AT 620 Review for Midterm 1 Part 2

AT 620 Review for Midterm #1 Part 2: Chapters 5 -7 Brenda Dolan October 19, 2005

Chapter 5: Atmospheric Aerosols

Atmospheric Aerosols G Aerosol: Small liquid or solid particles suspended in a medium (the atmosphere). They are very small particles that do not have appreciable fall speeds. G Cloud Condensation Nuclei (CCN): Aerosols that are activated to serve as cloud nuclei at realistic (low) supersaturations (S that would be found in atmosphere G Condensation Nuclei (CN): All aerosols in atmosphere, including those that are activated at high supersaturations an may not serve as cloud nuclei under normal atmospheric conditions

Atmospheric Aerosols

Aerosol Production Processes: G 1) Gas-to-particle conversion (mostly Aitken production) -vapors from plant exhaltations and combustion products -chemical reactions catalyzed by UV radiation -chemical reactions in small water droplets (clouds process air, and thus there can be more concentrations of particles that have been processed by a cloud)

Aerosol Production Processes: G 2) Mechanical disintegration of the solid and liquid earth surface (mostly large and giant production) Solid earth -organic particulates by plants (pollen, seeds, waxes, spores) -mechanical and chemical disintegration of vegetation free rocks and soils -volcanic emissions -particles injected into the atmosphere by industrial processes (paper mills, steel mills)

Aerosol Production Processes: G 2) Mechanical disintegration of the solid and liquid earth surface (mostly large and giant production) Ocean -production of spray droplets at the crest of breaking waves (minor) -bursting air bubbles that are present at ocean’s surface (few in number but fairly large in size) -Primarily water-soluble sulfates G 3) Extraterrestrial sources (minor source of giant and large)

Aerosol Production Processes: G In terms of mass weighting, natural particles are the greatest sources of aerosols, while anthropogenic particles are minor sources. In terms of numbers, anthropogenic can be quite extensive. G Aerosol concentrations

Aerosol Distributions: G In general, aerosol concentrations drop off with height G Junge layer -Abrupt increase in aerosol concentrations in the lower stratosphere -Changes in time and season, but is observed world-wide -Possibly a result of volcanic eruptions G In the ocean, aerosol particles are not dominated by sea -salt particles, but rather oxidation of DMS G Aerosol concentrations can be variable over the oceans, but are significantly less than continental concentrations

Measuring aerosols: G 1) Electrical aerosol analyzer (EAA)—measure mass and size of aerosols based on their measured mobility in applied electric field G 2) Optical counters and nephelometers—concentration and size distribution of aerosols is determined by the amount (intensity) of scattered light. G 3) Direct impaction instruments—coated slides are swept through volumes of aerosols. Used for large particles (>0. 1 µm) G 4) X-ray techniques—evaluate composition of aerosols depending upon the radiation given off G 5) Aitken nucleus counter —expansion chamber used to create high supersaturations, then they are counted optically.

Aerosol removal processes: G Aitken particles 1) Coagulation: brownian motion causes particles to collide and self-collect 2) Capture by cloud droplets: either by condensation of vapor on surface, or direct impact on aerosol by a cloud droplet G Giant particles 1) Sedimentation: Dry deposition due to relatively large fall velocities 2) Precipitation scavenging: collection efficiencies are large

Aerosol removal processes: G Large particles: The Greenfeld gap G Large aerosols have the longest life because there is no efficient sink for them. Their fall velocities are not large enough in most cases for dry deposition, and they are in the size range where coaguation is not efficient. G 1) Some dry deposition G 2) Some precipitation scavenging

Aerosol removal processes: G Coagulation G Brownian motion: irregular movement of aerosol particles due to thermal bombardment by air molecules G Smoluchowski’s equation for Coagulation gain G Collection kernel loss

Aerosol removal processes: G Smoluchowski’s equation for Coagulation G Describes the change in size spectrum of aerosols G since particles that are moving irregularly have a finite probability of colliding and coagulating with one another and particles with relatively large mobilities collide and coagulate more readily, we need to define some efficiency that is related to mass. This is the diffusivity, D. G Diffusivity is inversely proportional to r and proportional to T

