Aerosols and the Climate System
Peter J. Adams
Center for Atmospheric Particle Studies (CAPS)
Department of Civil and Environmental Engineering
Department of Engineering and Public Policy
Carnegie Mellon University
Alpine Summer School: Climate, Aerosols, and the Cryosphere
20 June 2012
2
Outline
• Introduction and Overview
• Aerosols: Size, Chemistry, Behavior
• Aerosol Climate Impacts
• Bigger Picture Context
• Aerosols: a major uncertainty in climate science
• Research Challenges
• (Aerosol Properties)
• Mie theory: optics; aerosol-sunlight interactions
• Kohler theory: aerosol-cloud interactions
3
Atmospheric Particles (“Aerosols”)
SEM Images: Pittsburgh Air Quality Study Gary Casuccio, R.J. Lee Group, Monroeville, PA
Small Suspended Particles -- few nm to 10s mm -- Complex shapes -- >1000s compounds -- Multiphase -- Many sources
Diesel Soot
1 mm1 mm1 mm1 mm
Coal Combustion
0.2 μm0.2 μm0.2 μm
0.2 μm0.2 μm0.2 μm
Biological
4
Introduction
• Formation mechanisms • Primary: particles directly emitted by a source (e.g. smoke from
combustion)
• Secondary: gas-phase oxidation products that form particles
• Composition • Wind-blown: Mineral dust, sea spray
• Combustion: elemental and organic carbon
• Atmospheric chemistry: sulfate, nitrate, secondary organic carbon
• Size • Diameters are ~1 nm to ~10 mm
• Most behavior is size-dependent
• Growth by condensation and coagulation important
5
Atmospheric Particles: A Primer
Nucleation Mode
(1-10 nm)
Ultrafine Mode
(10-100 nm)
«Aitken» mode
Fine Mode
(0.1-1 mm)
«Accumulation»
mode
Coarse Mode
(1-10 mm)
Cloud Drops
(10+ mm)
6
Atmospheric Particles: A Primer
Nuc Mode
(1-10 nm)
UF Mode
(10-100 nm)
Fine Mode
(0.1-1 mm)
Coarse Mode
(1-10 mm)
Cloud Drops
(10+ mm)
10 nm 100 nm 1 mm 1 nm 10 mm
contributes most to aerosol
number concentrations
surface area
mass conc.
always when present when present
7
Sources of Particles
• Wind-blown
• Primary (directly emitted) particles
• Sea-salt and mineral dust
• Larger than 1 mm (mostly)
• Aerosol nucleation (new particle formation)
• Clustering of supersaturated gases to form particles
• Secondary particle formed in atmosphere from
precursors: H2SO4(g) (...and other species)
• Smallest identifiable «particle» is 1 nm
• Combustion
• «Primary» particle directly emitted by a source
• ...really nucleation in or near source
• Sizes typically 30-300 nm
8
Atmospheric Particles: A Primer
Nuc Mode
(1-10 nm)
UF Mode
(10-100 nm)
Fine Mode
(0.1-1 mm)
Coarse Mode
(1-10 mm)
Cloud Drops
(10+ mm)
nucle
ation combustion
win
d-b
low
n
9
Knudsen Number
• Continuum regime (Kn << 1; coarse mode)
• air may be treated as a continuum (standard fluid
dynamics and mass/heat transport)
• Transition regime (Kn ~ 1; fine mode)
• empirical correction factors
• Kinetic regime (Kn >> 1; ultrafine mode)
• particles small → kinetic theory of gases
Dp
l
Dp: Particle
diameter
l : Mean free path
of air molecules pD
Knl2
10
Condensation: Kinetic Regime
spk cccRJ 2
c
Particle
Condensing Gas Jk: Net flux to one
particle
(condensation
minus evaporation)
: accomodation
coefficient
Rp: Particle radius
: Mean velocity of
condensing gas
: bulk
concentration of
condensing gas
cs: “surface”
(equilibrium)
concentration of
condensing gas
cflux (per unit area) particle cross
sectional area
Comments:
•J is condensation rate to particle (molecules s-1)
•J ~ Rp2 in kinetic regime
11
Condensation: Continuum Regime
• Diffusive flux from bulk gas to particle surface
(or vice versa) governs transport
• Since flux per area ~ 1/Rp, overall condensation
rate is ~Rp (not Rp2)
Particle )(4 sgpc CCDRJ
CsC
Jc = (diffusive flux per area)(surface area) ~ Rp2 (?)
