Why the aerosol are important for clouds

1
Why the aerosol are important for clouds and Why
clouds are important for the aerosol Jeff
Snider University of Wyoming, Laramie jsnider_at_uwyo
.edu
2
Outline - Basic physics of cloud/aerosol and
aerosol/cloud interactions Properties of aerosol
particles that make them good nuclei Cloud
condensation nuclei (CCN) Ice nuclei Aerosol-to-
CCN closure studies Populations of aerosol and
hydrometeors (i.e., droplets, drops and
crystals) Importance of clouds to the aerosol
3
(No Transcript)
4

• In the atmosphere, H2O vapor is often
  subsaturated
• However, saturated conditions exist within clouds
• Some cloud environments can be supersaturated
• Vapor state (subsaturated, saturated or
  supersaturated) is quantified with vapor pressure
  or absolute humidity (or vapor mixing ratio)

5

• Below 0 oC, interesting thermodynamic effects
  take place
• Ice is the stable phase, but liquid water can
  coexist
• Like water, ice has a temperature-dependent
  saturation vapor pressure
• The saturation vapor pressure over liquid water
  exceeds that over ice

6

• Below 0 oC, the saturation vapor pressure over
  liquid water is larger than that over ice
• From the perspective of the ice hydrometeors,
  mixed phase clouds are supersaturated. The ice
  can grow by deposition
• The water saturation ratio increases by 10 for
  every 10 degree of supercooling