Aerosol removal processes: G Wet Removal Mechanisms G Phoretic effects G Condensation and evaporation of vapor molecules can effect the collection of aerosols, because aerosol particles being bombarded by vapor molecules experience a force directed toward the droplet surface, which enhances the coagulation between cloud droplets and aerosol particles. G Diffusiophoresis: Enhanced diffusion of aerosols to drop, enhancing the collection kernel G Thermophoresis: Diffusion of heat away from growing droplet, which inhibits collection of aerosols

Aerosol removal processes: G Wet Removal Mechanisms G Phoretic effects are most important for aerosol particles between 0. 1 µm> diffusiophoresis G This results in a reduced rate of aerosol particle scavenging by a cloud droplet growing by vapor deposition G This results in an enhanced rate of aerosol particle scavenging by a cloud droplet that is evaporating

Aerosol removal processes: [DRAW]

Aerosol removal processes: G Hydrodynamic capture G A large drop settling through smaller drops will sweep out a volume and collect aerosols with some efficiency, E G Depends on size of drops and size of aerosols G Most efficient for large and giant aerosols duet to large Vt and cross-sectional area

Cloud Condensation Nuclei: G ~1% of aerosol mass serve as CCN in continental air, while 10 -20% serve as CCN in maritime G Chemical composition determines the best CCN G hygroscopic G wettable G solubility

Cloud Condensation Nuclei: G CCN measurement techniques G Thermogradient diffusion chamber: G Two wetted plates are held at different temperatures, molecular diffusion not convection leads to: G linear variation of T between plates G linear variation of vapor pressure (e) between plates G Saturation vapor pressure varies exponentially with T G Saturation can be changed by changing the temperature of the two plates

Cloud Condensation Nuclei: [DRAW]

Cloud Condensation Nuclei: G World-wide measurements of CCN G continental air masses are richer in CCN than maritime air masses G Concentrations of CCN increase with supersaturation as expected G Remote ocean air contains the fewest CCN G Typically NCCN~100 cm-3 at 1% supersaturation for maritime airmasses The relationship between CCN and supersaturation is exponential:

Cloud Condensation Nuclei: G Spatial and temporal variation of NCCN G CCN can vary over several orders of magnitude over short periods of time G proximity to CCN sources G wind direction and wind speed (air mass could switch to maritime or continental) G precipitation (cloud formation depletes CCN, precipitation scavenges CCN) G CCN concentrations diminish with height away from ground; but inversions could trap CCN

Cloud Condensation Nuclei: G Properties of CCN G Theory: Na. Cl and large particles serve as CCN G Observations: not Na. Cl, but sulfides and sulfur compounds; even particles down to 0. 02 µm can serve as CCN G Type of could system can influence type of activated CCN (low S, weak verticall motion, etc. ) G In reality, atmospheric nuclei are composed of a mixture of particles G Number of CN and CCN are not well correlated

Chapter 6: Observed Microstructure of Warm Clouds

Cloud droplet distributions G CSk combined with radiative cooling through ascent G Distribution is no just due to aerosols type or air mass, but also depends on velocity and liquid water content G when drop concentrations are smaller, drops can grow larger G when there are lots of CCN, they grow smaller (competing for the water) G higher vertical velocities lead to higher concentrations G Activated spectra – bimodal distribution G non-activated spectra – mono-modal distribution G Continental: mean=11. 2 µm, mode=12 µm, more narrow droplet size-spectra G Maritime: larger mean and modal diameters, broader droplet-size spectra, but lower concentrations

Raindrop size spectra G Marshall-Palmer distribution G Slope depends on rainfall rate G Assume that N 0 can be specified G Generalized gamma distribution G The concentration of raindrops is much smaller than the concentration of cloud droplets G Rain drops are obviously much larger than cloud droplets G This implies that only a few cloud droplets make it into raindrops [LABEL]

Fog size distributions [DRAW]

Chapter 7: Theory of Cloud Droplet Growth

Growth by vapor deposition G Growth by vapor deposition (Diffusional growth) G Assumptions G steady state—no accumulation of vapor density G surface of drop is exactly saturated

