A21C-0748 0800h
Equilibrium Sizes of Cloud Droplets Condensed on Multi-sized CCN After In-size-Class and Inter-size-Class Competitive Growth
(Introduction) The equilibrium size of a cloud droplet condensed on a CCN is estimated by traditional Köhler model derived from thermodynamic equilibrium between droplet and gas phases (Pruppacher and Klett, 1980). However, Köhler model is based on the idealistic assumption that a cloud droplet grows in an infinitely large reservoir of water vapor at constant pressure and constant temperature. In application to realistic air parcels, this brings about two serious faults in size estimation (Shiba et al, 2003). The first is failure to consider competitive growth of plural droplets (i.e., effect of CCN number density). The second is limitation of the maximum allowable saturation ratio to get a definite droplet size. A new model applicable to competitive growth on multi-sized CCN has been developed, taking account of both water vapor reduction and temperature rise to remove above faults. (Mathematical Model) For multi-sized CCN of $n$ size-classes, $n+2$ governing equations are solved to estimate radius $a_{i}$ ($i=1, 2,., n$), saturation ratio $S$ and temperature $T$. They are derived from (1) $n$ equilibria between droplet chemical potential and vapor one, (2) conservation of mass, and (3) conservation of heat energy. (Model Calculations of Cloud Droplet Size) To do demonstrative calculations, three virtual air parcels (Types 1, 2, and 3) are arranged, combining CCN size [(small, medium, large) = (0.1, 0.5, 1.0) $mu$m] with CCN number [(most, average, least) = (750, 500, 250) cm$^{-3}$]. Types 1, 2, and 3 are small-CCN-rich parcel (750, 500, 250), even-CCN-distribution parcel (500, 500, 500), and large-CCN-rich parcel (250, 500, 750), respectively. Parenthesized CCN numbers are put in up-sized order of CCN. The more air parcel contains large CCN, the smaller droplet size becomes in any size-class regardless of parcel type. Droplet sizes in small and medium size-classes seem to be contrary to idea of in-size-class competition that less competitive growth produces larger cloud droplets. From the viewpoint of in-size-class competition we expect that small size-class droplets in large-CCN-rich air parcel becomes larger than that in small-CCN-rich air parcel, because number of small CCN in large-CCN-rich air parcel is less than that in small-CCN-rich air parcel. If cloud droplet size were controlled exclusively by in-size-class competition, as for small size-class, Type 3 air parcel [which contains the least (250 cm$^{-3}$) small CCN] would have the largest size droplet and Type 1 air parcel [which contains the most (750 cm$^{-3}$) small CCN] would have the smallest one. However, in any size-class, air parcel order according to droplet size is the same as that in the largest size-class (i.e., regardless of size-class, Types 1, 2 and 3 are in down-sized order of droplet). This means that the largest CCN controls not only in-size-class competition with themselves but also inter-size-class competition with smaller CCN. (Conclusions) Model calculations show that: (1) Equilibrium size of cloud droplet on multi-sized CCN is controlled by large size CCN; (2) Inter-size-class competitive growth brings about a kind of nonlinearity in mapping of CCN size distribution to cloud droplet size distribution.
A21C-0749 0800h
Sensitivity of a Cloud-Resolving Model to the Bulk and Explicit Bin Microphysical Schemes
A cloud-resolving model is used to study sensitivities of two different microphysical schemes, one is the traditional bulk type, and the other is an explicit bin scheme, in simulating a mid-latitude squall line case (PRE-STORM, June 10-11, 1985). Simulations using different microphysical schemes are compared with each other and also with the observations. Both the bulk and bin models reproduce the general features during the developing and mature stage of the system. Furthermore, the observations and the well-proven bulk scheme simulation serve as validations for the newly incorporated bin scheme. However, it is also shown that the bulk and bin simulations have distinct differences, most notably in the stratiform region of the squall line system. Weak convective cells exist in the stratiform region in the bulk simulation, but not in the bin simulation. These weak convective cells in the stratiform region simulated in the bulk scheme model are remnants of the stronger convections previously at the leading edge of the system, sustained by horizontal vorticity generated by its own cool pool near the surface. The bin simulation, on the other hand, has a horizontally homogeneous stratiform cloud structure, which agrees better with the observations. Examinations of the downdraft core strength, the potential temperature perturbation, and the evaporative cooling rate show that the differences between the bulk and bin models are due mainly to the stronger low-level evaporative cooling in the convective zone simulated in the bulk microphysical scheme, which is unrealistic because of the assumptions made in raindrop size distribution. Further sensitivity tests that reduce the evaporation rate in bulk scheme artificially produce more upright convective core and less weak cores in stratiform region. However, they produce weaker upper level outflow and consequently less stratiform rain area. The addition of a more realistic raindrop breakup scheme in the bin scheme results more realistic radar reflectivity and stronger surface rainfall. Despite the increase of the rain evaporation and strengthening of the near surface cool pool, bin scheme with rain breakup shows homogeneous stratiform rain. These sensitivity tests prove the robustness of the bin microphysical scheme and the difficulty of tuning the limited parameters in the bulk microphysical scheme to realistically reproduce detail structures in a mid-latitude squall line case study.