Fick’s Law: (diffusive flux per area)
p
ssg
R
CC
L
CC
r
cD
L (thickness of
diffusive layer)
12
Condensation: Particle Growth Rates
• More useful to think about particle growth rates
• GR = dDp/dt (diameter change with time, nm/h)
• Depends on condensable gas concentrations
• GR const w/ Dp (small particles; kinetic regime)
• GR ~ Dp-1 (large particles; continuum regime)
• A moderately fast growth rate
• GR = 5 nm/h for kinetic (...lower for continuum)
• Expected growth due to condensation
• = (5 nm/h) x (1 week) = 800 nm (0.8 mm)
• Nucleation and ultrafine mode particles can
grow to fine mode
• Fine mode does not grow to coarse mode
13
Atmospheric Particles: A Primer
Nuc Mode
(1-10 nm)
UF Mode
(10-100 nm)
Fine Mode
(0.1-1 mm)
Coarse Mode
(1-10 mm)
Cloud Drops
(10+ mm)
nucle
ation combustion
win
d-b
low
n
cond. cond.
• Note: condensation arrows show behavior of individual particle counts (aerosol number distribution)
• Looking at mass budgets of condensable gases
• Condensation ~ total particle surface
• ...mostly to fine mode (some UF and coarse)
14
Coagulation
• Basic expression for coagulation rate between
two particle size classes
jiijcoag NNKJ
Jcoag: coagulation rate (collisions cm-3 s-1)
Ni and Nj: number concentrations (cm-3) of particles in each size class
Kij: coagulation coefficient for size class i with size class j
• Note similarity to bimolecular reaction rate, r=k[A][B]
• Functional form of Kij depends on physical processes causing collision • Brownian motion (main thing for atmospheric aerosols)
• Particle diffusion is the most important process
15
Coagulation
• Brownian coagulation coefficient, Kij
• Kij maximum between very small (highly diffusive/mobile) and very large (large cross sections) particles
• → “coagulational scavenging” of ultrafines by accumulation mode (and coarse if present)
• Resulting particle lifetimes w.r.t coagulation
• Ultrafine particles: 1-10 hours
• Fine mode: few days
21212 DDDDK ppij
Dpi: particle diameters
Di : particle diffusivities
: non-continuum correction factor
particle size (target area) particle mobility
16
Atmospheric Particles: A Primer
Nuc Mode
(1-10 nm)
UF Mode
(10-100 nm)
Fine Mode
(0.1-1 mm)
Coarse Mode
(1-10 mm)
Cloud Drops
(10+ mm)
nucle
ation combustion
win
d-b
low
n
cond. cond.
coag.
coag.
coag.
17
Dry Deposition: General Observations
• Ultrafine mode
• Diffusion gives higher vd than accumulation mode
• But coagulation tends to dominate losses under most
conditions
• Accumulation mode • Minimal deposition velocity
(timescale is ~months)
• Wet deposition matters more
• Coarse mode • Dry deposition is dominant
removal (timescale ~1 day)
Depositio
n V
elo
city
Particle Diameter
Acc. Ultra
-fine Coarse
18
Atmospheric Particles: A Primer
Nuc Mode
(1-10 nm)
UF Mode
(10-100 nm)
Fine Mode
(0.1-1 mm)
Coarse Mode
(1-10 mm)
Cloud Drops
(10+ mm)
nucle
ation combustion
win
d-b
low
n
cond. cond.
coag.
coag.
coag.
dry
dep
dry
dep
19
Cloud Condensation Nuclei (CCN)
• In a particle-free atmosphere, a strong supersaturation
(~400% relative humidity) is required to nucleate new,
pure water droplets
• Instead, cloud water condenses onto pre-existing
particles: cloud condensation nuclei (CCN)
Clear Sky
(RH < 100%) Cloudy Sky
(RH > 100%)
CCN
(~100 nm)
Other
particles
(aerosols) Cloud
droplets
(~10 mm)
Activation:
water condenses
on CCN to form
cloud droplets
20
Cloud Processing
• Only larger (100 nm or more) and more soluble
particles activate into cloud droplets
• From aerosol point of view, fate is
• cloud droplet precipitates → aerosol wet deposition
cloud droplet evaporates → «cloud processing»
• Cloud processing
• Aqueous chemistry • SO2 oxidation into sulfate
• formation of secondary organic aerosol (SOA) in aqueous?
