Zhanqing Li Dept of Atmos. & Oceanic Science University of Maryland

Zhanqing Li

Dept of Atmos. & Oceanic Science University of

Maryland

Applications of Aerosol Remtoe Sensing Products

in Climate Studies

0 1 10 100 10001960

1970

1980

1990

2000

Publication per year on "Aerosol AND satellite"

Year

1967: Sekera, aerosol from satellites polarization meas. 1967: ATS III

1972: ATS III dust transport Carlson and Prospero,

1976: dust - Landsat, Fraser

1972: Landsat,

1980s: study of transport aerosol species - effect on climate; stratospheric, aerosol overland: Stowe, McCormicK

1999: MODIS, MISR

2002: GLI

1990s: analysis of POLDER, ATSR, methods for MODIS, MISR, radiation budget

2000s: aerosol climatologies, aerosol-cloud interaction

1996: POLDER

1991: ATSR

1984: Earth Radiation budget satellites

1975: GOES-VISSR

1981: AVHRR afternoon

1965: stratospheric aerosol profiles from Vostok 6 -

Rosenberg and Tereshkova

1979: SAGE

2003: GLAS

The effects of aerosol pollution on clouds

• Since the 1960s, measurements have provided numerous and consistent pieces of evidence that an increase in pollution leads to increases in Cloud Condensation Nuclei (CCN) [a sub-set of atmospheric aerosols].

Indirect Effect Haywood and Boucher Revs. Geophys. (accepted) 2000

1) Increased CCN - reduces reff

2) Drizzle suppression - increases LWC

3) Increased cloud height

4) Increased cloud lifetime

‘First’ indirect effect

‘Second’ indirect effect

eff

LWP

r2

3~

Ice nuclei (IN) are a much smaller sub-set of atmospheric aerosols than CCN.

Their role in precipitation formation in certain clouds is critical.

Finding correlation between the concentrations of ice nuclei and ice crystals in clouds is difficult because of the low concentrations of IN and the numerous mechanisms by which ice crystals can form including ice multiplication mechanisms.

Ice in clouds

How do clouds form?

Clouds form in regions of the atmosphere where water vapor is supersaturated. We focus on liquid water clouds.

Water vapor supersaturation is generated by cooling (primarily through expansion in updraft regions and radiative cooling)

Cloud droplets form from pre-existing particles found in the atmosphere (aerosols). This process is known as activation.

Aerosols that can become droplets are called cloud condensation nuclei (CCN).

CCN that activatesinto a cloud drop

Aerosol particlethat does not activate

Cloud

Köhler curve

0

where)3

(

asgiven valuecritical a has

11 Define

ationSupersatur

6.6Equation See

1

*2

1*

3

*

3

dr

ds

a

br

rr

b

r

a

e

eSS

r

b

r

a

e

e

s

hr

s

hr

Activity SpectrumLet Nc be the number of particles per unit volume that are activated to become cloud droplets.

Data from cloud chamber measurements are often parameterized as

Nc = C (S-1)k

where C and k are parameters that depend on air mass type.

Rogers gives:

Maritime air: 30 < C < 300 cm-3; 0.3 < k < 1

Continental air: 300 < C < 3000 cm-3; 0.2 < k < 2

Thus, for the same saturation ratio, one would expect to find small numbers of CCN per unit volume in maritime air and large numbers per unit volume in continental air.

How can humans affect clouds?

By changing CCN; cloud properties are a strong function of their concentration.

This phenomenon is known as aerosol indirect effect.

The aerosol indirect effect can lead to climatic cooling by:

• Increasing cloud reflectivity (albedo)

• Increasing cloud lifetime & coverage.

Clean Environment

CCN

Lower Albedo

Polluted Environment (few CCN)

(more CCN)

CCN

Higher Albedo

Asian pollution plumes.

Is the indirect effect globally important?

Biomass burning in the Amazon.

Pollution is a global problem. CCN are emitted together with greenhouse gases.

• Aerosol-cloud interactions take place at smaller spatial scales than global climate models can resolve, and must be parameterized.

• Aerosol-cloud interactions are complex; many aspects are unknown or poorly understood.

• Climate models provide important but limited information about clouds and aerosols.

Why is the Indirect Effect Poorly Characterized?

AerosolSize Distribution and Chemical Composition

Cloud Radiative Properties

Cloud Droplet Number and Size?? Well

Well

Defined

Defined

This problem has historically been reduced to finding the relationship between aerosol number concentration and cloud droplet number concentration. Empirical relationships are often used.

