Production of Ice in Tropospheric
Clouds
by Will Cantrell and Andrew Heymsfield
Presentation by:
Emily Riley
27 February 2007
Outline:
• Homogeneous Nucleation
– Theory
– Assumptions
• Heterogeneous Nucleation
– Different Types of Particles
– Preactivation
– Contact Nucleation
• Secondary Production
– Hallett-Mossop process
– Other Mechanisms
Homogeneous Nucleation Theory
Nucleation Occurs:
energy contribution from bulk > energy contribution surface
Nucleation Rate:
J = Joexp(-E/kT)
• Jo - long explanation
• E - energy barrier
– f(, , c, LH, )
• f(T)
• k - Boltzmann’s constant
Homogeneous Nucleation Theory
• Assumes:
(1) Water molecules randomly arrange into an ice
structure
(2) Liquid transitions directly to hexagonal ice
crystal
(3) occurs in pure water
(4)The initial ice fragment forms in the bulk
So, Are Assumptions Good?
Problem: Nucleation rate disagreed
with observations
Assumption (1):
• As T decreases, J
increases more than the
theory predicts! Why?
• As T decreases, the nature
of Hydrogen bonding
results in water clustering
• The clusters grow with
decreasing T, while the
bond links become more
linear
Pruppacher (1995)
Assumption (2)
• There can be an
intermediate step from
liquid to hexagonal
crystals
• Cubic crystals
– occur T < -70C
– Can cause clouds to
dehydrate more
effectively in the upper
trop.
Riikonen et al. (2000)
Murphy (2003)
Assumption (3)
• Homogeneous nucleation can occur in solution drops
– Solute becomes dissolved in water such that the
drop is like pure water
• Use a modified temperature - T* = T + T
– T - melting pt. depression
–  - Fudge factor that relates fz pt. depression to T
Assumption (3)
• Koop et al. (2000)
– J in an aqueous solution is independent of the
nature of the solute, only depends on water
activity
– Good or bad…Don’t know??
• Seifert et al. (2003a) & Cziczo et al. (2004)
– While independent of solute nature, may
depend on solute size
Assumption (4)
• Djikaev et al. (2002)
– Nucleation could be favored at the surface as
opposed to the “bulk”
– Only important for D < 1m
Heterogeneous Nucleation:
• Basics
• Particle and Surface Characteristic Studies
• Preactivation
– Two Studies
• Contact Nucleation
– One Study
Heterogeneous Nucleation:
• May significantly impact radiative properties of
clouds
• J equation still holds, but now E is decreased due
to substrate
• Common substrates - dust, fly ash, soot
• Interesting substrates - long-chain alcohols,
testosterone
Let’s look at some studies...
The importance of dust:
• Zuberi et al (2002)
– Found ammonium sulfate drops with dust
(kalonite & montmorillonite) immersed in them
had freezing temperatures 10C higher than
drops without dust
• Hung et al. (2003)
– Found ammonium sulfate drops with aerosols
(corundum & hematite) raised the freezing
temperature 6C
Let’s look at some studies...
The importance of the surface character:
• DeMott et al. (1999)
– Coated soot particles with sulfuric acid
– For T > -53C untreated, monolayered, and
multi-layered soot particles initiated ice
formation about the same
– For T < -53C multi-layered soot particles
formed ice better
Let’s look at some studies...
The importance of the surface character:
• Gorbunov et al. (2001)
– If the surface contains chemical groups capable
of forming hydrogen bonds with water
molecules, the soot’s ice-forming potential
could be increased.
– Such surfaces were 3x more efficient as ice
nuclei at -20C than surfaces without such
capabilities
Preactivation : “memory effect”
• Increase in the effective freezing T after ice
nucleator has catalyzed the phase transition once
or cooled below -40C
• Effect lost if T (of the system) exceeds some
threshold
• Occurs due to an ordered, ice like layer of water
molecules on the substrate
• Basically, aged nucleators more efficient
Yet more studies...
Preactivation : “memory effect”
• Seeley and Seidler. (2001)
– Long chain alcohols exhibit preactivation
– Alcohols could act as effective ice nucleators
above the melting point
Yet more studies...
Preactivation : “memory effect” continuted:
• Rosinski (1991) & Rosinski and Morgan (1991)
– Conditioned Particles
1. Particles exposed to SS H2O(v) at T < 0C
2. Resulting drops evaporated
3. Particles exposed to H2O(v) SS w.r.t. ice but not
H2O(l)
– Result: conditioned particles formed ice
crystals by deposition, while non-conditioned
particles did not
Contact Nucleation:
• Freezing of a drop initiated by contact with an
aerosol particle
• Substance have a different freezing threshold than
when they act as deopostion, condensation, or
immersion nuclei
– Indicates freezing mechanism different for
different modes
Secondary Production:
• Important - Ice production frequently
exceeds IN concentration
• How:
– Hallet-Mossop process (most common)
– Mechanical Fracture
– Ice multiplication
Paquita Explains the H-M
process….
Paquita Explains the H-M
process….
Paquita Explains the H-M
process….
Hallett and Mossop (1974), Mossop and Hallett (1974),Mossop (1976)
Studies confirming the Hallett-Mossop
process...
• Hogen et al. (2002)
– Observed ice crystal concentrations up to 1000 l-1 in UK
convection at T = -6C
– Typical heterogeneous IN concentrations were 3-5 orders
of magnitude lower than the measured IN concentrations
• Ovtchinnikov et al. (2000)
– Cumulus clouds in NM
• Modeling studies - confirmed these observations
So, is it always the H-M process?
• According to Hobbs and Rangno (many papers),
Nope. Why not?
– Clouds glacieated much faster than H-M process could
explain
– Crystal habits observed were often not compatible with
the T range in which H-M operates
– High concentrations of small ice particles appeared
concurrently with frozen drizzle drops
• Conclusion - Origin of ice is a mystery…sounds
like a job for more research!
Other Mechanisms:
• Mechanical Fracture
– Occurs particularly in dendritic crystals
• Ice multiplication
– Occurs when large drops shatter upon freezing
Conclusion:
• Homogeneous Nucleation - Best
Understood
• Secondary Production - Somewhat
Understood
• Heterogeneous Production - “Progress is
desperately needed”