• Other • Scavenging of «interstitial» (unactivated) aerosol particles
• Reduction in N concentration due to cloud collection proc.
21
Atmospheric Particles: A Primer
Nuc Mode
(1-10 nm)
UF Mode
(10-100 nm)
Fine Mode
(0.1-1 mm)
Coarse Mode
(1-10 mm)
Cloud Drops
(10+ mm)
nucle
ation combustion
win
d-b
low
n
cond. cond.
coag.
coag.
coag.
dry
dep
dry
dep
activation
cloud proc.
wet
dep
22
Size
• Nucleation mode (Dp < 10 nm) • Source: formation of new particles from super-saturated vapors
• Sinks: coagulate w/ larger particles or grow to Aitken mode by condensation (t ~ hours)
• Aitken mode (10 nm < Dp < 100 nm) • Sources: growth of nucleation mode, combustion emissions, (some) sea-
spray
• Sinks: coagulate w/ larger particles, grow to accumulation model by condensation, removal by precipitation (wet deposition) (t ~ hours to days)
• Accumulation mode (100 nm < Dp < 1 mm) • Sources: growth of Aitken mode, combustion emissions, sea-spray, some
mineral dust
• Sinks: removal by wet deposition (major) and dry deposition (minor) (t ~ 4 to 8 days)
• Coarse mode (1 mm < Dp < 10 mm) • Sources: wind-blown sea-spray and mineral dust
• Sinks: dry deposition (major) and wet deposition (minor) (t ~ 1 to 3 days)
23
Aerosol Mass Budgets
(Tg/yr) (days) (Tg)
Aerosol Source Lifetime Burden
EC 7.7 8.1 0.2
Sulfate 120 5.7 1.9
Organics 190 4.7 2.5
Sea-salt 1800 0.8 3.9
Mineral Dust 2400 2.7 17.6
Burden = Source x Lifetime
• Coarse-mode particles
• large sources • Short-lived
(gravitational settling)
• Little or no anthropogenic contribution
24
Aerosol Mass Budgets
(Tg/yr) (days) (Tg)
Aerosol Source Lifetime Burden
EC 7.7 8.1 0.2
Sulfate 120 5.7 1.9
Organics 190 4.7 2.5
Sea-salt 1800 0.8 3.9
Mineral Dust 2400 2.7 17.6
Burden = Source x Lifetime
• Fine-mode particles • Medium sources • Longer lifetimes • Large anthropogenic
contribution
25
Aerosol Mass Budgets
(Tg/yr) (days) (Tg)
Aerosol Source Lifetime Burden
EC 7.7 8.1 0.2
Sulfate 120 5.7 1.9
Organics 190 4.7 2.5
Sea-salt 1800 0.8 3.9
Mineral Dust 2400 2.7 17.6
Burden = Source x Lifetime
• Small source • Long lifetime • Major source of
anthropogenic absorption
26
Sulfur Cycle
DMS
(dimethyl
sulfide)
SO2(g)
H2SO4(g)
SO42-
(aerosol)
OH, NO3
radicals
OH radicals
condensation
onto existing
particles
nucleation of
new particles
SO2(aq)
dissolution
into cloud
drop
SO42- (aq)
H2O2
evaporation
of cloud drop
•DMS emitted by phytoplankton (10-20 Tg S/yr)
•SO2 emitted mostly by power plant combustion (70 Tg S/yr), volcanos (5 Tg S/yr)
•SO2 and sulfate undergo both dry/wet deposition
27
Organic Aerosol
emissions
Primary organic aerosol (POA) is
directly emitted in particle phase
Oxidation of VOCs
creates secondary
organic aerosol (SOA)
Volatile organic compounds
(VOCs) are gas-phase emissions
28
Aerosol Cycles
• Coarse, wind-blown particles
• Sea-salt and mineral dust
• Wind-blown emissions
• Large dry deposition, some wet deposition
• Elemental carbon
• No chemical production/loss
• Budget is simple: emissions = deposition
• «mixing state» (e.g. coated with other species?)
affects absorption efficiency
• effective as an ice nuclei?
29
Size and Composition
Nuc Mode
(1-10 nm)
UF Mode
(10-100 nm)
Fine Mode
(0.1-1 mm)
Coarse Mode
(1-10 mm)
Cloud Drops
(10+ mm)
condensable gases:
sulfate, nitrate,
secondary organic
aerosol
onto aerosol surface
area
combustion: EC
and primary
organic aerosol
wind-blown: NaCl and
crustal (Si, Ca, Fe)
30
Aerosol Number Budget (highly uncertain!)