Quantification of the Indirect Effect

Goal Goal Couple all aerosol-cloud-radiation interactions within a Couple all aerosol-cloud-radiation interactions within a framework of parameterizations appropriate for global framework of parameterizations appropriate for global models.models.

““Input” variables (from GCM)Input” variables (from GCM)• Cloud liquid water content.Cloud liquid water content.

• Aerosol size distribution and chemistry.Aerosol size distribution and chemistry.

• Wind fields.Wind fields.

• Static stability/turbulence.Static stability/turbulence.

““Output” variables (to GCM)Output” variables (to GCM)• Droplet number, distribution characteristicsDroplet number, distribution characteristics

• Cloud optical propertiesCloud optical properties

• Cloud coverage, subgrid statisticsCloud coverage, subgrid statistics

Aerosol-Cloud Interaction ModulesAerosol-Cloud Interaction Modules

Simplest aerosol-cloud interaction module: correlationsD

ropl

et C

once

ntra

tion

Aerosol sulfate concentration

(Boucher & Lohmann, 1995)

Very large variability.

Why?

• Meteorology

• Cloud microphysics

• Chemical composition

• etc…

Pro: Very simple relationship to implement. Fast computation.

Con: Large predictive uncertainty, without chance of improving.

Fig. 12

Predicted and Observed CCN

0

500

1000

1500

2000

0 500 1000 1500 2000

CN

CC

N

Ramanathan et al., 2000; composite scheme

Cantrell et al., 2001; INDOEX KCO

Cantrell et al., INDOEX Sagar Kanya

0.29 Na1.25 Sk

0.33 Na1.14 Sk

S : Super saturation =0.5%k = 0.76

y = 0.0027x0.643

R2 = 0.87

0.010

0.100

1.000

10 100 1000 10000

CCN0.4 [cm-3]

AO

T5

00

Remote Marine

Remote Continental

Polluted Marine

Polluted Continental

Andreae, ACPD 2008

Maximum at AOD ~ 0.25Giant CCN shift max to greater AOD

Unaccounted “chemical” effects on droplet activation

Slightly soluble compounds (Shulman et al., 1996):They add solute to the drop as it grows; this facilitates their ability to activate.Examples: organics (succinic acid), CaSO4.

Soluble gases (Kulmala et al., 1993):They add solute to the drop as it grows; this facilitates their ability to activate.Examples: HNO3, HCl, NH3.

A(g)

A(aq)

A(g)

A(aq)

A(g)

A(aq)


Surface-active soluble compounds (Facchini et al., 1999):They decrease surface tension of droplets; this facilitates their ability to activate.Examples: organics (succinic acid, humic substances).

The departure from pure water values can be very large!

Surface tension change isdifferent for each CCN.

C(mol l-1)

1e-4 1e-3 1e-2 1e-1

Sur

face

tens

ion

(dyn

e/cm

)

50

55

60

65

70

75

Droplet concentrationrange at activation

Surface tension data from cloud and fog water samples.

Pure water

Charlson et al., Science, 2001


Film-forming compounds (e.g., Feingold & Chuang, 2002):They can slow down droplet growth. Once the film breaks, rapid growth is resumed:

Examples: hydrophobic organics.

Such substances do not necessarily alter droplet thermodynamics; they affect the kinetics of droplet growth.

If present, such substances can strongly affect droplet number.

Film breaks

water molecule

water molecule

Slow Rapid

... advectionevapactivationdrop QQQ

dt

dN

Uncertainties can be decreased by using first principles. Cloud droplet balance:

Activation is the direct aerosol-cloud microphysical link. Two types of

information are necessary for its calculation:

- Aerosol chemistry and size distribution (CCN)

- Representation of subgrid dynamics in cloud-forming regions.

Embedding a numerical activation model is too slow; must parameterize.

Physically-based aerosol-cloud interaction modules

0)()( dwwNwp

dt

d

Probability of updraft w Activated droplets for updraft w

Mechanistic parameterizations: underlying ideas

Approach:• Assume an aerosol size distribution and chemical composition below cloud.

• Aerosols rise into cloud.

• Expansion generates cooling and supersaturation.

• Aerosols activate into droplets.

• Köhler theory links aerosols to CCN properties.

aerosol

activation

drop growth

S

Smax

t

Major challenge:Derive expression for the condensational growth of CCN; include within the supersaturation balance for the parcel, and solve for the maximum.