Growth by vapor deposition G Heat budget for a drop growing by vapor deposition: G Internal energy: G No heat storage (du=0): G Three mechanisms for heating G 1. Condensation G 2. Molecular diffusion G 3. Radiative heating (cooling)

Growth by vapor deposition G Total heat budget (thermodynamic equation): G Combined equation for growth by vapor deposition:

Growth by vapor deposition G Combined equation for growth by vapor deposition: G Rate of change of radius decreases as drop gets bigger (doesn’t favor growth of large droplets) G Growth rate increases if saturation ratio increases G Growth rate increases over a solution G Growth rate decreases due to curvature G Radiation can either increase or decrease the growth rate G net effect of this is that drops can cool enough at the top of the could to grow by vapor deposition. Bigger droplets cool more by OLR

Growth by vapor deposition G Combined equation for growth by vapor deposition: G Assumptions made in deriving this equation: G Transfer of heat and moisture are by steady-state diffusion G The vapor density at the droplet surface is that under which the droplet persists in equilibrium G There is no disturbance of vapor field by neighboring droplets G There is no disturbance of vapor field by motion of the droplet G There is no additional source of heat to or from the droplet other than radiation G Heat storage on the droplet is negligible

Growth by vapor deposition G Combined equation for growth by vapor deposition: G Also, this is an assumption that this continuous diffusion rather than discrete. Thus we can modify the diffusion coefficient with a condensation (accommodation) coefficient, and similarly, thermal diffusion coefficient. G Large drops also ventilate, which can enhance evaporation and condensation G Large drops can also evaporate

Growth by vapor deposition G Growth example: G Narrows the droplet spectrum in time G shows 1/a dependence G solution effects enhance growth [DRAW]

Growth by vapor deposition G Growth of a population by condensation: G As drops grow, they remove S, but as air rises, S increases

Growth by vapor deposition G Growth of a population by condensation: G In fog, S is lower and only the most chemically active and huge aerosols are activated G Small drops get “starved” of H 2 O, never reach S large enough to grow G If updraft increases, peak saturation would also increase (cool air faster, takes drops long time to use up H 2 O) thus smaller drops also activate G In general, it takes days to grow drops to precipitation sizes by vapor deposition alone!!!! TOO SLOW!!

Collision-Coalescence Growth G Collection kernel: (units m 3/s) G E 1: Coalescence efficiency G E 2: Collision efficiency

Collision-Coalescence Growth G E is very small (especially for small drops) at the beginning because small drops sweep around droplet (following streamlines) G E~1 for a broad range of a 1/a 2 ratios G Can have efficiencies greater than 1 G E drops off as a 1/a 2 approaches 1 G E spikes as a 1/a 2 is very nearly 1. This is due to wake capture G wake capture: as drops are close to same size, hydrodynamic flow fields interfere and drops slip around each other. But this really doesn’t matter because when a 1 and a 2 about equal, the difference in their terminal fall speeds is so small it decreases the collection kernel. [DRAW]

Collision-Coalescence Growth G Continuous Growth Model or Accretion Model G if we also assume that a 1>>ai and v 1>>vi G Assume coalescence efficiency of unity G Continuous accretion model is applicable when collector droplet is much larger than collected droplets G Fails because it requires an initial broadening of droplet spectra to get drops large enough to be efficient collectors G A given droplet will always grow to the same size when falling through the same droplet population

Collision-Coalescence Growth G Quasi-stochastic model G Uses Smolokoskies equation to predict a unique spectrum after some time dt G Use a Monte Carlo distribution G a type of “bin” model that predicts the time rate of change of mass or volume (not radius) gain loss

Collision-Coalescence Growth G Quasi-stochastic model G Similar to aerosols, but K increases as you get to larger droplets (K decreases for smaller aerosols) G Integration limit on Gain term accounts for combinations (don’t want to double count) G Implies that higher droplet concentrations = higher rate of collection but for a given LWC higher concentrations lead to smaller droplets G ie: for same LWC, a cloud with less concentration will grow larger drops than one with more concentration G aerosol # can really affect the cloud/precipitation processes G polluted clouds => much smaller drops G Increasing the LWC can greatly accelerate the collection process