A21C-0750 0800h
Water-soluble organic carbon in coarse particles in a coastal environment: Role of cloud processing
Water-soluble organic carbon (WSOC) components constitute a significant fraction of atmospheric aerosol carbon mass, ranging from 5% to 83%. Because of their hydrophilic nature, water-soluble organic compounds are active in modulating aerosol's role to act as cloud condensation nuclei. Despite the abundance of WSOC, sources and formation pathways of WSOC are still poorly understood. In this study, we measured the size distribution characteristics of coarse mode WSOC and investigated its relationship with major inorganic constituents in the coastal city of Hong Kong in summer 2004. Atmospheric particles were collected into five size bins in the coarse mode ($>$18, 18-10, 10-6.2, 6.2-3.2, 3.2-1.8 micrometer) and one size bin in the fine mode ($<$1.8 micrometer). The WSOC concentrations in two coarse bins (1.8-3.2 and 3.2-6.2 micrometer) show strong correlations with the WSOC in the fine mode ($<$1.8 micrometer). In the two coarse bins and the fine size bin, WSOC was well correlated with non-sea-salt sulfate while showed no correlation with either sea salt or crustal species. WSOC concentrations in particles larger than 6.2 micrometer were not correlated with the fine mode WSOC concentrations, and found to have a constant value of ~ 0.3 microgram per cubic meter. We postulate that cloud processing, known to be the dominant formation pathway for fine mode sulfate, was mainly responsible for the production of WSOC in the two coarse sizes of 1.8-3.2 micrometer and 3.2-6.2 micrometer and in the fine particles. On the other hand, the nearly constant presence of WSOC on larger coarse particles ($>$ 6.2 micrometer) was likely a result of primary emissions.
A21C-0751 0800h
Cloud Condensation Nucleus (CCN) Activation Properties of Biogenic Secondary Organic Aerosol
Organic compounds are known to comprise a significant fraction of the atmospheric aerosol population and have been found to contribute to the concentration of cloud condensation nuclei (CCN). Much of this organic material is secondary in nature; secondary organic aerosol (SOA) is formed when volatile organic compounds are oxidized to form less volatile products, which then condense into the aerosol phase. Many organic compounds found in the atmosphere, of both anthropogenic and biogenic origin, have been found to produce SOA. Such reactions typically result in complex mixtures of products, only a fraction of which have been identified. Thus while there have been several studies exploring the potential for organic particles to act as CCN (including some of the compounds identified in SOA products), there have been almost no direct investigation of the potential CCN activity of SOA. This paper presents the results of a series of experiments measuring directly the CCN activity of SOA produced by the ozonolysis of several common biogenic compounds. Six compounds were studied: five monoterpenes ($\alpha$-pinene, $\beta$-pinene, $\Delta^{3}$-carene, limonene, terpinolene) and one terpinoid alcohol (terpinen-4-ol). The chosen monoterpenes represent an estimated 87% of global monoterpene emissions, while the terpenoid alcohols make up approximately 25% of the other biogenic compounds capable of forming SOA. In each experiment, SOA was generated under controlled conditions at the Caltech indoor facility. Over several hours, CCN concentrations were measured at supersaturations ranging from 0.27% to 0.80%. These data are compared to simultaneous particle concentration and size distribution observations to determine the relationship between particle diameter and CCN activity. The analysis indicates considerable variation in CCN activity among the experiments; possible causes for such variability are explored.