Solution:• Depends on the approach used in each parameterization. (e.g. Nenes and Seinfeld, JGR, 2003)

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.1 1 10

Updraft Velocity (m/s)

Act

ivat

ion

Fra

ctio

nNumerical Simulation (s.t. effects present)Parameterization (s.t. effects present)Parameterization (s.t. effects absent)

Nenes and Seinfeld, in pressNenes and Seinfeld, JGR, 2003

N & S (2003) evaluation: compare with numerical model

Underprediction: common to many parameterizations

0.001

0.010

0.100

1.000

0.001 0.010 0.100 1.000

Activation ratio (Parcel Model)

Act

ivat

ion

rat

io (

Gh

an P

aram

etri

zati

on

)SM1

SM2

SM3

SM4

SM5

TM1

TM2

prist

ine

pollu

ted

Nenes and Seinfeld, JGR, 2003

Abdul-Razzak et al. parameterization “family”

= 1.0

Satellite observation of aerosol

indirect effect in the Black Sea.

Red: clouds with large drops.

White: clouds with small drops.

Observational evidence of indirect effect

Rosenfeld et al., Science

Power plant

Lead smelter

Port

Oil refineries

Observational evidence of indirect effect

Rosenfeld et al., Science

Satellite observation of aerosol

indirect effect in the Black Sea.

Red: clouds with large drops.

White: clouds with small drops.

DER-AOD relationship

Yuan et al. (2008, JGR)

AIE efficiency distribution

Yuan et al. (2008)

AIE efficiency determining factor

Global Analysis Region Latitude

rangeLongitude range

Dominant Aerosol/Cloud

Types

Period AIE efficiency

Sample size

North Atlantic 10-20N 20-40 W Dust, Stratocumulus

June-August, 2002 Negative 99,978

South Atlantic 5-20S 5E-20W Smoke, Stratocumulus

June-August,2002 Negative 100,377

Southern Pacific

5-25S 75-105W Sea salt, sulfate and pollution,

Stratocumulus

August-October,2002

Negative 74,216

Indian Ocean 12-20N 60-70E Dust with pollution, Trade cumulus

June-August, 2002 Negative 94,023

India 13-24N 70-85E Mixture of sulfate, dust, sea salt and smoke, cumulus

June-August,2002 Neutral 53,888

Amazonia 8S-12N 44-76W Mainly smoke August-October,2002

Negative 672,421

Southeastern China

23-43N 100-120E Mixture, cumulus June-August,2002 Positive 179,533

Student-t test indicates except India the difference among different loading of aerosols are statistically significant at least at the 95% level

h = 2.72 N

0

1000

2000

3000

4000

5000

6000

0 500 1000 1500 2000

War

m R

ain

Dep

th (

m)

Average Droplet Concentration (cm ) -3

More cloud drops deeper cloud for onset of rain

Can the slowing of auto-conversion result in increasingprecipitation?

Conceptual model:

Graphics by Robert Simmon, NASAHAIL

Conceptual model:

Graphics by Robert Simmon, NASAHAIL

Annual average lightning density [flashes km-2]Lightning prevail mostly over land, whereas rainfall is similar over land and ocean, indicates fundamental differences between continental and maritime rainfall.

Why is Continental - Maritime classification so fundamental?

Why is Continental - Maritime classification so fundamental?

TRMM annual average rainfall amount [mm / day]

There is little relation between lightning and rainfall amount

Global effects of pollution on precipitationGCM-- estimates 0 to - 4.5% change in global mean precipitation over the last

100 years due to the direct and indirect aerosol effects.

The differences among models over land range from -1.5% to -8.5%.

Global N. Hemisphere S. Hemisphere

Ocean Land

A lot have been done concerning aerosol’s impact on rainfall

Rainfall Suppressed by Aerosols

Rosenfeld, 1999; Rosenfeld, 2000; Andreae, 2004;

etc

Khain, 2004 ， 2005; Tao 2007, Fan, 2007; Van den Heever, 2007

etc

Rainfall Enhanced by Aerosols

Koren, 2005;Lin, 2006;Bell, 2007;

etc

Observational studies

Modeling studies

But little has been done for rain frequency!