Collision-Coalescence Growth G Problem of Initial Broadening G Problem: How to get droplets to a size where collisioncoalescence can kick in G Initial droplet spectra is 4µm to 12µm, so how do we get to sizes 25 -30µm to make collision-coalescence productive? G Growth by vapor deposition tends to narrow the droplet spectrum G Need broad spectrum of sizes or else the velocities and sizes will be too similar for Collision and Coalescence collection kernel. Thus problem of initial broadening is not just creating drops large enough for Collisioncoalescence to begin

Collision-Coalescence Growth G Problem of Initial Broadening G 1) Turbulence influences on condensation growth via fluctuation supersaturations G mixing process is inhomogeneous (get pockets of clear air—spaghetti strings) G parts of cloud may be rising and others falling on a small scale leading to evaporation of some drops, leading to relatively larger drops in some areas G Fine scale eddies can centerfuge particles out of regions, increasing the S leading to faster growth of particles that remain G 2) Role of GCCN G Can act as “Coalescence Embryos” if soluble, wettable, and large G very small concentrations (similar to raindrop concentrations) G depends on CCN concentration if GCCN is important, because lower CCN clouds drizzle actively without the presence of these GCCN—maritime clouds are prolific collision and coalescence machines and GCCN presence doesn’t matter

Collision-Coalescence Growth G Problem of Initial Broadening G 3) Turbulence influences on droplet collision and coalescence G --> Enhance collision efficiencies (small drops can cross streamlines) G --> Enhance collection kernels (accelerated by air movements) G --> Producing inhomogeneities in droplet concentrations G 4) Radiative broadening G assumes droplet stays around top of cloud for a long enough time (strat/fog) G cooling decreases satruation vapor pressure at the surface of the drop, leading to faster growth than drops in the middle of the could G can offset the 1/a dependence since larger drops radiate more G limited to certain classes of clouds, but it is most easy to quantify G Broadening mechanisms are difficult to quantify because of difficulty of studying turbulence.

Drops G Drop terminal velocity G Vt is classified for 3 regimes G Vt flattens outs because drops deform G Accumulation zone or “balance zone”—in steady updraft, drops can grow to reach max terminal velocity G Drop breakup G When one drop breaks up, can get a spectrum of drop sizes G 1) Hydrodynamic instability of large drops due to natural oscillations G 2) Collisions with other drops G ring mode (collision with other drops) G bag mode (air trapped inside, blows up like baloon) G 3) Langmuir’s chain reaction theory

Drops G 3) Langmuir’s chain reaction theory G when a drop grows by collection to the point that it breaks up, it will produce fragments which are still precipitation sized. Those fragments then collect smaller droplets and be come large enough to break up, . . . G Requirements: G 1. Cloud remains in steady state for extended periods G 2. Updraft is strong enough to support lots of hydrometeors G Can be prolific in some warm clouds, leading to bursts of precipitation G Effects raindrop size-distribution G Effects vr=> effects distribution of water G Effects water loading, water in supercooled regions, etc. G Increases number of collectors for a given LWC

Summary of warm-rain G Colloidally unstable: warm-based cloud that has a large value of cloud-base saturation mixing ration and has a potential for condensing a significant amount of LW. G Colloidally stable: larger concentration of CCN and smaller saturation mixing ratio at cloud base leads to a lower potential for liquid water production and must distribute the limited LWC over more droplets, lowering its potential for creating precipitatioin. G Maritime, warm-based clouds: more likely to produce warm rain (fewer, bigger droplets and more likely to have broader droplet spectrum) G Continental, cold-based cloud: activates a larger concentration of CCN and has smaller saturation mixing ratio at cloud base. G Can really have a mixture of all of the above in real clouds

Summary of warm-rain G GCCN can move a cloud that is colloidally stable to colloidally unstable G need to know composition of pollution to understand its influences on the production of warm rain G can effect cloud base temperature G depends on size and hygroscopicity of aerosols G hard to predict if any cloud will rain on any given day. . .