Ultrafine Mode Fine Mode (CCN)
Emissions = 58
cm-3 day-1
100 nm
Growth = 12.2 cm-3
day-1
Deposition = 9
cm-3 day-1
Coagulation = 52
cm-3 day-1
Emissions = 3.2
cm-3 day-1
Deposition =
15.4 cm-3 day-1
GISS-GCM; global average; Sulfate, sea-salt, carbonaceous, dust
(Binary) Nucleation
J1 = 310 cm-3 day-1
J10 = 14 cm-3 day-1
Number
Size
31
Outline
• Introduction and Overview
• Aerosols: Size, Chemistry, Behavior
• Aerosol Climate Impacts
• Bigger Picture Context
• Aerosols: a major uncertainty in climate science
• Research Challenges
• (Aerosol Properties)
• Mie theory: optics; aerosol-sunlight interactions
• Kohler theory: aerosol-cloud interactions
32
Aerosols Scattering Sunlight
Dust and smoke over Australia (Terra)
33
Aerosols Absorbing Sunlight
Kuwaiti oil fires
photo courtesy of Jay Apt (via Steve Schwartz)
34
Clouds and Climate
Atmosphere
Earth
•fluxes are W m-2
•Width of arrow proportional to flux
342
77
30
168
67
• 23% of incoming sunlight reflected by atmosphere (mostly by clouds)
• Without cloud reflection, the Earth would be ~15° C warmer
(absorbed)
(reflected)
35
photo courtesy of Amy Sage
Cloud Optics: Surface Area
In clouds, reflection and scattering are
proportional to surface area
36
Aerosol Cloud Reflectivity Effect
Polluted air mass
•Higher CCN concentration
•Lower Transmittance
•Higher reflectivity
•Less precipitation?
•Longer cloud lifetime?
Clean air mass
•Lower CCN concentration
•Higher Transmittance
•Lower reflectivity (albedo)
•Better chance of
precipitation?
•Shorter cloud lifetime?
37
photo courtesy of Amy Sage
Cloud Optics: Surface Area
For a given amount of liquid water (or ice):
• More pollution/CCN → More cloud droplets → More surface area
• → More scattering → Brighter cloud → Cooler Earth
Brighter polluted
cloud
(More CCN)
Darker clean
cloud
(Few CCN)
38
Aerosols and Clouds
AVHRR satellite “false color” image
Red: darker clouds (large droplets)
Green: brighter clouds (small
droplets)
Blue: clear sky
Power plant
Lead smelter
Port
Oil refineries
Rosenfeld, Science (2000)
39
How direct is direct?
• Direct effect: scattering/absorbing sunlight
• Semi-direct effect:
• aerosol absorption heats atmospheric layer
• warmer air → lower relative humidity → less/no cloud
• Indirect effect: modifying cloud properties
• “brightness (first) effect”
• “lifetime (second) effect”
40
Forcing
Shortwave Longwave
Top-of-
atmosphere
(TOA)
Tropopause
Surface
Shortwave Longwave
Reference Atmosphere Perturbed Atmosphere
41
Forcing
• Change from reference to perturbed state • Anthropogenic forcing: e.g. pre-Industrial to now
• Future forcing: e.g. 2000 to 2100
• All-aerosol forcing: with vs without aerosols (includes natural
aerosol forcing, e.g. satellite)
• At an altitude • TOA or tropopause usually
• Surface forcing for hydrological cycle impacts
• Wavelength range • Except dust, aerosols generally have minimal impacts on IR
fluxes (so SW is usually whole story)
• Without feedbacks* • for GHGs, common to account for stratospheric adjustment
• *semi-direct and cloud lifetime require clouds to change
42
Forcing
• Ultimately, forcing is (or is not) a useful quantity to the extent
that it is a predictor of global temperature change
• Key parameter is l,“climate sensitivity”
• Implicit assumption is that sensitivity
• does not depend on the kind of forcing (GHGs, ozone, aerosols, etc)
• does depend on the particular climate model (stronger or weaker climate
feedbacks)
• Generally, aerosol forcings obey this assumption except
• black carbon absorption (triggers cloud changes, i.e. semi-direct effect)
• need to take care in defining/calculating cloud lifetime forcing
FT lglobal average
temperature
change
global average
radiative forcing
43
Climate “Forcing”
Radiative Forcing (W m-2)
Long-lived GHGs:
+2.6 W/m2 (+/- 10%)
Aerosol effects:
strong, but
uncertain cooling
44
Climate “Forcing”
Radiative Forcing (W m-2)
On the chart:
• Direct effect
• Cloud albedo (1st)
indirect
Not on chart:
• Semi-direct
(forcing or
feedback)
• Cloud lifetime
(2nd) indirect
45
Aerosol Forcing Constraints
• «Forward» calculations
• Emissions → aerosol conc → forcing
• Potential to give very strong cooling (i.e. more than
offsetting GHG forcing)
• «Reverse» calculations
• Observed T → what net forcing required?
