Climatology of Aerosol Optical Properties over the High Arctic

Auromeet Saha1, N.T. O’Neill1, R.S. Stone2

K. Strong3, C. Viatte3, A. Herber4

1 CARTEL, Universite de Sherbrooke, Sherbrooke, Quebec, Canada2 CIRES/NOAA ESRL, Boulder, Colorado, USA

3 University of Toronto, Toronto, Ontario, Canada4 Alfred Wegener Institute for Polar & Marine Research, Germany

• Aerosols affect the Earth’s radiative balance through direct and indirect interactions with shortwave (SW) and longwave (LW) radiation at the top of the atmosphere, at the surface, as well as within the atmosphere by providing a radiative heating (or cooling) in the aerosol layers.

• Aerosols scatter (sulfates) and absorb (Soot/Black Carbon and Dust) the incoming solar radiation (Aerosol Direct Effect).

• Scattering aerosols reflect the incoming solar radiation resulting in atmospheric cooling. However, warming may result where the surface albedo is very high (Pueschel and Kinne, 1995) or if the sulfate is internally mixed with absorbing aerosols (Jacobson, 2001).

• Coarse mode aerosols such as mineral dust act like greenhouse gases by absorbing and emitting longwave (LW) radiation.

• Aerosols can influence the microphysical properties of clouds (Indirect aerosol effects) and thereby impacts the energy balance, precipitation, and the hydrological cycle. Depending on the composition and size, aerosols can serve as cloud condensation nuclei (CCN) or ice-forming nuclei (IN).

• Indirect effects of aerosols on cloud properties typically cause surface cooling (Quinn et al., 2008) but can also warm the surface through interactions with terrestrial (LW) radiation (Lubin and Vogelmann, 2006). The warming is expected to dominate during Arctic winter (Lubin and Vogelmann, 2007).

Introduction: Aerosols & Arctic Haze

• Long-range transport of pollution from mid-latitudes is a major source of aerosols to the Arctic, with a winter-spring maximum known as Arctic haze (Quinn et al., 2009). High aerosol concentrations in the Arctic during winter-spring can be due to efficient transport from mid-latitudes, reduced vertical mixing, and lack of precipitation (Shaw, 1995; Quinn et al., 2007; Garrett et al., 2010).

• Unlike long-lived CO2, aerosols have an average lifetime in the atmosphere from days to weeks, but exhibit a somewhat longer lifetime in the Arctic.

• Absorption of solar radiation by aerosols in the atmosphere is enhanced by highly reflective snow- and ice-covered surfaces in the Arctic. Deposition of light-absorbing aerosols (such as Soot) on snow or ice can decrease the surface albedo.

• The shortwave (SW) component of the energy balance can be affected by light absorbing aerosols through the atmosphere and in the ice/snow surface.

• With respect to the longwave (LW) component, the aerosol effect in the Arctic differs from other regions because of the cold atmospheric temperatures, frequent occurrences of temperature inversion, and low water vapor content.

• The longwave (LW) radiation plays a particularly important role in the energy balance in the Arctic because of prolonged polar nights with little or no sunlight.

• Hygroscopic growth of particles leads to absorption of terrestrial radiation, inducing a direct warming effect that can be particularly efficient during polar night (Ritter et al., 2005).

Introduction: Aerosols & Arctic Haze

High Arctic Sites

MODIS-Terra (250 m) July 8, 2004

OPAL

PEARL

Eureka

PEARL - Polar Environmental Atmospheric Research Lab. It is operated by the Canadian Network for the Detection of Atmospheric Change (CANDAC).

PEARL is located at Eureka, Nunavut (80N, 86W) on Ellesmere Island in Canada's high Arctic, 450 km north of Grise Fiord, the most northerly permanent settlement. PEARL is 1,100 km from the North Pole.

ØPAL - Zero altitude Polar Atmosphere Laboratory. It is located about 15 km southeast of the PEARL ridge lab which is at an elevation of 610 m.

PEARL

Instruments & Data used in this Study (Eureka)

Cimel Sun Photometer

Brewer SpectrometerBruker Fourier Transform Spectrometer

AHSRL

Cimel Sun Photometer (AEROCAN / AERONET)

Aerosol optical properties: Aerosol optical depth (AOD), and Effective radius of the fine & coarse mode.

The fine mode (f & reff,f) can typically be associated with forest or agriculture fire smoke, and pollution aerosols, while the coarse mode (c & reff,c) is an indicator of super-micron sized particulates such as dust, marine aerosols, and ice or water phase cloud particulates.

