LWP increases w/ aerosol McFarquhar et al., 2004; Matheson et al., 2005; Xue and Feingold, 2006: cloud fraction, LWP can decrease w/ aerosol
but net radiative effect not always reported 2)
..might it be possible to measure/evaluate cloud susceptibility S instead ?
S = δ(cloud albedo; Acld)/δ(CDNC) (Twomey, 1977) ~ Acld(1-Acld)/(3*CDNC) (Platnick & Twomey, 1994); conservative scattering useful for remote sensing & arguably climate models
Large-Eddy Simulations of Trade Wind Cumuli: Investigation of Aerosol Indirect Effects Huiwen Xue and Graham Feingold NOAA/ESRL, Boulder, Colorado “The effects of aerosol on warm trade cumulus clouds are investigated using a large-eddy simulation with size-resolved cloud microphysics. As expected, increases in aerosols cause a reduction in precipitation and an increase in the cloud-averaged liquid water path. However, for the case under study, cloud fraction, cloud size, cloud-top height, and depth decrease in response to increasing aerosol concentration, contrary to accepted hypotheses associated with the second aerosol indirect effect. The complex responses are determined by Competing effects of precipitation and droplet evaporation associated with entrainment…”
JAS, June 2006, p. 1605-
Large-Eddy Simulations of Trade Wind Cumuli: Investigation of Aerosol Indirect Effects Huiwen Xue and Graham Feingold NOAA/ESRL, Boulder, Colorado “The effects of aerosol on warm trade cumulus clouds are investigated using a large-eddy simulation with size-resolved cloud microphysics. As expected, increases in aerosols cause a reduction in precipitation and an increase in the cloud-averaged liquid water path. However, for the case under study, cloud fraction, cloud size, cloud-top height, and depth decrease in response to increasing aerosol concentration, contrary to accepted hypotheses associated with the second aerosol indirect effect. The complex responses are determined by competing effects of precipitation and droplet evaporation associated with entrainment…”
JAS, June 2006, p. 1605• size-resolved cloud µphysics -> careful calculation of radiative parameters τ, ω,P(Θ) • interesting cloud physical effects -> not universal. BOMEX-specific.
Summary of XF06 results
cloud
τ > 0.001
cloud LWP > 10 g m-2
domain
first (Twomey) AIE
suppression of precipitation AIE
domain
but also: decreased cloud size & fraction with increasing aerosol concentration, explained as more efficient evaporation&cooling-induced entrainment
Summary of XF06 results
cloud
τ > 0.001
cloud LWP > 10 g m-2
domain
domain
first (Twomey) AIE
for this case:
suppression of precipitation AIE
but also: decreased cloud size & fraction with increasing aerosol concentration, explained as more efficient evaporation&cooling-induced entrainment
• how does the net shortwave cloud radiative effect vary with Naer ? • is a 3D radiative transfer calculation necessary ? • even if 3DRT effects are significant for the domain-albedo, can the susceptibility (δalbedo/δCDNC) be approximated ?
LES model: (6.4 km)2 domain;Δx=100m; Δz=40m 48 realizations at each of 5 Naer values ~ dynamic variability RT model: Community Monte Carlo (I3RC project; Cahalan et al., ‘05) 3D & ICA calcs; 2 sun angles; many photons; aerosols+clouds; monochromatic calc @ 0.64 µm cloud radiative representation? • accurate τ w/ Qext=2 • ω =1 • P(Θ): Mie, reff =10
LES model: (6.4 km)2 domain;Δx=100m; Δz=40m 48 realizations at each of 5 Naer values ~ dynamic variability RT model: Community Monte Carlo (I3RC project; Cahalan et al., ‘05) 3D & ICA calcs; 2 sun angles; many photons; aerosols+clouds; monochromatic calc @ 0.64 µm overhead Sun
cloud radiative representation? • accurate τ w/ Qext=2 • ω =1 • P(Θ): Mie, reff =10
Naer = 25-2000 /mg
solar zenith angle=600
ICA MC
aerosol DRF
Focusing on AIE only: overhead Sun
Sun angle = 600
ICA 3D 3D
ICA
Focusing on AIE only: overhead Sun
Sun angle = 600
ICA 3D 3D
ICA
Yes, 3DRT effects matter for the domain albedo
Dotted line: mean cloud fraction (23%) * Fcld the impact of cloud fraction changes (due to evaporationentrainment) reduces radiative impact by ~ 1/3 - 1/2 could help explain some of obs-model AIE RT
cloud susceptibility
δAcloud/δNcdnc ~ Acld(1-Acld)/(3* δNcdnc) S
So
Select scenes w/ LWP between 30-40 g m-2 overhead Sun
So: dotted
ICA MC
So ~ S ! even with 3DRT effects built in
Sun angle = 60
Area-weighted cloud susceptibility S: δAdomain/δNcdnc =CF* δAcloud/δNcdnc + Acloud* δCF/ δNcdnc Mean area-weighted S
1)
1) net 2)
S changes due to 2) cloud fraction changes
Area-weighted cloud susceptibility S: δAdomain/δNcdnc =CF* δAcloud/δNcdnc + Acloud* δCF/ δNcdnc Mean area-weighted S
1)
At low Naer fractional cloud changes dominate area-weighted S, at higher Naer dominated by cloud S
1) net 2)
S changes due to 2) cloud fraction changes
SUMMARY: for this case • how does the net shortwave cloud radiative effect vary with Naer ? First AIE (Twomey effect) dominates albedo response to aerosol increases, despite coincident reductions in cloud fraction, size, and depth (contrasts to McFarquhar et al.,2004)
•is a 3D radiative transfer calculation necessary ? YES. 3D RT effects significant, reduce indirect RF to 60-90% of ICA contrasts to some observational studies • even if 3DRT effects are significant for the domain-albedo, can the susceptibility (δalbedo/δCDNC) be approximated ? YES. cloud susceptibility less sensitive to 3D effects, well approximated by So (reassuring for satellite applications) area-weighted cloud susceptibility dominated by cloud fractional changes at low aerosol concentrations, cloud susceptibility at high aerosol conc.
currently ATEX model output being similarly examined; different relationships expected