• Net anthropogenic forcing on IPCC chart really
more constrained by reverse calculations than
forward calculations
46
Surface Dimming and Hydrological Cycle
• Reduces sunlight reaching surface
• Surface forcing can be greater than top-of-atmosphere (TOA)
• Slows evaporation of water from surface
• Drier climate, reduced hydrological cycle
Ramanathan et al. 2001
47
Aerosols: Other Effects
• Snow/Ice albedo modification
• Black carbon («soot») on snow and ice causes them
to darken
• May have high regional importance for melting
glaciers and ice pack
• Nutrient deposition to ocean
• Some oceanic ecosystems limited by various
micronutrients (e.g. Iron)
• Deposition of iron (dust, geoengineering) can
stimulate ocean biota, uptake of atmospheric C
• Ice nuclei
• Ice formation requires an aerosol surface (ice nuclei,
e.g. mineral dust) between 0 and -40 °C
48
Outline
• Introduction and Overview
• Aerosols: Size, Chemistry, Behavior
• Aerosol Climate Impacts
• Bigger Picture Context
• Aerosols: a major uncertainty in climate science
• Research Challenges
• (Aerosol Properties)
• Mie theory: optics; aerosol-sunlight interactions
• Kohler theory: aerosol-cloud interactions
49
• Global warming of 1.1 to 6.4°C by 2100 predicted
50
Climate Change Uncertainty
• “Climate sensitivity” is a key parameter
• Key parameter is l,“climate sensitivity”
• Blackbody Earth (no feedbacks): 0.2 to 0.3 °C per W/m2
• Climate models: 0.3 to 1 °C per W/m2 (1.5 - 4.5 °C for 2xCO2)
• Water vapor feedback and ice albedo feedback increase sensitivity
• In climate models, representation of cloud feedback is largest
source of uncertainty
• Uncertainty in aerosol forcing plays a key role in
• future projections
• interpretation of past
FT lglobal average
temperature
change
global average
radiative forcing
51
Strategies for Determining Climate Sensitivity
• Strategy 1: Climate Models (Theory) • Use a computer simulation to predict how much
temperature will change per amount of forcing
• Basic chemistry/physics of climate system are well known
• Problem is cloud feedback (parameterizations)
• Strategy 2: Earth in the past (Observations) • Example: Industrial Revolution (1800) to now
• We know greenhouse gas forcing very well
• We know temperature increase pretty accurately
• This tells us climate sensitivity
• Problem is concurrent changes in reflecting “haze” particles (aerosols)
52
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5
Forcing (W m-2
)
Tem
pera
ture
Ch
an
ge (
K)
Aerosols and Climate Uncertainty
High
sensitivity
Low
sensitivity
GHG forcing
20th century T
increase
Aerosol (haze)
+ GHG forcing
53
Two Uncertainties: Forcing / Sensitivity
Andreae et al. 2005
T = (high sensitivity) x (low forcing)
T = (low sensitivity) x (high forcing)
To be consistent with past warming, must be
within dotted lines
54
Climate Models: Sensitivity / Aerosols
Figure from Kiehl et al., GRL v34
doi:10.1029/2007GL0313832007
55
Climate Sensitivity and Aerosols
• Climate models “hindcast” the 20th century warming of ~0.6K • High sensitivity; low total forcing (strong aerosol
cooling)
• Low sensitivity; high total forcing (weak aerosol cooling)
• If we knew aerosol forcing better, we could use the 20th century temperature record to infer climate sensitivity
• Tuning / cheating by climate models?
