A22B-01 INVITED 10:20h
Effects of Aerosols on Clouds
Aerosols make an important contribution to the heat balance of the Earth's atmosphere. However, a potentially more important effect of atmospheric aerosols is in their role in the formation of cloud particles and precipitation. Field observations of the effects of aerosols on clouds and precipitation will be critically reviewed. Suggestions will be made for studies to further elucidate these effects.
A22B-02 10:44h
MODIS Observations of Ship Tracks: The Response of Marine Stratus to Smoke Particles
Ship tracks offer a unique opportunity for studying the response of marine stratus to increased particle concentrations. The study of ship tracks avoids the difficulties encountered in regional and global studies of cloud systems affected by haze in that in order to deduce the aerosol effects in the regional studies the response of clouds to the thermodynamics of the air mass in which the haze is imbedded must also be accounted for. With ship tracks, polluted clouds are compared with unpolluted clouds separated by only a few kilometers thereby insuring that the thermodynamics of the environments are on average the same and the only difference is the elevated particle concentrations in the polluted clouds. While marine stratus represent a specific cloud system having behavior that may not be representative of all low-level cloud types, such clouds nonetheless offer opportunities for testing the fidelity of at least a subset of cloud models. The use of MODIS observations to study ship tracks follows an earlier study based on AVHRR data. The earlier study demonstrated that polluted clouds lost on average approximately 20 percent of their liquid water. The finding contradicts the widely accepted theory that precipitation is suppressed in polluted clouds and thus should lead to increased cloud liquid water, cloud lifetimes, and cloud cover. The MODIS observations and the new analysis schemes developed for use with the MODIS data include several improvements on the earlier study with AVHRR data. First, MODIS radiances at 1.6 and 2.1 microns provide additional measures of cloud droplet radii thereby providing additional assessments of the changes in cloud liquid water which for the AVHRR observations were restricted to the use of 3.7-micron radiances to obtain droplet radii. Second, the new analysis procedures account for partial cloud cover in the 1-km fields of view and thus provide an objective means for estimating the growth in cloud liquid water due to increased cloud cover fractions for the polluted clouds. Third, a new automated scheme for comparing the polluted and nearby unpolluted clouds allows for comparison of clouds within 5 km, typically the autocorrelation length for cloud properties within marine stratus, thereby enhancing the contrast between the polluted and unpolluted clouds against the variability of the properties exhibited by marine stratus. Preliminary analyses are restricted to 1-km pixels found to be overcast by marine stratus. The results obtained with the MODIS radiances confirm the earlier findings that most polluted clouds lose liquid water. Based on recent LES model simulations of marine stratus, the loss is thought to result from the entrainment of the dry free tropospheric air above the clouds leading to a drying of the boundary layer where the clouds are polluted. Where the clouds remain unpolluted, the entrained dry air is less effective in drying the boundary layer. Analyses are underway to determine whether support for these model results can be found in the column water vapor amounts derived from the MODIS 0.905 and 0.935-micron radiances.
A22B-03 10:56h
Smoke Suppression of Clouds in Brazil
We present a study of smoke aerosol effects on cloud formation in the biomass burning regions of Brazil and investigate the reasons for the progressive reduction in cloud fraction with increasing smoke optical depth that has been observed by remote sensors. Numerical simulations of cumulus clouds are performed with a large eddy simulation (LES) coupled to an aerosol/cloud microphysical model and a radiation model. A sounding from the Smoke, Aerosols, Clouds, Rainfall and Climate (SMOCC) experiment is used as a framework for the numerical experiments. The model shows that the vertical distribution of smoke is important in determining the effect on clouds. Elevated smoke layers result in a strong reduction in cloud fraction whereas smoke confined mostly to the boundary layer has little effect on cloud fraction. Heating associated with smoke embedded in droplets has a negligible effect on the clouds for the conditions studied. On the other hand, the observed and modeled reduction in surface latent and sensible heat fluxes asssociated with smoke, is by itself sufficient to reduce cloud fraction significantly. This has important implications for land/ocean differences in the cloud response to smoke.
