A22D-01 INVITED
What determines aerosol number concentration for cloud droplet nucleation and radiative forcing?
Aerosol number concentrations, size distribution and chemical characteristics are all important for determining the cloud droplet concentration and hence the radiative forcing of anthropogenic aerosols. Aerosol number concentrations and CCN concentrations result from emissions of primary particles together with nucleation of new particles, their coagulation to accumulation mode aerosols, and the condensation of gas phase precursors onto pre-existing particles. Significant uncertainty in determining the forcing by anthropogenic aerosols has remained because the total number of new particles generated in the present-day atmosphere that grow to CCN sizes is unknown. Here, we examine the importance of binary homogenous nucleation of H2SO4(g) and a first- order nucleation scheme based on measured rates for determination of aerosol number, cloud droplet number, and indirect radiative forcing by anthropogenic aerosols.
A22D-02 INVITED
Cancellation of aerosol indirect effects in stratocumulus through cloud thinning
It is generally assumed, and climate models show, that suppression of precipitation by increasing atmospheric aerosols results in an increased cloud albedo (e.g. increased cloud thickness, liquid water content, coverage, lifetime) in addition to that expected from the Twomey effect. However, recent high resolution model results and some observations challenge this notion. Here, I demonstrate that partial cancellation of the Twomey effect by the suppression of precipitation can be found in even the simplest of physical cloud models (mixed layer model) allowing an exploration of the factors that determine the strength of the aerosol indirect effect (AIE) in stratocumulus clouds. The results show that in order for a model to successfully represent AIEs it must be able to correctly treat (a) the response of precipitation to both cloud droplet concentration and liquid water content; (b) the entrainment process; (c) evaporation in the layer below cloud. The ability of our current generation of climate models to adequately resolve these processes is questionable. http://www.atmos.washington.edu/~robwood/papers/mlm/aie.pdf
A22D-03
Implications of cloud resolving model simulations of aerosol indirect effects on cloud radiative properties and precipitation to global climate
There is mounting evidence from my group's modeling studies of aerosol impacts on cloud radiative properties and precipitation as well as a number of other research groups, that once the precipitation process is modified, the cloud response is highly nonlinear. In some cases the clouds may respond to pollution aerosols by becoming optically thicker and raining less while in other cases just the opposite happens. This is true for marine stratocumulus clouds, tradewind cumuli, and cumulonimbi. In this talk I provide examples of such responses. The implication of this work is that cloud responses to pollution aerosols are very difficult to parameterize realistically in general circulation models and moreover that those responses are quite unpredictable at the present time. This is not to say that pollution aerosol impacts on the climate system is not important, but that quantifying that impact is a great deal more challenging than we have been lead to believe.
A22D-04
Efficiency of primary and secondary aerosols in indirect aerosol effects
Atmospheric aerosols play an important role in the global climate system through modifications of the global radiation budget: directly, by scattering and absorption of radiation and indirectly, by the modification of cloud properties and abundance. In particular the indirect aerosol effects on clouds are subject to large uncertainties. Global aerosol-cloud climate models allow quantitative estimates, albeit uncertain, of anthropogenic indirect aerosol effects. In this study we investigate the indirect aerosol effects through modeling studies with the ECHAM5-HAM aerosol- climate model with microphysical representation of aerosol-cloud interactions: we utilize its prognostic aerosol size-distribution, mixing state and cloud droplet number concentration to explicitly couple the aerosol and cloud systems via an explicit, Koehler theory based, aerosol activation scheme. Our focus is on the different efficiencies of primary and secondary aerosols in perturbing the global radiation balance - as our previous work has indicated non-negligible differences in the ability to form accumulation mode sized particles, as surrogate for cloud condensation nuclei. We present the modeling results in synergy with satellite observations of aerosol and cloud parameters, providing strong observational constraints on the simulated global aerosol-cloud interactions. Our results help reduce uncertainties in estimates of the indirect aerosol effects and provide valuable information for the necessary level of detail in the microphysical process representation in global models of aerosol-cloud interactions.
A22D-05
What does the common metric of the first aerosol indirect effect really measure?
