Regimes within the First Regimes within the First Aerosol Indirect Effect

Mark A. Miller, Maureen DunnMary Jane Bartholomew Pavlos KolliasMary Jane Bartholomew, Pavlos Kollias

Brookhaven National Laboratory

Thanks: Pete Daum, Mike Jensen, Andy Vogelmann, Christine Chiu,

Dave TurnerDave Turner

Theory of the First Aerosol Theory of the First Aerosol Indirect Effect

• In an adiabatic ascent, an increase in the droplet number concentration at constant liquid water mixing ratio results in a liquid water mixing ratio results in a decrease in the average droplet radius and an increase in the reflection of incoming

li h (T 1974 1991)sunlight (Twomey, 1974, 1991).• Key issues for implementation in real

atmosphere:atmosphere:– Isolated parcel (no mass or heat exchange)– Adequate supersaturation (updraft velocity) q p ( p y)

to activate new Cloud Condensation Nuclei (CCN)

K Q i !Key Question!

• Does the adiabatic cloud droplet nucleation process, at constant p ,liquid water path, actually determine the optical properties of p p pthe cloud?

A Basic ExperimentA Basic Experiment

rewmax

NN

2NFOV-τ

<200 m

#/cm

3

FMWR-LAERI-L

Cloud Chamber

%supersaturation[f(wmax)]

95-GHz-wmax

Cloud Chamber

The One-Third LawThe One-Third Law

• A reduction in the global in the global value of reby 2-μm by 2 μm would offset the doubling Liu and Daum, 2002

-indirect warming effectgof CO2 (Slingo 1990)

g

High LWPHigh LWP

Low LWPLow LWP

LWP and UpwellingLWP and Upwelling

Updraft Limitations?Updraft Limitations?

S d C l iSummary and Conclusions• Observed LWP-dependent regimes within

the First Aerosol Indirect Effect R d th t “O thi d L ” b • Recommend that “One-third Law” be applied selectively in coastal stratus– may require modification to account for may require modification to account for

subadiabatic clouds– Otherwise, an overestimate of AIE

N d d • Need more extensive data set• New technique for number density is

promisingpromising

THE END

Jensen et al. (2007)5-years MODIS Terra—300x300 km boxAlgorithm to remove cloud edge pixels>30,000 data points

LWP RegimesLWP Regimes

Droplet Microphysics in GCMsDroplet Microphysics in GCMs“One-Third Power Law”

Liquid Water Path[PROGNOSTIC]

Shortwave Transmission

1 3L⎡ ⎤

Shortwave TransmissionDepends on re and L

eLr C

N zβ ⎡ ⎤= ⎢ ⎥Δ⎣ ⎦Light Scattering

Properties of Cloud Droplets

Cloud Droplet Number Density[Sometimes PROGNOSTICRelative

Dispersion

Cloud Droplets

[sometimes parameterized]Dispersion

(sigma/mean)[Parameterized]

Results from GResults from G--1 Flights During MASE1 Flights During MASE

Increasing aerosol concentrations observed with increasing Increasing aerosol concentrations observed with increasing g gg gdroplet concentrations and a decrease in the effective droplet concentrations and a decrease in the effective radiusradius..

LWP d D i lLWP and Drizzle

M i S lMarine Stratocumulus

J. Norris1954-1992Synoptic Ship

Observations0 10 20 30 40 50 %

Aerosol Shortwave Indirect Aerosol Shortwave Indirect Effects

• First (or Twomey Effect): “We can with some confidence predict that clouds in the future will contain more droplets per unit volume and, other things being equal, these droplets per unit volume and, other things being equal, these droplets must become smaller” Twomey, S., 1991 Aerosols, clouds, and radiation. Atmos. Env., 25A, 2435-2442. (originally proposed in 1974)

• Second (or Albrecht Effect): “Increases in aerosol concentrations over the oceans may increase the amount of low level cloudiness through a reduction in drizzle—a process low level cloudiness through a reduction in drizzle—a process that regulates the liquid water content and energetics of shallow marine clouds” Albrecht, B.A., 1989: Aerosols, cloud microphysics, and fractional cloudiness. Science, 245, 1227-p y1230.

Th Adi b i Cl d M d lThe Adiabatic Cloud Model

• Many past studies use an adiabatic cloud model to compute the adiabatic cloud droplet number density, which is subsequently linked to effective radius to examine the Aerosol First Indirect Effectexamine the Aerosol First Indirect Effect– Penner et al., 2005; Feingold et al., 200X;

Brengiuer et al., 2005; ect.Br ng u r t a ., 5; ct.– Assume the nucleation properties of the

aerosol particles

Eff f MEffects of Mixing

• Homogeneous versus heterogeneous mixing– AIE impacts apparently different

Thi l d h t ?– Thin clouds—heterogeneous?• Wood: Cancellation of aerosol indirect

effects through cloud thinningeffects through cloud thinning– Cloud base height is critical factor– Zcb<400 m feedbacks cancel Twomey– Consistent with low LWP results

• SGP: New results that support homogeneous mixing in clouds with higher homogeneous mixing in clouds with higher LWP

MODIS Analyses: Effective cloud diameterMODIS Analyses: Effective cloud diameter

RiftRift

SheetSheet

Rift Rift ZonesZones

re vs LWP response varies with air mass

T lf d J W (1996)Telford, J.W.(1996)