A22B-04 11:08h
On the Determination of CCN from Satellite: Challenges and Possibilities
We are using aerosol size distributions measured in the size range from 0.01 to 10+ um during TRACE-P and ACE-Asia, results of chemical analysis, measured/modeled humidity growth and stratification by air mass types to explore correlation between aerosol optical parameters and aerosol number concentration. Size distributions allow us to integrate aerosol number over any size range expected to be effective cloud condensation nuclei (CCN) and provide definition of a proxy for CCN (CCNproxy). Because of the mixed nature of the accumulation mode aerosol and the link between volatility and solubility this CCNproxy can be linked to the optical properties of these size distributions at ambient conditions. This allows examination of the relationship between CCNproxy and the aerosol optical properties expected to be seen by satellites. Relative increases in coarse aerosol (e.g. dust) generally add little particle number to effective CCN but significantly increase scattering detected by satellite and drives the Angstrom exponent to approach zero. This has prompted the use of a so-called aerosol index (AI) based upon the product of the scattering and the non-dimensional Angstrom exponent, both capable of being inferred from satellite observations. This biases the AI to be closer to scattering values generated by particles in the accumulation mode (Angstrom exponent about 1 to 2) that dominate particle number and is therefore dominated by sizes commonly effective as CCN. Our measurements demonstrate that AI does not generally relate well to measured CCN proxy unless the data is suitably stratified. Multiple layers, complex humidity profiles, dust with very low Angstrom etc. mixed with pollution and size distributions differences in pollution and biomass emissions appear to contribute most to method limitations. However, we demonstrate that these characteristic differences result in predictable influences on AI. These results suggest that new satellite and model capabilities can be integrated to improve on satellite retrieval of CCN.
A22B-05 11:20h
Aerosol-Cloud Interactions Over the North Pacific Ocean: an Integrated Assessment Using Aircraft, Satellites and a Global Model
Interactions between aerosols and the cloud systems of the North Pacific Ocean were observed by aircraft during the Cloud Indirect Forcing Experiment (CIFEX) in April 2004. The CIFEX project seeks to determine the impact of aerosol indirect effects on the radiative forcing of highly reflective North Pacific clouds under the influence of aerosols traveling across the ocean basin from Asia. Toward this end, CIFEX aircraft observations from the Northeast Pacific of aerosol and cloud microphysics are blended with coincident satellite observations of cloud properties from MODIS and cloud radiative forcing from CERES. The satellite observations are then compared with global model simulations of aerosol indirect forcing over the entire North Pacific basin. During April 2004 the U. Wyoming King Air research aircraft sampled aerosol and cloud microphysical parameters including aerosol and cloud particle sizes and concentrations, cloud liquid water amounts, and cloud structure using the Wyoming Cloud Radar. A range of clean and polluted conditions were observed by the aircraft during the period, in addition to two major Asian dust storm events. CN concentrations below stratus clouds varied from 25 to 300 cm-3. A variety of cloud systems were sampled as well, ranging from shallow stratus and stratocumulus clouds to mixed-phase precipitating cumulus. Under pristine conditions, many shallow clouds were observed to be drizzling, suggesting that Northern Pacific Ocean cloud systems may be highly susceptible to the influence of aerosols. Clouds in this region are responsible for a large cooling of the ocean surface. The magnitude of shortwave cloud radiative cooling exceeded -80 W m-2 over much of the North Pacific during the experiment. Stratus cloud drop concentrations varied from 25 to 150 cm-3 and are correlated with the concentration of accumulation mode aerosols below cloud base. Mean cloud albedos vary from 0.3 to 0.5 for these same clouds, and MODIS observations of cloud drop effective radius compare well with the in-situ measurements. To facilitate the integrated interpretation of these observations and test modern parameterizations of aerosol-cloud interactions in global models, simulations are performed with the NOAA/GFDL AM2 global atmospheric model under the conditions observed during CIFEX. The variability of observed radiative forcing of North Pacific cloud systems are compared with a simulation using a new parameterization based on a prognostic equation for cloud drop number concentration, and a separate simulation using a common diagnostic parameterization based on an empirical relationship between sulfate concentration and cloud drop number concentration.