• …or climate models as suite of possible realities
56
Uncertainty in Future Projections
Both high/low aerosol forcings
consistent with past 100 years
Andreae et al. 2005
Future diverges:
• GHGs accumulate
(long-lived)
• Aerosols don’t
(short-lived)
• Aerosol sources
being controlled
57
Air Quality vs Climate Mitigation
Brasseur and
Roeckner 2005
Constant
GHGs/aerosols
Anthro. sulfate
removed T = 0.8 K
(air quality
improvements
cause quick
warming, as
large as 20th
century)
58
Soot as Climate Mitigation?
• Black carbon
(“soot”) controls
have been
proposed for climate
mitigation
• Potentially cheap,
fast response
• (Somewhat)
quantified warming
effects: +0.3 to 1
W/m2
• Offsetting cooling
effects less well
studied
Kuwaiti oil fires (photo
courtesy of Jay Apt)
Sunlight Absorption
Cloud Burnoff
Snow/Ice Darkening
Cloud Brightening: CCN
Co-emitted Reflecting
Particle Species: SO2,
POA, SOA
Warming Cooling
59
Outline
• Introduction and Overview
• Aerosols: Size, Chemistry, Behavior
• Aerosol Climate Impacts
• Bigger Picture Context
• Aerosols: a major uncertainty in climate science
• Research Challenges
• (Aerosol Properties)
• Mie theory: optics; aerosol-sunlight interactions
• Kohler theory: aerosol-cloud interactions
60
NH/SH
mixing
intra-
hemispheric
mixing
Challenges and Uncertainties
• Need to characterize particle • mass/number concentration
• size distribution: ~10 nm to 10 mm
• chemical composition: >hundreds compounds
• mixing state
• interactions with clouds
• Highly variable in space and time:
century decadal annual daily monthly hourly
Mean CO2
residence
(well
mixed)
Mean
aerosol
residence
(not well
mixed)
61
Aerosol Variability
62
Old Challenge: Aerosol Monitoring
• Global SO4 models: same emissions, chemistry understood
• Global sulfate burdens differ by factor of 2
• Agree with surface (mostly N America and Europe) measured sulfate within ~20%
• Differences in remote areas and above PBL
• Errors in model burden will lead to similar errors in forcing
• Monitoring aerosol burdens essential to keep models ok
Barrie et al., 2001
63
Model-Measurement Integration
Models
Field Campaigns
Remote Sensing
In-situ Monitoring Networks
toxics.usgs.gov
MODIS (oceanmotion.org)
NASA ER-2 (cimss.ssec.wisc.edu)
Spatial coverage
-vs-
Physical-chemical
properties
Temporal
coverage
Exploration,
synergy -vs-
64
Field Campaigns
• Chemical
weather
• Non-
representative
sampling(?)
• Long-term data
sets preferred for
model evaluation
• Better at
exploration
(process insights)
than monitoring
(burden, forcing)
NASA ER-2 (cimss.ssec.wisc.edu)
65
Outline
• Introduction and Overview
• Aerosols: Size, Chemistry, Behavior
• Aerosol Climate Impacts
• Bigger Picture Context
• Aerosols: a major uncertainty in climate science
• Research Challenges
• (Aerosol Properties)
• Mie theory: optics; aerosol-sunlight interactions
• Kohler theory: aerosol-cloud interactions
66
Particle Optics: Introduction
• Energy balance
• Incident = Transmission + Scattering + Absorption
• Define extinction
• Extinction = Scattering + Absorption
• Extinction is the loss of the direct beam
Incident light,
Io(l)
Transmission,
I (l)
Scattering
Absorption
67
Particle Optics: Beer’s Law
• Beer’s Law
• dI(l) = -bext(l) I dx
• Extinction coefficient: bext (m-1)
• Depends on wavelength and air composition
Incident light,
Ix(l)
Transmission,
Ix+dx(l)
dx
68
Particle Optics
• Simple case: monodisperse aerosol (all
particles have same size)
• bext = Qext r2 N
• Qext (l,r) = extinction efficiency (no units)
• r = particle radius (m)
• N = particle number concentration (m-3)
• Qext: extinction efficiency
• relative to particle cross section
• Qext = 1 (every photon incident on particle undergoes
extinction)
69
Mie Theory: Regimes
• Mie theory gives Qext(l,r)
• Three regimes of behavior
• 1) Rayleigh: r << l (e.g. air molecules)
• Qext ~ r4 / l4
• 2) Mie: r ~ l (e.g. aerosols)
• Qext is complicated
• 3) Geometric: r >> l (e.g. cloud droplets)
• Qext ~ 2
70
Particle Optics
• Simple case: monodisperse aerosol (all
particles have same size)
• bext = Qext r2 N
• Geometric optics (e.g. clouds)
• Qext ≈ 2 (not a function of r)
• bext = 2 r2 N (proportional to total surface area!)