A21C-0752 0800h
Regional and Season Variations in Stratus Cloud Properties from MODIS Observations
Marine stratus clouds represent a climatologically significant influence on the global energy and water cycle. By possessing a higher albedo than the underlying ocean surface, these clouds cause a significant decrease in the amount of solar radiation absorbed in the ocean's mixed layer. This radiative impact is influenced by cloud micro- and macrophysical properties that are not fully understood and a realistic representation of their structure and properties is essential to obtain realistic climate model simulations. We use four full years of cloud retrievals from MODIS for five prominent marine stratus regions to compile statistics of stratus cloud macro- and microphysical properties. We employ a rigorous, automated quality control procedure that determines the presence of stratus and removes portions of scenes that are affected by sunglint, contain cirrus or frontal clouds, or contain other retrieval errors. We also investigate the possible effects of partially cloudy pixels by compiling statistics both including and excluding cloud edge pixels. We identify more than 31,000 GCM-grid-sized regions for which we define measures of central tendency and variability of important cloud properties (e.g. optical depth, liquid water path and cloud particle effective radius), and bulk scene properties including the cloud fraction and a measure of mesoscale structure. The analysis explores the seasonal and regional differences in cloud properties, their sub-grid scale properties, and how these vary as a function of cloud macrophysics. Our results show several interesting regional and seasonal differences in cloud properties, which can be used for model validation and development of stratus parameterizations.
A21C-0753 0800h
Isoprene Forms Secondary Organic Aerosol Through Cloud Processing: A Model Study
Cloud processing of water-soluble organic vapors has been proposed as a pathway for the formation of organic particulate matter (PM) in the atmosphere (i.e., secondary organic aerosol; SOA). The simulations described below suggest that cloud processing of isoprene is a substantial contributor to atmospheric oxalic acid and SOA formation. Isoprene, a biogenic volatile organic compound (VOC) of global importance, forms highly water-soluble glyoxal, methylglyoxal, and glycolaldehyde in the gas phase. These carbonyls favorably partition into cloud droplets where they oxidize to organic acids (e.g., glyoxylic acid, glycolic acid, pyruvic acid, and oxalic acid). In this study we developed a box model to examine the importance of isoprene chemistry to in-cloud formation of SOA. The box model incorporates gas- and aqueous-phase chemistry and phase transfer of relevant water-soluble species. Simulations were conducted under clean conditions typical of the tropical Amazon (i.e., high emission flux of isoprene and low emissions of anthropogenic pollutants). Simulation results show that isoprene forms a substantial amount of organic acids through cloud processing. More than 80% of oxalic acid is expected to remain in the particle phase after cloud evaporation. This study suggests that cloud processing of isoprene is a substantial contributor to atmospheric water-soluble SOA that can alter the microphysics of cloud condensation nuclei (CCN). This work also discusses the sensitivity of the model to important model parameters.
A21C-0754 0800h
Summary of the Chemistry Transport in Deep Convection Cloud Modeling Workshop Intercomparison
The cloud chemistry case of the 6th International Cloud Modeling Workshop investigated transport of ozone (O$_3$), carbon monoxide (CO), and NOx in deep convection as simulated by several cloud-scale chemistry models. The purpose of the intercomparison was to assess the capability of each model to transport chemical species from the boundary layer to the upper troposphere including the entrainment of free tropospheric air. Parameterizations of lightning-produced NOx and transport of the soluble species nitric acid, hydrogen peroxide, and formaldehyde were also compared. The results from several models can now be used to document the variability among reputable cloud chemistry convective models for a given storm. Six models simulated the 10 July 1996 STERAO storm, which was observed in northeastern Colorado. Observations of CO, O$_3$, and NOx in the anvil were compared to modeled mixing ratios of these species. Generally, the model results were in good agreement with each other and with the observations for CO and O$_3$. Models that included the production of NOx by lightning were able to produce NOx mixing ratios of similar magnitude as observed values. Analysis of the anvil winds and species concentrations give an estimate of the mass flux of species to the upper troposphere. The fluxes from preliminary model results replicate analyzed fluxes from the observations to about $\pm$30% for air and $\pm$25% for CO. As part of documenting the variability among cloud-scale chemistry models, we will further archive characteristics of the model results. These characteristics include cloud area, cloud top, mass fluxes into the storm, into the UT, and onto the ground, and scavenging fraction of soluble species.