A22B-06 11:32h
Satellite-Based Assessment of the Aerosol Effect on Global Warm Cloud Properties
We present characteristics of global warm cloud properties and warm-rain process in conjunction with the aerosol index (AI) and the lower-tropospheric stability (LTS). The Tropical Rainfall Measurement Mission (TRMM) Microwave Imager (TMI) and Visible/Infrared Radiance Imager (VIRS) simultaneously derive cloud-top droplet effect radius, column droplet radius, cloud fraction, cloud liquid water path, cloud optical depth, and a warm rain index. These cloud properties are clustered by different bins of LTS, and then compared with GOCART-derived AI and MODIS-derived AI values. Results indicate that the characteristics of the aerosol-cloud interactions significant vary between different cloud types and thermodynamic environments. This indicates that the radiative effect and warm-rain process due to aerosols have significant heterogeneous regional climate forcing effects.
A22B-07 11:44h
Clouds Aerosols Internal Affaires: Increasing Cloud Fraction and Enhancing the Convection
Clouds developing in a polluted environment have more numerous, smaller cloud droplets that can increase the cloud lifetime and liquid water content. Such changes in the cloud droplet properties may suppress low precipitation allowing development of a stronger convection and higher freezing level. Delaying the washout of the cloud water (and aerosol), and the stronger convection will result in higher clouds with longer life time and larger anvils. We show these effects by using large statistics of the new, 1km resolution data from MODIS on the Terra satellite. We isolate the aerosol effects from meteorology by regression and showing that aerosol microphysical effects increases cloud fraction by average of 30 presents for all cloud types and increases convective cloud top pressure by average of 35mb. We analyze the aerosol cloud interaction separately for high pressure trade wind cloud systems and separately for deep convective cloud systems. The resultant aerosol radiative effect on climate for the high pressure cloud system is:-10 to -13 W/m2 at the top of the atmosphere (TOA) and -11 to -14 W/m2 at the surface. For deeper convective clods the forcing is: -4 to -5 W/m2 at the TOA and -6 to -7 W/m2 at the surface.
A22B-08 INVITED 11:56h
Aerosol Interactions With Deep Convective Clouds and Their Climate Impacts Through the Hydrological Cycle
Convective maritime clouds typically have well developed warm rain that much of it precipitates before ever freezing, moderate updrafts, little supercooled water and relatively high glaciation temperature. In contrast, continental clouds have much less warm rain, stronger updrafts, deeper mixed phase zone and more than 10 times more lightning for the same rainfall amount. These differences have been attributed traditionally to the differences between land and sea surfaces and to the greater relative humidity over oceans. Recent evidence show that the enhanced small CCN aerosols over land contribute significantly to the "continental" nature of the clouds, whereas clouds developing in pristine air over land have much more "maritime" nature. These differences are associated with similarly profound differences in the latent heating profile, where the continental clouds deposit the released latent heat higher in the atmosphere. Furthermore, greater instability is consumed for the same rainfall amount in continental convection, requiring greater recovery time of the convection by the radiative cooling of the troposphere. Sensitivity studies have shown that regional and global circulation systems are quite sensitive to such changes. These pollution aerosols changes are less obvious over ocean, mainly because polluted clouds over ocean are "hygroscopically seeded" by sea salt aerosols that enhance the warm rain processes. In spite of the reduced effect, over ocean, recent satellite (MODIS) analyses of cloud properties and aerosols show that cloud cover and vertical development of the convective clouds are much larger for greater aerosol amounts. Such aerosol increased cloud cover and vertical height can be explained by the aerosol effect on slowing down the precipitation forming processes, and the dynamic feedbacks on the cloud development. We can see these effects over oceans and not over land because aerosols cannot be detected by satellite as accurately over land. However, over land, where most CCN are anthropogenic, these effects are likely even greater than over ocean. Therefore, it is probable that anthropogenic aerosols have been already profoundly impact the precipitation and through that the cloud cover, cloud vertical development, hydrological cycle and hence the global circulation. Realistic simulations of these effects with GCMs is a major challenge, because much of the underlying cloud physics is still missing in them.