AEROCAN (Canadian Sunphotometer Network)

Resolute Bay

Saturna Island

Bratt’s Lake

Waskesiu

ChurchillKuujjuarapik

HalifaxSherbrookeEgbert

Pickle Lake

Fort McMurray

Kelowna

Chapais

Eureka (PEARL, OPAL)

Toronto

Yellowknife

Iqaluit

Sable Island

Lethbridge

AEROCAN is a full fledged sub-network of AERONET and benefits from all services that AERONET offers.

AEROCAN has a few unique features that go beyond AERONET protocols:

- All instruments use the high frequency, 3-minute mode (O’Neill mode).

- All data transmission is done using FTP.

- Complementary AOD processing chain in preparation for real timeingestion into an aerosol assimilation model (currently on hold).

- System wide display tools for real-time QA & troubleshooting.

- Separate prototype MySQL database for data QA (and science).

- Satellite calibration facility being developed (on hold with EC freezes).

AERONET vs AEROCAN

AEROCAN* (Canadian Starphotometer Sites)

Egbert

Eureka

Sherbrooke

Arctic High Spectral Resolution Lidar (AHSRL)

AHSRL measures the vertical profiles of backscatter coefficient () and depolarization ratio ()

The δ profiles permit one to distinguish between sub-micron smoke or pollution aerosols (low δ values) and irregularly shaped super-micron particles such as ice crystals and coarse mode aerosols such as dust (high δ values).

Bruker Fourier Transform Spectrometer(CO Column concentration)

Brewer Spectrometer (SO2)

Instruments & Data used in this Study (Barrow & Alert)

NOAA Sun Photometer SP02 used at Barrow & Alert (aerosol optical properties)

wavelength photometer deployedtraps heat and stabilizing temp

OPAL – CIMEL 327 PEARL – CIMEL 401

Sunphotometers installation in Eureka

runwayOPAL

PEARL Ridge Lab

This dual placement was designed to study the layer between the two sites as well as provide an element of redundancy for the AOD measurements.

OPAL

PEARL

~12 km~0.61 km

SAFFIRE

OPALPEARL

0.00

0.04

0.08

0.12

0.16

AO

D (5

00 n

m)

OPALPEARL

Eureka Aerosol Optical Depth Climatology

2007-2011

0

50

100

150

MAR APR MAY JUN JUL AUG SEP

Nob

s (D

ays)

OPALPEARL

0.00

0.05

0.10

0.15

0.20

0.25

AO

D (5

00 n

m)

Egbert (1998-2011)

CARTEL(1995-2010)

0

100

200

300

JAN

FEB

MAR AP

R

MAY JU

N

JUL

AUG

SEP

OC

T

NO

V

DEC

N obs

(Day

s) EgbertCARTEL

Mid-Latitude AOD Climatology

0.00

0.50

1.00

1.50

2.00

Prec

ipita

ble

Wat

er (c

m)

OPALPEARL

0

50

100

150

MAR APR MAY JUN JUL AUG SEP

Nob

s (D

ays)

OPALPEARL

Eureka Precipitable Water Climatology

0

1

2

3

Prec

ipita

ble

Wat

er (c

m)

Egbert (1998-2011)

CARTEL (1995-2010)

0

100

200

300

JAN MAR MAY JUL SEP NOV

Nob

s (D

ays)

EgbertCARTEL

Mid-latitude Precipitable Water Climatology

0.00

0.05

0.10

0.15

0.20

N L N L N L N L N L

AO

D (5

00 n

m)

OPAL

PEARL

0

10

20

30

MA

R

AP

R

MA

Y

JUN

JUL

AU

G

SE

P

MA

R

AP

R

MA

Y

JUN

JUL

AU

G

SE

P

MA

R

AP

R

MA

Y

JUN

JUL

AU

G

SE

P

MA

R

AP

R

MA

Y

JUN

JUL

AU

G

SE

P

MA

R

AP

R

MA

Y

JUN

JUL

AU

G

SE

P

Nob

s (D

ays) OPAL

PEARL

2007 2008 2009 2010 2011

Seasonal and inter-annual variation of AOD over EurekaBiomass burning from Forest & Agricultural Fires in Russia & Kazakstan

Kasatochi volcanic aerosols

Sarychev volcanic aerosols

Biomass burning aerosols during Spring 2008 (ARCTAS)*

*Saha et al., GRL, 2010

0.0

0.1

0.2

0.3

0.4

9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

PEARL (Level 1.0), poly. order = 2, = 0.5 m

Total Fine Coarse

m-1 str-1)

Day of Month (April 2008)

Alti

tude

(km

)A

ltitu

de (k

m)

AHSRL aerosol backscatter cross section ()

AHSRL particulate circular depolarization ratio ()

Fine mode aerosols were associated with biomass burning (agricultural and forest fires) from E/SW Russia & northenKazakhstan.

