Effect of aerosol on warm convective clouds:Aerosol-cloud-surface flux feedbacks in a new coupledlarge eddy modelHongli JiangCooperative Institute for Research in the Atmosphere/NOAA Earth System Research Laboratory, Boulder, Colorado, USAGraham FeingoldNOAA Earth System Research Laboratory, Boulder, Colorado, USAReceived 27 April 2005; revised 25 October 2005; accepted 8 November 2005; published 5 January 2006.[1] We present a new large eddy simulation model that comprises coupled componentsrepresenting size-resolved aerosol and cloud microphysics, radiative properties of aerosoland clouds, dynamics, and a surface soil and vegetation model. The model is used toinvestigate the effect of increases in aerosol on liquid water path LWP, cloud fraction,optical depth, and precipitation formation in warm, continental cumulus clouds. Sets ofsimulations that either neglect, or include the radiative properties of a partially absorbingaerosol are performed. In the absence of aerosol radiative effects, an increase in aerosolloading results in a reduction in precipitation. However, the clouds do not experiencesignificant changes in LWP, cloud fraction and cloud depth; aerosol effects on LWP andcloud fraction are small compared to the dynamical variability of the clouds at any givenaerosol concentration. Reasons for this response are discussed. When aerosol radiativeeffects are included, the modification in atmospheric heating profiles, and the reduction insurface latent and sensible heat fluxes resulting from the presence of these particles, have asignificant effect on cloud parameters and boundary layer evolution. For the caseconsidered, there is a significant reduction in the strength of convection, LWP, cloudfraction and cloud depth. Cloud optical depth responds non-monotonically to the increasein aerosol. These results indicate that in continental regions surface processes must beincluded in calculations of aerosol-cloud-precipitation interactions. Neglect of thesesurface processes may result in an overestimate of the second aerosol indirecteffect.Citation: Jiang, H., and G. Feingold (2006), Effect of aerosol on warm convective clouds: Aerosol-cloud-surface flux feedbacks in anew coupled large eddy model, J. Geophys. Res., 111, D01202, doi:10.1029/2005JD006138.1. Introduction[2] The aerosol-cloud-climate system is a complex one,comprising myriad feedbacks that challenge our ability topredict the radiative response of clouds to changes inaerosol. In addition to the direct effect of aerosol onradiation (the ‘‘direct effect’’), a host of ‘‘aerosol indirecteffects’’ have been proposed. These include the ‘‘firstindirect effect’’ [Twomey, 1974] which considers the re-sponse of cloud drop size and reflectance to a change inaerosol with reference to a constant liquid water content(LWC); the ‘‘second indirect effect’’ [Albrecht, 1989] whichproposes that an increase in aerosol will reduce the ability ofa cloud to precipitate, increase cloud liquid water, andextend cloud coverage and lifetime; the ‘‘semidirect effect’’[Grassl, 1975; Hansen et al., 1997] that considers thereduction in cloudiness due to the presence of absorbingaerosol in the atmosphere ; and various other indirect effects[e.g., Jacobson, 2002; Lohmann and Feichter, 2005] thatawait further elucidation.[3] The first indirect effect, or the closely related aerosoleffect on cloud drop number and size, has been identified innumerous in situ observations [Warner and Twomey, 1967;Durkee et al., 2000; Brenguier et al., 2000], by satelliteremote sensors [Kaufman and Nakajima, 1993; Han et al.,1998; Bre´on et al., 2002; Nakajima et al., 2001] andsurface-based remote sensors [e.g., Feingold et al., 2003;Kim et al., 2003 ] but quantification of the magnitude of thiseffect remains an elusive goal. This is illustrated by therange of observed relationships between cloud drop con-centration Ndand accumulation mode aerosol concentrationNaderived from field studies [see, e.g., Ramanathan et al.,2001]. The range of Ndversus Narelationships is due invarying, and uncertain degrees to aerosol concentration,size distribution, composition and updraft velocity [e.g.,Twomey, 1959; Leaitch et al., 1996; Facchini et al., 1999;Nenes et al., 2002; Feingold, 2003]. Improved understand-JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 111, D01202, doi:10.1029/2005JD006138, 2006Copyright 2006 by the American Geophysical Union.0148-0227/06/2005JD006138$09.00D01202 1of12ing of the Nd Narelationship is not a sufficient criterionfor assessment of the first indirect effect, which must alsoinclude the effects of spatial heterogeneity and three-dimensional cloud structure.[4] The second indirect effect relaxes the reference toconstant LWC and opens a very broad range of possibilitiesof cloud response to aerosol via dynamical, radiative, andeven surface flux feedbacks. Observational assessments ofthe second indirect effect, including perturbations to cloudlife-cycles, are very difficult to achieve, but there is evi-dence of aerosol-induced reduction in precipitation, partic-ularly associated with biomass burning [Warner, 1968;Rosenfeld, 1999]. A number of modeling studies havepointed out that even the sign of these responses is unclear[Stevens et al., 1998; Jiang et al., 2002; Feingold andKreidenweis, 2002] and dep endent am ongst oth ers, ontemperature, humidity and stability parameters, both in theboundary layer and above [Ackerman et al., 2004; Lu andSeinfeld, 2005]. In the marine stratocumulus environment,increases in aerosol result in decreases in cloud liquid waterpath (LWP) when dry air resides above the boundary layer,whereas moister conditions above the boundary layer resultin an increase in LWP with increasing aerosol [Ackermanet al., 2004]. Stevens et al. [1998] showed that when theair above the stratocumulus-c apped boundary laye r i smoist, a small amount of drizzle promotes higher LWPby stabilizing the boundary layer and reducing entrainmentrates. Jiang et al. [2002] showed that polluted aerosollayers residing above stratocumulus clouds reduced pre-cipi tation as well as LWP by r educing the supply ofmoisture from cumulus penetrating into stratocumulus. Intheir simulations, cloud albedo was almost unaffected bythe increases in aerosol because increases in drop numberconcentration were accompanied by decreases in LWP.Such effects are likely highly sensitive to the thermody-namic state of the atmosphere .[5] The semidirect