While rain amount and frequency change in harmony in general, the

impact of aerosol on initiation of rain is likely to be more significant than rain amount, as the latter is dictated more by dynamics and abundance of

available water

Datasets Used

• Daily ARM SGP data 2003-2008 (~20000 data samples)

• Most complete and highest quality measurements of aerosol, cloud, atmospheric state

• Key variables used:– Aerosol CN number concentration on the ground

– Tipping bucket rain gauge

– LWP from microwave radiometer

– Cloud bottom and top heights from cloud radar & lidar

– NOAA/NCAR Reanalysis

– MODIS cloud particle size

Rainfall Frequency for clouds with different liquid water path at SGP

(All-Season Data)

0.00%

10.00%

20.00%

30.00%

40.00%

50.00%

60.00%

0-1000 1000-2000

2000-3000

3000-4000

4000-5000

5000-6000

CN Number Concentration (1/cm^3)

Ra

infa

ll F

req

ue

nc

y f

or

Clo

ud

s

wit

h H

igh

an

d M

od

era

te L

WP

0.00%

0.40%

0.80%

1.20%

1.60%

2.00%

Ra

infa

ll F

req

ue

nc

y f

or

Clo

ud

s

wit

h L

ow

LW

P

LWP:>0.8mm

LWP:0.4-0.8mm

LWP:0.0-0.4mm R2 = 0.7803

R2 = 0.9088

R2 = 0.037

5

10

15

20

25

0-1000 1000-2000

2000-3000

3000-4000

4000-5000

5000-6000

CN Number Concentration (per cm^3)

Su

rfa

ce

te

mp

era

ture

(D

eg

ree

Ce

lsiu

s)

LWP:>0.8mm

LWP:0.4-0.8mm

LWP:0.0-0.4mm

970

972

974

976

978

980

982

0-1000 1000-2000

2000-3000

3000-4000

4000-5000

5000-6000

CN Number Concentration (/cm^3)

Su

rfa

ce

Pre

ss

ure

(h

p)

LWP:>0.8mm

LWP:0.4-0.8

LWP:0.0-0.4

0

10

20

30

40

50

60

0-1000 1000-2000

2000-3000

3000-4000

4000-5000

5000-6000


Wat

er V

apo

r (c

m)

>0.8

0.4-0.8

0.0-0.4

3

3.5

4

4.5

5

5.5

6

6.5

0-1000 1000-2000

2000-3000

3000-4000

4000-5000

5000-6000

CN Number Concentration (per cm^3)

Su

rfa

ce

Win

d S

pe

ed

(m

/s)

LWP:>0.8mm

LWP:0.4-0.8mm

LWP:0.0-0.4mm

T P

WV Wind

Cloud Thickness for clouds with different cloud base heights

0

500

1000

1500

2000

2500

3000

3500

0-1000 1000-2000

2000-3000

3000-4000

4000-5000

5000-6000

CN number concentration (1/cm^3)

Clo

ud

Th

ickn

ess

(m

)

CBH:<1km

CBH:1km-2km

CBH:2km-4km R2 = 0.1055

R2 = 0.9718

R2 = 0.9159 y = 638.06x + 1258.6

R2 = 0.9169

y = 97.05x + 2321.2

R2 = 0.5543

y = 56.399x + 3819.7

R2 = 0.1658

0

1000

2000

3000

4000

5000

6000

0-1000 1000-2000

2000-3000

3000-4000

4000-5000

5000-6000


Clo

ud

To

p H

eig

ht

(m)

CBH:<1km

CBH:1km-2km

CBH:2km-4km

Linear (CBH:<1km)

Linear (CBH:1km-2km)


Cloud Base Heights

y = -10.06x + 2969.9

R2 = 0.2024

y = 12.978x + 1404.7

R2 = 0.586

y = 7.9095x + 330.05

R2 = 0.5579

0

500

1000

1500

2000

2500

3000

3500

0-1000 1000-2000

2000-3000

3000-4000

4000-5000

5000-6000


Clo

ud

Ba

se H

eig

ht

(m)

CBH:<1km

CBH:1km-2km

CBH:2km-4km



Linear (CBH:<1km)

Cloud Top Heights(for clouds of cloud base <1km)

All Seasons Summer Only

y = 237x + 4160.4

R2 = 0.8169

y = 32.296x + 964.41

R2 = 0.7245

0

1000

2000

3000

4000

5000

6000

7000

0-1000 1000-2000

2000-3000

3000-4000

4000-5000

5000-6000


Clo

ud

To

p H

eig

ht

(m)

CTH:>2km

CTH:<2km

Linear (CTH:>2km)

Linear (CTH:<2km)

y = 907.49x + 3297.4

R2 = 0.82

y = 22.14x + 1159.4

R2 = 0.2692

0

1000

2000

3000

4000

5000

6000

7000

8000

9000

10000

0-1000 1000-2000

2000-3000

3000-4000

4000-5000

5000-6000


Clo

ud

To

p H

eig

ht

(m)

CTH:>2km

CTH:<2km

Linear (CTH:>2km)

Linear (CTH:<2km)

For clouds with CBH<1km, clouds are classified into two categories with cloud top heights greater (blue) and less than (red) 2km.