Coarse‐mode events were likely due to ice particles whose nucleation may have been associated with the presence of smoke, or possibly dust.

f correlates well with variations in the presence of small , while cvalues correlate with in the presence of large .

Barrow

T hule

PEA RL

Yellowknife

Ny-A lesund

Hornsund

SERb

(c)

Ka(e)

(b)(a)

(e)2008 04 12 1500 UTC 2008 04 28 1800 UTC

SERaKb

SER2

2008 04 24 0900 UTC

SER2

*Saha et al., GRL, 2010

AOD

GEM-AQ AOD simulation for April 2008

*Saha et al., GRL, 2010

0

0.1

0.2

0.3

103.4 103.5 103.6 103.7 103.8 103.9 104 104.1

f ine mode OD

coarse mode OD

total OD

0

0.1

0.2

0.3

104.4 104.5 104.6 104.7 104.8 104.9 105 105.1

f ine mode OD

coarse mode OD

total OD

April 12 April 13

Fine Mode (smoke) event of April 12-13 observed at Eureka

fine modeSmoke layers

*Saha et al., GRL, 2010

Aqua, April 12, 1405 UT ThuleEureka

Terra - fusion of April 12, 1155 and 1200 UT images.

MODIS true color images (fusion of Terra and Aqua) showing the Pan-Arctic view of the April 12 plume crossing Eureka (part of the SERa plume movement).

MODIS Imagery for April 12 Smoke Plume crossing Eureka

*Saha et al., GRL, 2010

Ny-Alesund

MODIS Aqua Visible, 0930, April 24, 2008, 500 m. resolution

Smoke surrounding Ny-Alesund (Kb event of April 24). While blinking between the Aqua visible image (R,G, B = bands 1, 4, 3) and the Aqua SWIR image (R,G, B = bands 7, 2, 1), the fine-mode smoke largely disappears in the SWIR image over water.

Ny-Alesund

MODIS Imagery for April 24 Fine Mode Event at Ny-Alesund

*Saha et al., GRL, 2010

Smoke plume event at Barrow. The Aqua true-color image has been CMYK enhanced (linear stretch applied to the C, M, Y channels). The higher reflectivity of clouds versus snow indicates that the plume lies above a low-lying cloud. This was confirmed in comparisonswith the HSRL data of Ferrare et al. (2009). The red circle indicates the position of Barrow.

MODIS Aqua Visible, 2220, April 19, 2008, 250 m. resolution

MODIS Imagery for April 19 Fine Mode Event at Barrow (Alaska)

*Saha et al., GRL, 2010

108 109 110 111

0.0

0.1

0.2

0.3

0.4

0.5

0.6

Aeronet Level 1.0

D f Y (2008)

17-19 Apr-2008PEARL

AOT

()

Total Fine Coarse

HYSPLIT Back Trajectories

OMI Aerosol Index Gobi Desert

CALIOP profiles of Total Attenuated Backscatterand depolarization ratio

Coarse Mode Event observed at Eureka during April 17-19

For this high event, c varies in a low frequency (diurnal) fashion, unlike typical cloud effects whose high frequency cexcursions form the basis of most cloud screening algorithms. The smoke plume positions inferred from MODIS imagery and synoptic- dynamics combined with an analysis of the lidar / sunphotometry data and AOD simulations by Pierce et al. (2009) suggest that these features were a complex mixture of smoke and ice crystals and that the smoke particles (or possibly dust) might have acted as Ice Nuclei.

Eureka

Precipitating Ice Crystals

100.5 100.6 100.7 100.8 100.9 101.0 101.1 101.2 101.3

0.0

0.1

0.2

0.3

0.4

0.5

Barrow (Screened), poly. order = 2, = 0.5 m

Total Fine Coarse

The AOD time series plot over Barrow on April 9 shows significant low-frequency c variation with enhanced backscatter and high cross-polarization profiles. Using airborne lidar measurements acquired near Barrow on April 9, Ferrare et al. (2009) concluded that ice particles were making a significant contribution to the AOD.