Results from GResults from G--1 Flights During MASE1 Flights During MASE

Increasing cloud droplet concentrations observed Increasing cloud droplet concentrations observed Increasing cloud droplet concentrations observed Increasing cloud droplet concentrations observed coincidently with a decrease in the drizzle coincidently with a decrease in the drizzle concentration.concentration.

erHomogeneous

Mi iHeterogeneous mixing

e

erMixing

Extreme case

Secondary activation

Enhanced growth

Underlying mechanism

Faster Mixing Uniform evaporation

Nucleation Coalescence

aN and aL aN=1 aN = aL aN > aL aN < aL

Mixing function

Mixing does not change N

but reduce the sizes

Stronger mixing results

in more droplets

Stronger mixing results in less but

bigger droplets

er

er

sizes droplets,

Response of

Depending on ab and aL

independent of aL

decreases with decreasing aL

increases with decreasing aL

( )1/3e ea Lr rβα α= e ear rβα=

1/3

Le ea

N

r rβααα⎛ ⎞

= ⎜ ⎟⎝ ⎠

1/3

Le ea

N

r rβααα⎛ ⎞

= ⎜ ⎟⎝ ⎠Formula

AIE Effect ? No change More AIE effect

Less AIE effecteffect

Impacts of Cloud Dynamics on Impacts of Cloud Dynamics on p yp yCloud MicrophysicsCloud Microphysics

re

N=288

10.1 Adiabaticity

Adiabaticity, αAdiabaticity, αα α = 1: adiabatic= 1: adiabaticα α < 1: sub< 1: sub--adiabaticadiabatic

Bins of CloudBins of CloudThicknessThickness

10.1 Adiabaticity

Blue: 50Blue: 50--600m600mGreen: 600Green: 600--1000m1000m

τ

Collaborators: B.G. Kim, Steve Schwartz10.1 Adiabaticity

Impacts of Cloud Dynamics on Impacts of Cloud Dynamics on p yp yCloud MicrophysicsCloud MicrophysicsBins of Adiabaticity

Blue: 0.1-0.9 reGreen: 0.9-1.5

Cloud Depth100 1000Cloud Depth100

R2=0.49

R2 0 92τ R2=0.92

Collaborators: B.G. Kim, Steve Schwartz100 Cloud Depth 1000

The Entity-Type The Entity Type Entrainment Mixing Process

J.W. Telford

b t t d

Stage 1

subsaturated

1. Droplets evaporate into subsaturated air2 C li

cloud

2. Cooling3. Descending vortex4. Saturation

saturates

Stage 1 Stage 2

CCN( )( )0 0c ic edV V C C n+ + +

well-mixedsaturates

1. Reduction in size and number of entrained droplets prior to saturation2. Droplets enter turbule after saturation with no change in size3. Drastic changes in number concentration from parcel to parcelg p p

Stage 1 Stage 2

CCN( )( )0 0c ic edV V C C n+ + +

well-mixedsaturates

1 Evaporation to smaller sizes than at same pressure during undiluted ascent1. Evaporation to smaller sizes than at same pressure during undiluted ascent2. Same amount of water has to evaporate from less droplets3. Continued entrainment of larger undiluted drops adds a supply of larger drops4. Parcel contains a wide range of diameters

Stage 3

descent

Stage 1 Stage 2

CCN( )( )0 0c ic edV V C C n+ + +

well-mixedsaturates

1 Rise to accommodate newly descending parcels1. Rise to accommodate newly descending parcels2. Few biggest droplets acquire a disproportionate fraction of new water3. Large droplets formed4. Mocks giant nuclei

Stage 3 Stage 4

descent re-ascent

Observables Checklist• Stage 1

– Reduced reflectivity in pockets near cloud top– Considerable variability in reflectivity near cloud topConsiderable variability in reflectivity near cloud top

• Stage 2 and 3– Large variations in number density from parcel to parcel

Descent– Descent– Lower reflectivity and wider spectral width in the

downdrafts relative to the updrafts• Stage 4• Stage 4

– Possible enhanced reflectivity in some updrafts due to the presence of exceptionally large droplets.

288 3

Continental288 cm-3Number

Density

d159 cm-3

Mean andStandardDeviation

At a Given Wavelength A Cloud Particle At Different Wavelengths

ar ar

At Different Wavelengthsd

to R

ada

d to

Rad

a

Ret

urne

d

Ret

urne

d

Ener

gy R

Ener

gy R

E E

Size of Cloud Particle Radar Wavelength

F d l R l i hiFundamental Relationships

( )( )0

1/

2 1g

D Igττ↑

−=

+Shortwave Albedo

( )2 1 g τ+ −

3 L

2eQ ≈Assume

How much total liquid in the cloud?

Liquid Water Path (LWP)

32 w e

Lr

τρ

=How much total liquid in the cloud?

How much surface area does the liquid have?

Effective Radius--Mean radius for scattering

R l f Hi h LWPResults for High LWP

R l f Hi h LWPResults for High LWP

Remotely Sensing the Cloud Remotely Sensing the Cloud Droplet Distribution

Mode Radius??

Width??

+Hei

ght

Width??

++R d E h

H

Particle Size+

Broadband IrradianceRadar EchoIntensity

Total Liquid Water(Micr v R di m t r)

Number Concentration??

(Solar Radiometer)

(Microwave Radiometer)

Eff i R diEffective RadiusLight Scattering C S tiCross-Section~ r2

Mean

r2

r2-weighted

Particle Size

r weighted arithmetic mean radius