7
Bergeron, 1935 - Points to the importance
mixed-phase cloud, the favored growth of ice in
such environments, and consequence for
precipitation
8
Description of conditions in warm
clouds- Thermodynamic - water vapor is in
equilibrium with liquid water Dynamic - water
vapor amount exceeds equilibrium
Description of conditions in cold
clouds- Thermodynamic - water vapor is in
equilibrium with ice Dynamic - water vapor amount
exceeds equilibrium (Note often liquid water and
ice coexist....so-called mixed phase clouds)
9
How to apportion the latent heat?
Rate of vapor mass transfer to a droplet of
diameter D
Rate of sensible heat mass transfer to a droplet
of diameter D
Combining, and linearizing the Claussius-Clapeyron
equation-gt Maxwell-Mason droplet growth equation
10
Köhler theory provides two things 1) A
connection between wet diameter and saturation
ratio at the interface 2) A connection between
critical saturation ratio and dry diameter
11
We have two equations that account for single
particle growth via condensation- 1) Droplet
size is related to time and ambient
conditions The Maxwell-Mason Equation 2) Bdry
condition at the droplet/air interface is related
to properties of the nucleus The Koehler Equation
12
Characteristic of an active cloud condensation
nucleus -Large particle size -Contain
materials (solute) that dissolve in
water -Contain solutes that dissociate in
solution -Contain many solute molecules -Contain
solutes that reduce the energetic cost of
forming an interface
13
Some active ice nuclei have lattice dimensions
similar to ice-
Substance a axis dimension, nm c axis dimension, nm Temperature to nucleate ice, oC
Ice 0.45 0.74 0
Substance a axis dimension, nm c axis dimension, nm Temperature to nucleate ice, oC
AgI 0.46 0.75 -4
CuO 0.47 0.51 -7
Kaolinite 0.52 0.72 -9
Substance a axis dimension, nm c axis dimension, nm Temperature to nucleate ice, oC
Bacteria -- -- -3
14
(No Transcript)
15
Crystal concentrations observed in some clouds do
depend on temperature in a manner consistent with
generation via nucleation But, information is
needed for describing the connection between ice
nuclei sources, nuclei activity spectra and ice
crystal concentrations. This is lacking
16
With ice there are additional complications Ice
can form via secondary processes Collision
between graupel and snow Shattering of freezing
drops Ice from one cloud can also "seed" a
neighboring cloud
17
Aerosol-to-CCN Closure Studies
How well do the predicted and observed CCN
concentrations compare?
18
Why Closure Studies?
Chemical Transport Model use a mass balance,
constrained by aerosol source and sink processes,
to derive aerosol size spectra.
Parameterization, based on observation, is often
used in the models.
Questions How reliable are the observational
data sets used in the parameterizations? Does
systematic error in the measurements alter the
sign or sensitivity of the model prediction to
alterations in aerosol properties? Under what
circumstances are the simplifying assumptions OK?
A common assumption is that the particles are
spherical, often they are not
Picture from Alexei Kiselev
19
The Wyoming static diffusion CCN Instrument
20
DMT (Scrips/Caltech/GT) Continuous-flow CCN
Instrument
21
Two CCN instruments
Developer Chamber type Operating
principle Calibration CCN detection Non-ideality
1 Non-ideality 2 Non-ideality 3
University of Wyoming Static Non-linearity of
ew(T) Snider et al. (2006) Scattering from
ensemble of droplets Temp. difference between
outer and inner wall Wall material may alter
ew(T) Activation spectrum broadening
Developer Chamber type Operating
principle Calibration CCN detection Non-ideality
1 Non-ideality 2
Droplet Measurement Technologies (DMT) Continuous
flow Dissimilarity of vapor and heat
diffusivities Lance et al. (2006) Single particle
scattering Temp. difference between outer and
inner wall Concentration bias
22
Wyoming Static Thermal Diffusion CCN Instrument
23
DMT (Scrips/Caltech/GT) Continuous-flow CCN
Instrument
Aerosol flow stream is surrounded by sheath
flow H2O vapor diffuses (inward) faster than
sensible heat Maximum supersaturation is near
exit to OPC Activated (growing) CCN are counted
in OPC Resistance to heat flow across
wall Efficiency (Th'-Tc')/(Th-Tc) 0.7
Efficiency is evaluated in laboratory studies
24
DMT Calibration Th-Tc 5.35 oC, Qtot0.5 L/min,
P 0.8 atm, SAR 10
Particles of known size and composition are
produced in a DMA Ammonium sulfate is preferred,
but there are issues Koehler theory used to infer
particle SSc from the DMA-selected Dd Small test
particles (i.e. particle SSc gt maximum chamber
SS) -gt no activation Large test particles (i.e.
particle SSc lt maximum chamber SS) -gt complete
activation Activated fraction 0.5 defines the
maximum chamber SS Outer wall temperatures,
efficiency, chamber model -gt from max chamber SS
25
(No Transcript)
26
Snider et al., in press, Journal of Atmospheric
and Oceanic Technology
27
Snider et al., in press, Journal of Atmospheric
and Oceanic Technology
28
Wyoming calibration ?T 2.2 oC, Tt20 oC, P
0.8 atm
From measurement of activated fraction
versus sphere equivalent diameter, the size at
50 activation is determined - gt D50