Cloud Top Heights

y = 327.17x + 1168.6

R2 = 0.9727

y = 222.86x + 1848.1

R2 = 0.9344

y = 12.256x + 4290.7

R2 = 0.0428

0

1000

2000

3000

4000

5000

0-1000 1000-2000

2000-3000

3000-4000

4000-5000

5000-6000


Clo

ud

To

p H

eig

ht

(m)

CBH:<1km

CBH:1km-2km

CBH:2km-4km

Linear (CBH:<1km)



y = 638.06x + 1258.6

R2 = 0.9169

y = 97.05x + 2321.2

R2 = 0.5543

y = 56.399x + 3819.7

R2 = 0.1658

0

1000

2000

3000

4000

5000

6000

0-1000 1000-2000

2000-3000

3000-4000

4000-5000

5000-6000


Clo

ud

To

p H

eig

ht

(m)

CBH:<1km

CBH:1km-2km

CBH:2km-4km

Linear (CBH:<1km)



All Seasons Summer Only

Competition of two opposite effects

Responses ofRainfall frequency to increasing CN

Meteorological effects

Invigoration EffectsIncrease rainfall

Suppress rainfall

Not always increase

Depend criticallyon cloudbase !!

Microphysical Effect

Depend criticallyon available water

The WMO/IUGGINTERNATIONAL AEROSOL PRECIPITATION SCIENCE ASSESSMENT GROUP

(IAPSAG)

Aerosol Pollution Impact on Precipitation:A Scientific Review

Zev Levin, ChairmanWilliam Cotton, Vice Chairman

Approved by the WMO - May. 2007

Recommendations• Implement a series of international projects targeted toward

unraveling the complex interactions among aerosols, clouds, and precipitation.

• WMO/IUGG should take the lead in such projects together with other UN and International Organizations.

• Some of these could be sponsored and financially supported by the countries involved. For example:

– Study the effects of an evolving industrial economy such as China on precipitation.

– Study the effects of biomass burning and dust in some of the African regions.

1) Better characterization of aerosols

– Emission inventories• Size, number concentrations

– Chemical processes, physical properties and instrumentation

• Accurate knowledge of the chemical processes leading from gas pollution to CCN

• The ability of different types of particles (e.g. mineral dust, biomass smoke, biogenic, carbonaceous) to act as CCN, GCCN, and IN as a function of aerosol size, origin, and air mass history.

The WMO/IUGG can play a key coordination role in encouraging that the following recommendations are implemented.

• Develop new and innovative instruments and measurements to determine CCN, GCCN and IN concentrations as a function of particle size, composition and supersaturation.

• Emphasis should be placed on understanding the different modes of ice nucleation.

• Global coordination of observational networks is needed for more complete coverage of global aerosols (ground-based

remote sensing methods e.g. AERONET) .

• More accurate assessment is needed from satellites of the aerosol distribution, concentration and properties.

2) The effects on clouds and precipitation:

– Design experiments to better understand the role of ice in precipitation development;

– Multi-year measurements from space of precipitation patterns along with retrievals of cloud nucleating aerosols are needed to assess both regional and global impacts of aerosol pollution on precipitation.

– Improved satellite measurements of Liquid Water path (LWP) and Ice Water Path (IWP), which define the potential for precipitation – with pollution modulating how much will reach the ground.

• Models should be used to provide a quantitative answer as to the relative effects of aerosols versus environmental parameters (temperature, Relative humidity, wind shear, land-surface properties, etc.) on precipitation.

• New methods are needed to estimate precipitation amounts with high enough accuracy to be able to resolve changes due to pollution.

– The high variability in precipitation amounts from GCMs stresses the need to improve representation of aerosol and cloud processes to be able to answer with some confidence the question on the effects of pollution on precipitation.

– Detailed knowledge of ice formation in clouds is still lacking, requiring more laboratory, modeling, and field studies.

Prof. John Seinfeld, Caltech

Prof. Athanasios Nenes, George Tech

Prof. W. Cotton, CSU

Prof. D. Rosenfeld, Hebrew University

Prof. V. Ramanathan, UCSD

Prof. Z. Li, UMD

Contributors

Documents

Zhanqing Li Dept of Atmos. & Oceanic Science University of Maryland