http://box.mmm.ucar.edu/individual/barth/TracerTransportDeepConvection.html
A21C-0755 0800h
Volatile Chemical Retention During Wet-growth Riming
Interactions of chemicals with the ice phase in tropospheric clouds are poorly understood. Recent research indicates that volatile chemical partitioning during cloud hydrometeor freezing and riming may significantly impact cloud transport of soluble chemicals (e.g. HNO3, H2O2, SO2, aldehydes, and carboxylic acids) important to acid precipitation, the oxidizing capacity of the troposphere, and global climate change. In previous work, we have developed theory-based models and parameterizations of chemical partitioning during non-rime freezing and dry-growth riming. In this work, we develop a steady-state model for partitioning during wet-growth riming. Our model is adapted from a model used to predict the salinity of sea-spray ice and is based on solute mass and rate balances over a wet-growth riming hail particle. It accounts for drop impingement on the hydrometeor, hail growth, evaporation from hail, and shedding of liquid water. It also accounts for chemical partitioning between the water, air, and ice phases. Our model predicts that the retention efficiency (the ratio of a solute's mass in hail to that in the impinging supercooled drops) for wet-growth riming will be a function of the liquid water content of the hail, the solute effective Henry's law constant, the effective ice-water distribution coefficient, and the rates of impingement, growth, and shedding. Application of this model within cloud-scale convective storm simulations will help elucidate the impacts of cloud transport and processing on tropospheric chemistry.
A21C-0756 0800h
Experimentation and modelling of mineral aerosol dissolution as source of transition metals in cloud droplets
Even at nano-molar concentrations, transition metals (TMI) could play an important role in the radical chemistry of the atmospheric liquid phase. For instance, cloud chemistry model calculations suggest that depletion of HOx by reactions between TMI and HO2/O2- radicals significantly slows down O3 production in polluted clouds. TMI are transferred into the liquid phase from aerosol particles by dissolution processes which can be a slow reaction. The dissolution kinetic of the solid phase thus competes with chemical kinetics in the homogeneous aqueous phase. It is therefore of importance to consider the evolution of TMI concentrations into cloud droplets in order to quantify the atmospheric impact of aerosols on the aqueous chemistry. Mineral particles including soil-derived particles and fly-ash are important sources of TMI in the troposphere.In order to parameterize the dissolution kinetic and concentrations of TMI from mineral particles into cloud droplets, we have conducted experimental laboratory simulations which mimic particles/water interactions occurring into droplets. These simulations were carried out in an open-flow reactor for typical atmospheric conditions (pH, ionic strength.). Data on TMI dissolution kinetic are provided for two generic kinds of mineral matrices from anthropogenic and natural sources: alumino-silicated and carbonaceous particles (dust, fly-ash, or urban particles). The concentrations of TMI released depend on pH, matrix type and particle-water contact time. The metals coming from carbonaceous particles are adsorbed impurities or salts and are very soluble with dissolution hardly dependent on pH. On the opposite, the metals dissolved from alumino-silicated particles are less soluble, notably the ones constitutive of the matrix network (Fe, Mn), and their dissolution is highly influenced by the pH. However, at a given pH, the results on the kinetic of dissolution emphasize that whatever the matrix, the TMI dissolution rates decrease with the contact time following a double first order kinetic. The kinetic constant of dissolution is determined for each metal from each particle type and at various pH. It appears that these constants are also directly related to the mineral matrix types. From these observations, the metals dissolution is parameterized as a function of particle-water contact time and particle composition. This parameterization could be included in cloud chemistry models.
A21C-0757 0800h
Aerosol size distribution in the tropical/subtropical upper tropsophere: First observation of in-situ new particle formation in cirrus clouds
Particle nucleation (formation of solid or liquid particles directly from the gas phase) in the upper troposphere and lower stratosphere is an important step in the chain of events that produce clouds, but the mechanisms are poorly understood. Previous studies show that new particle formation takes place in the outflows of marine stratus and cumulus clouds, because cloud scavenging leads to a reduction in the surface area of pre-existing aerosol particles. However, observations of particle formation in clouds are scarce. Here we show measurements indicative of in-situ new particle formation occurring in cirrus clouds. Measurements were made at altitudes from 7 to 16 km over Florida with instruments on the WB-57F aircraft during Cirrus Regional Study of Tropical Anvils and Cirrus Layers-Florida Area Cirrus Experiments (CRYSTAL-FACE) in July 2002. Size-resolved ice crystal particle concentrations and water vapor concentrations were measured to help identify the presence of cirrus clouds. Surprisingly high concentrations of ultrafine particles, diameters (D$_{p}$) from 4 to 9 nm (N$_{4-9}$), were measured in interstitial cirrus cloud aerosol. About 72 % of the in-cloud samples showed new particle formation events with the average N$_{4-9}$ of 3.0x10$^{3}$ cm$^{-3}$, whereas $\sim$56 % of the out-of-cloud samples had events with the lower N$_{4-9}$ of $\sim$1.3x10$^{3}$ cm$^{-3}$. Both size distributions of the in- and out-of- cloud samples with new particle formation events showed a high peak at D$_{p}$ $<$10 nm and another peak at D$_{p}$ $\sim$20 nm. The magnitude and frequency of new particle formation events seen in cirrus clouds were also higher than those previously observed in the tropical/subtropical upper troposphere in the absence of clouds. These results suggest that cirrus clouds may provide favorable conditions for particle formation, such as low temperatures, high relative humidity, high OH production (due to high water vapor), cloud electricity and atmospheric convection; but at present we are unable to identify particle nucleation mechanisms in clouds, a condition where high concentrations of surface area of ice crystal particles exist.