• Mie regime (e.g. aerosols)
• Qext ~ r; approx for fine mode (plus range of r)
• bext ~ r3 N (proportional to total mass!)
71
Outline
• Introduction and Overview
• Aerosols: Size, Chemistry, Behavior
• Aerosol Climate Impacts
• Bigger Picture Context
• Aerosols: a major uncertainty in climate science
• Research Challenges
• (Aerosol Properties)
• Mie theory: optics; aerosol-sunlight interactions
• Kohler theory: aerosol-cloud interactions
72
What Makes a CCN: Kohler Theory
• But Peqm above a particle depends on:
1. Temperature (Clausius-Clapeyron)
2. Dissolved solute (Raoult Effect)
• Decreases water vapor pressure
• Facilitates condensation / activation
3. Surface tension (Kelvin Effect)
• CCN / cloud droplets have significant surface free energy
• Destabilizes liquid water phase
• Increases water vapor pressure
• Inhibits condensation / activation
• Activation is a competition between (2) and (3):
o
p
lOHeqm P
RTD
vxP
4exp2
p
l
OHo
eqm
RTD
vx
P
PS
4exp2
Raoult Kelvin
PH2O
Peqm
73
99.6%
99.8%
100.0%
100.2%
100.4%
100.6%
0.1 1 10 100
Wet Diameter [mm]
Re
lati
ve
Hu
mid
ity
What Makes a CCN? (Kohler Theory)
Kelvin effect – Surface Tension - Curvature
Overall Kohler theory: Combination
of Kelvin and solute effects
Solute effect – Raoult’s Law
Slide courtesy of Jeff Pierce
sea-salt particle (50 nm dry diameter)
74
Kohler Theory
0.1 mm dry diameter
Sc ~ 0.13%
0.07 mm dry diameter
Sc ~ 0.2%
0.04 mm dry diameter
Sc ~ 0.45%
Sea-salt (NaCl) particles: Number of CCN depends on:
1) Number of particles
2) Their sizes (size distribution)
3) Their composition (solubility)
75
Aerosol Activation
Diameter
Num
be
r
• “Activation” = formation of cloud droplet
• involves a competition between solute
and surface tension effects
• A particle will activate if it has enough solute to overcome its surface tension
Depends on number
concentration above
“critical diameter”
Important factors: number, size, composition
76
Particles and Climate
“Direct Effect”
•Depends on aerosol mass
concentrations and
composition
•Easier to predict
•Well represented in IPCC
models doing long (century)
climate simulations
•“Traditional” models
“Indirect Effect”
•Depends on number
concentrations, size, and
composition
•Harder to predict: time-
consuming, poorly understood
processes
•Simulations typically 1-5 years
•“Microphysical” models
(developed over last 10 years)
77
Mixing State
• Mixing state: distribution of compounds across different particles
• Example: composition of 100 nm particles (somewhere) is 50% sulfate and 50% organics • «Internally mixed»: every 100 nm particle is half sulfate, half organic
• «Externally mixed»: half of the particles are pure sulfate; half are pure organic
• Reality is always somewhere between these ideal cases
• Internally/externally mixed refer to two particles of (approx) the same size • If a location has fine particles that are organic/sulfate and coarse particles that
are sea spray, this does not mean externally mixed
• Obviously, microphysical processes of condensation and coagulation govern conversion of externally mixed particles into internally mixed
• Observations show • External mixing near sources
• ...becomes more internally mixed over time
• Indirect effect: on global scale, internal mixing probably an ok assumption, more common than external mixing
• Direct effect: fairly insensitive to mixing state except
• For absorption efficiency of black carbon, aerosol mixing state matters (internally mixed absorbs more)
78
Conclusions?
• Aerosols are complicated
• ...but knowing key processes for each mode helps
• Large variety of climate impacts
• ...interactions with clouds tend to be most difficult
• Variety of physical properties
• number, size, composition, mass, mixing state, etc
• ...but usually only a subset is important for a given
problem