Snider et al., in press, Journal of Atmospheric
and Oceanic Technology
29
We concluded that the supersaturation determined
from temperature measurement, and a model of the
chamber, that the nominal supersaturation is a
factor of 1.6 larger than the supersaturation
evaluated from particle size and a Koehler
model The cause of this discrepancy...
Snider et al., in press, Journal of Atmospheric
and Oceanic Technology
30
DMT / Wyoming CCN comparison experiments in
Leipzig, November 2005 Simple (ammonium sulfate)
and complicated (soot-coated) test
particles Determinations of Critical
Supersaturation in both instruments Table shows
statistics (average and standard dev) for the DMT
/ Wyoming Ratio
Aerosol Ammonium Sulfate Soot Coated with Ammonium Sulfate Soot Coated with Levoglucosan
of Experiments 8 6 3
Average 1.01 1.01 1.03
Standard Deviation 0.06 1.10 -----
31
Droplets1000 and 50 cm-3, No Ice, zbase 500 m,
Tbase -10 oC
32
The physics of S(t) in a parcel model -
LWC lags the adiabatic liquid mixing ratio, and
more so when there are fewer droplets, i.e., 50
cm-3 In an adiabatic parcel their is no
supersaturation, vapor and liquid are at
equilibrium. In other words the supersaturating
effect of the cooling is exactly balanced by the
formation of liquid (dS/dt 0, and S1) It
follows that the initial rate of increase of S is
larger for the parcel with fewer droplets,
compared to the parcel with more droplets
33
The physics of S(t) in a parcel model (continued)
Also, the characteristic time for adjustment to a
steady state is longer in the case of the of the
simulation with fewer droplets, i.e., 50
cm-3 Hence, the peak saturation ratio is larger,
and it occurs higher in the cloud when there are
fewer droplets, i.e., 50 cm-3
34
Have I contradicted myself? Large cloud droplet
concentration -gt small maximum saturation
ratio Small cloud droplet concentration -gt large
maximum saturation ratio Does this imply that an
increase in nuclei (i.e., pollution) will
decrease the maximum saturation ratio enough to
decrease the droplet concentration? I.e., causing
a reverse of the first indirect effect of aerosol
on climate?
from Snider et al., JGR, 2003
35
Parcel Model Calculation - Sensitivity of droplet
concentration to aerosol and updraft
Parcel Model
w
Mcfiggans et al., Atmospheric Chemistry and
Physics, 2006
36
Droplets50 cm-3, Ice100 L-1, zbase 500 m
Tbase -10 oC
37
Cloud, and especially precipitation associated
with clouds, has a profound impact on the
aerosol!
Aerosol are removed from the atmosphere by
precipitation Coalescence scavenging Aerosol
scavenging by precipitation falling below
cloud Aerosol attachment to cloud and precip via
brownian motion Aerosol number concentration can
be decreased even if precipitation evaporates On
average there is a steady state between aerosol
source and aerosol removal In some cloud regimes
there is an imbalance between source and sink
Marine stratocumulus
38
Marine summertime clouds - DYCOMS-II
(2001) Marine stratus, July, cloud top
temperatures gt 0 oC, 300 km west of
California Aerosol Source Processes - Wind
speeds in the marine boundary layer (MBL) lt 10
m/s Characteristic time for sea salt aerosol
source 10 day Entrainment of free troposphere
(FT) into MBL characterized using tracers
Characteristic time for entrainment of FT aerosol
into MBL 10 day Aerosol Sink Processes
- Drizzle rates were surprisingly large (10
mm/day, 100 mm/day locally!) Coalescence
scavenging thought to dominate Aerosol source
rates lt Aerosol sink rates Aerosol
concentrations decrease Aerosol surface area
decreases, a threshold is reached, new particle
formation occurs Evidence for new particle
formation in the MBL on July 11, 2001
(RF02) Heavy drizzle, open-cell cloud structure
also documented in new particle region
39
Leon et al., Journal of Atmospheric and Oceanic
Technology, 2006
40
Drizzle is most intense in regions of rising air
motion (w 1 m/s) Ascent is driven by horizontal
convergence at the base of the MBL Coupling of
ascent and drizzle implies longer drizzle growth
times and enhanced removal of cloud droplets
(drizzle scavenging) compared to drizzle
formation elsewhere in the MBL
Leon et al., Journal of Atmospheric and Oceanic
Technology, 2006
41
We concluded 1) New aerosol particles
were formed in response to a depletion of the
preexisting aerosol surface area by
heavy drizzle. 2) Organization of the cloud
into open cell structures may be either
necessary for new particle formation or a
consequence of it.
Petters et al., Journal of Geophysical Research,
2006
42
Concluding Remarks - 1. Aerosol size spectra and
number concentration are influenced by
precipitation and this in turn influences the
properties of clouds 2. No measurement is
perfect, but through intercomparison
instrument bias uncovered and accounted for. 3.
Models need data sets for parameterization and
also for initialization. These data sets should
be as free of measurement bias as is
possible. 4. Through collaboration we reach our
objectives sooner and with greater understanding
of the consequence of our efforts. Acknowledgemen
ts - The group at Warsaw (Hanna, Tymon,
etc.) Markus Petters (Colorado State
University) David Leon (University of Wyoming)