A21C-0758 0800h
Internal Structure in Supercooled Water Aerosols and Their Role in the Formation of Ice Clouds
The chemistry and physics of processes occurring in supercooled cloud droplets is a sensitive function of the structure of the liquid. Such properties as uptake and diffusion coefficients, viscosity and ice nucleation behaviour all change rapidly with the amount of supercooling in samples of pure water. Direct measurements of these properties can only be made in micron-sized droplets. This presentation reports an experimental study of the nanostructure of supercooled water droplets in the temperature range 240-294 K. The experiments were done in a cryogenic flow tube coupled to an FTIR spectrometer. The flowtube permits us to create and manipulate aerosol particles having a desired size distribution at precisely known cryogenic temperatures. The size distribution and internal structure of the supercooled liquid droplets are obtained from the infrared extinction spectra of the flowing aerosols. These measurements show that ordered ice-like nanoclusters form inside droplets of supercooled water. The size and number of these clusters increase with decreasing temperature; they can occupy up to 40% of the droplet volume near freezing. The size distribution of the particles in a supercooled water aerosol is also observed to change dramatically upon freezing. This demonstrates directly the importance of mass transfer through the vapour phase - a process that also occurs during formation of ice clouds. The efficient transfer of mass from liquid droplets to nascent ice crystals occurs because supercooled water has a substantial vapour pressure, even at its freezing temperature. In our experiments, when a water aerosol freezes, we observe the rapid formation of large ice particles with size distributions that are generally not lognormal and frequently are multi-modal. This complicates the interpretation of the freezing kinetics significantly, necessitating the use of an aerosol microphysics model that takes into account the mass transfer as well as the nucleation and freezing rates. We will report our observations of the formation of dynamic nanoclusters and our measurements of freezing rates and mass transfer properties in supercooled water droplets for temperatures down to about 234 K, which is the ultimate freezing temperature of the water aerosol droplets used in this study.
A21C-0759 0800h
Nitric Acid in Cirrus and Polar Stratospheric Clouds
Cirrus clouds and PSCs were observed by in-situ instruments on the high altitude research aircraft Geophysica in the Arctic winter 2002/3. Cirrus clouds were detected during several flights at temperatures between 200 and 215K. Besides optical particle parameters, the water and nitric acid content in the gas and the condensed phase were measured. First results of the uptake of nitric acid in or on Arctic cirrus clouds will be presented and compared to other measurements and models. Also, at altitudes above 18 kilometers small nitric acid containing particles - probably nitric acid trihydrate (NAT) - with diameters up to 6 micrometers and low number densities ($<2*$10-4 cm-3) were measured in a PSC on 6 February 2003. Interestingly, the NAT particles formed in less than a day at temperatures more than 3.5 K above the ice frost point. The detection of particles in this unique experimental situation at threshold PSC formation conditions allows to constrain current hypothesis of NAT formation. We derive a nucleation rate of NAT nucleation on meteoritic smoke, which might explain the present observations. After growth, those NAT particles have potential to remove HNO3 from the stratosphere (denitrification), which enhances polar ozone loss. We compare the measured denitrification with results from global model simulations.
A21C-0760 0800h
Inference of Cirrus Cloud Properties Using ARM Data
The properties of cirrus clouds are inferred over the Atmospheric Radiation Measurement (ARM) site of the Southern Great Plaines. The background (clear-sky) optical thickness is first determined by matching the calculated radiance to measured radiance using the gamma-fitting technique developed by Smith et al. (1993). Next, we generate a look-up table of cloud optical thickness and particle effective size. We do this by inserting an ice cloud layer, and varying the optical thickness and particle effective size. We can then infer the cloud properties over our site by comparing the measured Atmospheric Emitted Radiance Interferometer (AERI) data with the look-up library.