Production of ice in tropospheric clouds

Production of ice in tropospheric clouds
Alexander Lowag
University of Miami


Motivation:

- ice: influence on various atmospheric processes, including:
+ precipitation
+ cloud electrification
+ radiative transfer

- liquid phase: number + chemical composition of aerosols --> number
of cloud droplets

- solid phase: physical / chemical principles of nucleation poorly understood

- secondary mechanisms: enhancement of ice concentration (number /
mass) , mechanism not completely known


- homogeneous nucleation: proceeds from a random event in the pure
liquid or solution without catalysis from foreign substance
- classical nucleation theory: energy barrier can be calculated under
the assumption that microscopic ice fragment forms in supercooled
liquid
- nucleation rate: J = A * exp (-∆E / kT)
- calculated rates in contradiction with measured ones (water droplets
at – 40º C)
- where is the mistake? tenet of theory? inputs to it?

- to calculate ∆E, density, viscosity, specific and latent heat and surface
tension must be known
- Pruppacher (1995): singularity of water at -45º C, extrapolation of
data aquired above 0º C
However: liquid water at -70º C (Bartell and Huang, 1994)
- Jeffrey and Austin (1997): analytic equation for state of water to
obtain ∆E



- homogeneous nucleation in Cu and Ci despite soluble aerosol
particles
- solute depresses melting and freezing point of ice
- modified temperature to obtain nucleation rate: T' = T + µ * dT
- Koop et al. (2002): nucleation rate only depends on water activity, not
on solute
-> Moehler et al. (2003) : freezing rate of sulfuric acid droplets
-> but: spread in empirically determined µ outside experimental
error margins (DeMott, 2002)


- classical nucleation theory based on assumptions that may prove to
limit its usefulness:
+ fragment initiating nucleation forms in the bulk liquid
However: nucleation rate should be proportional to volume
of liquid and not to how it is dispersed
+ embryo has same properties as bulk crystal (liquid to hexagonal
ice without intermediate stage)
However: nanometer-sized droplets at -70º C have cubic

structure
- cubic structure not stable -> consequences for ice in atmosphere


- heterogeneous nucleation: freezing occurs via catalysis by a foreign
body (T > -33º C)
- typical aerosols: mineral dust, emissions from aircraft (soot)
- equation for previous nucleation rate still applies, but ∆E is lower due
to substrate
- substrate forms a template on which water molecules adsorb in a
configuration similar to the crystaline structure of ice
- influence on radiative properties of Ci
- dominating nucleators: mineral dust, fly ash, metallic particles


- Zuberi et al. (2002): investigation of kaolinite and montmorillonite
-> freezing points of droplets with immersed dust 10 º C higher
- Hung et al. (2003): hematite and corundum aerosols in ammonium
sulfate solution
-> freezing point increased by 6º C
- De Mott et al. (1999): investigation of ice-forming activity of black
carbon particles
-> untreated particles acted as deposition / absorption nuclei for
T < -42º C
-> activation characteristics of treated particles were similar to
that of untreated ones
-> multiple layers: as well as or even better than untreated
particles

- Soot:
- surface character can have significant impact on its ice-nucleating
efficiency
-> chemical groups capable of forming hydrogen bonds increasing
ice-forming potential
-> particles with OH and carbonyl groups three orders of
magnitude more efficient than those without
- morphology of particle also important: “liquid” water down to
-73º C


- films of high-molecular-weight organic compounds
-> long-chain alcohols, testosterone
-> formation of a 2D hexagonal-like crystal at the air-water
interface
Memory effect (Preactivation):
- increase in effective freezing temperature, lost if temperature exceeds
some threshold typically close to melting temperature
- Seeley and Seidler (2001):
-> long-chain alcohols exhibit preactivation
-> ice formation / cooling below melting temperature not
required, i.e. layer already present at room temperature


Rosinski (1991): nucleation and evaporation of droplets
-> ice crystals due to deposition
- deactivation of ice nuclei: loss of ice-nucleating ability after
sublimation of ice
Contact nucleation: freezing of a droplet initiated by contact with an
aerosol particle

- many substances have different freezing thresholds as a contact nuclei
compared to being a condensation or immersion nuclei
-> different freezing mechanisms for different modes
- Seeley and Seidler (2001): examination of freezing characteristics of
volcanic ash



Secondary production:
- Hallett-Mossop process: production of splinters during riming


- consistent with ice concentration measured in Cu
- could be varified in many cases:
-> within convection in a frontal system off the coast of the
UK (Hogan et al., 2002)
-> in Cu in Mexico (Ovtchinnikov, 2000)
-> ice concentrations higher than what could have been
produced by Hallett-Mossop process


Outstanding problems:
- is freezing only a function of water activity?
-> ammoniated compounds seem to be an exception
- what is the mechanism underlying contact nucleation?
- secondary production mechanisms in the presence of liquid water
other than riming-splinting?
“In summary, the origin of ice in cumuliform clouds remains a mystery”

(Rangno and Hobbs, 1994)