The central point of the first aerosol indirect effect is that an increase in aerosol number concentration will increase droplet concentration, reduce effective radius (re), and enhance cloud albedo. Therefore, a good measure of the aerosol indirect effect should quantify the effect resulting from aerosol number concentration. In observation of the first aerosol indirect effect, a commonly used measure is the relative change of effective radius (re) with aerosol number concentration (Na): I = dln(re)/dln(Na). In this work, we show that although aerosol number concentration appears to be considered explicitly in this metric, the log operation minimizes the effect from aerosol number concentration. This metric actually measures a sum of two dispersion effects: effect on clouds of the relative dispersion of the pre-cloud aerosol size distribution, and the aerosol effect on the relative dispersion of the cloud droplet size distribution. We further show that a complete characterization of the first aerosol indirect effect requires at least three aerosol quantities: number concentration, mass concentration, and relative dispersion of the aerosol size distribution. Therefore, if resources are available, it is desirable to have one metric or several metrics that completely encompass the effects from all the three quantities.
A22D-06
Effects of Solar Heating on the Indirect Effect of Aerosols as Deduced from Observations of Ship Tracks
Inferring the indirect radiative effect of aerosols from observations is confounded by the response of clouds to many competing thermodynamic processes. These processes often overwhelm the effects of the particles. Observations of ship tracks, on the other hand, provide differences in cloud properties between polluted and nearby unpolluted clouds that are many times the differences among the unpolluted clouds on opposite sides of the track, thereby revealing the changes in the clouds are due solely to the haze. Marine stratocumulus affected by haze generally exhibit larger optical depths, and thus higher albedos, and smaller droplet radii than nearby unpolluted clouds. But this response is governed by the environment in which the clouds are embedded and also by the period over which the clouds are subjected to solar heating. One kilometer Moderate Resolution Imaging Spectroradiometer (MODIS) observations for Terra (morning) and Aqua (afternoon) were used to follow the morning to afternoon evolution of marine stratocumulus off the west coast of the U.S. that were affected by ship stack exhaust. The observations covered the summer months of 2002 and 2003. Low-level winds from NCEP re-analyses were used to identify the clouds common to both the Terra and Aqua observations. The 2002 and 2003 data contained several hundred ship track pairs in which the polluted clouds observed by the Terra MODIS were also observed by the Aqua MODIS. For overcast conditions, morning clouds had higher optical depths but the same droplet effective radius as the afternoon clouds. Consistent with the greater optical depths, liquid water amount was also higher (~10%) for the morning clouds. Under broken conditions, polluted clouds had more liquid water than nearby unpolluted clouds, whereas for overcast conditions polluted clouds had less liquid water. While work is underway to increase the ensemble of cases studied, the comparison of morning and afternoon clouds is revealing that while the changes in optical depth were the same for the morning and afternoon clouds, the decreases in droplet radius were greater for the afternoon clouds. This response appeared to be consistent with the afternoon breakup of the marine cloud layer. The unpolluted afternoon clouds had larger droplet radii than their morning counterparts, presumably from the growth of the droplets and formation of drizzle that promoted the breakup of the unpolluted clouds.
A22D-07
Constraining the effect of aerosol on warm cloud formation
Accurate knowledge of cloud droplet number concentration is required to predict cloud albedo and precipitation efficiency. Calculation of cloud droplet number from aerosol properties is challenging even in the simplest available framework of an adiabatic parcel rising at constant updraft. Using classical Köhler theory, a minimum of 12 parameters are required to initialize a simulation. Several of these parameters can be condensed into a single hygroscopicity parameter which can represent aerosol composition. A second grouping termed the ‘physicochemical activation parameter' can be constructed to combine the hygroscopicity parameter, surface tension, and the geometric mean diameter of the size distribution. The problem of cloud droplet formation can then be reduced to five unique parameters: aerosol number concentration, geometric standard deviation, updraft, condensation coefficient, and the ‘physicochemical activation parameter'. We modified a parcel model to survey the parameter space in terms of those five inputs. Using this approach we systematically establish the sensitivity of simulated cloud droplet number concentrations to variations in hygroscopicity, surface tension, particle size distribution, updraft, and condensation coefficient. We find that droplet concentrations are always three times as sensitive to changes in mode diameter as to changes in aerosol hygroscopicity. Evaluating the various influences on overall sensitivity of cloud droplet number concentration reveals that at least 70% of the overall sensitivity stems from sensitivity to size distribution parameters, 20% stems from sensitivity to updraft and condensation coefficient, and 10% stems from sensitivity to aerosol hygroscopicity and solution surface tension.