Coarse Mode Event observed at Barrow on April 9

*Saha et al., GRL, 2010

0

0.2

0.4

182.0 182.5 183.0 183.5 184.0

Opt

ical

Dep

th

fine mode ODcoarse mode OD

total ODAHSRL SOD

0.0

0.2

0.4

182.0 182.5 183.0 183.5 184.0Doy.UT/24

Opt

ical

Dep

th

0

5

10

SO2 [

DU

]fine mode ODBrewer #007Brewer #021Brewer #069

m-1 str-1)

July 1 July 2

(d)

(a)

(b)

(c)

Sarychev volcanic aerosols over the Arctic during Summer 2009*

0.00

0.05

0.10

0.15

f (5

00 n

m)

Pre-eruption Post-eruption

0.0

0.1

0.2

0.3

0.4

r eff

,f [

m]

*O’Neill et al., JGR 2012

Pre- and post- SAT f, and reff,f for four AEROCAN / AERONET high-Arctic stations

0PAL PEARL Thule Hornsund

Eureka

July 01, 2009 OMI: July 02, 2009

CA

LIO

P –

stra

tosp

heric

O

DO

MI –

colu

mn

conc

entra

tion

Envelope of CALIOP orbit lines

0.0

0.5

1.0

1.5

2.0SO2 [DU]

Illustration of Sarychev Plume (OMI SO2 vs CALIOP SOD)

*O’Neill et al., JGR 2012

1000

1500

2000

2500

CO

[mol

. cm

-2] x

1015

0

10

20

30

Nob

s (D

ays)

Variations of CO Total Column over Eureka

0

MA

R

AP

R

MA

Y

JUN

JUL

AU

G

SE

P

MA

R

AP

R

MA

Y

JUN

JUL

AU

G

SE

P

MA

R

AP

R

MA

Y

JUN

JUL

AU

G

SE

P

MA

R

AP

R

MA

Y

JUN

JUL

AU

G

SE

P

MA

R

AP

R

MA

Y

JUN

JUL

AU

G

SE

P

2007 2008 2009 2010 2011

CO being a tracer of biomass, one would expect a degree of correlation with the fine mode aerosols that are also a primary product of biomass burning.

y = 0.0027e0.0014x

R2 = 0.5781

0

0.1

0.2

0.3

1000 1500 2000 2500 3000

CO Column [mol. cm-2] x 1015

f (5

00 n

m)

2007

y = 0.0003e0.0026x

R2 = 0.7209

0

0.1

0.2

0.3

1000 1500 2000 2500 3000

CO Column [mol. cm-2] x 1015

f (5

00 n

m)

2008

y = 0.0099e0.001x

R2 = 0.5884

0

0.1

0.2

0.3

1000 1500 2000 2500 3000CO Column [mol. cm-2] x 1015

f (5

00 n

m)

2009

Sarychev

Fine Mode AOD vs CO

Shown in the figures are the variation of fine-mode AOD as a function of CO concentration for the three years of 2007, 2008 and 2009.

One can observe a moderate (2007), strong (2008) and insignificant (2009) correlation respectively for these three years. However if we separate out the Sarychev fine-mode AODs from the 2009 regression, we find a R2 value that is comparable to the other two years.

0.00

0.05

0.10

0.15

AOD

(500

nm

)

totalfine modecoarse mode

0.0

0.1

0.2

0.3

r eff,

f (m

)

PEARL

0

1

2

3

MAR MAY JUL SEP MAR MAY JUL SEP MAR MAY JUL SEP MAR MAY JUL SEP MAR MAY JUL SEP

r eff,

c (

m)

PEARL

Retrievals of fine & coarse mode AOD and Effective radius

2007 2008 2009 2010 2011

The fine-mode AOD shows the spring to summer decrease (confirms that the total AOD is fine-mode dominated).

The coarse mode AOD shows apparent springtime activity that could possibly be Asian Dust.

The fine-mode effective radius shows a relatively consistent variation except in 2009 where we know there was a strong Sarychev influence (O’Neill et al., 2012).

The coarse mode effective radius shows significant variations that could be real (possibly associated with marine aerosols) but which might, just as well, be associated with low-AOD retrieval artifacts.

reff,f ~ 0.2 m

00 04 08 12 16 20 24

0.00

0.05

0.10

0.15

0.20

0.25

Time (GMT)

AOT

()

Total Fine Coarse

Taklimakan Desert April 19, 2009

Illustration of Coarse Mode Event (Asian Dust)

Precipitating Ice crystals

Dust

April 27, 2009

0.0

0.2

0.4

0.6

0.8

18 19 20

AHSRL, = 532 nm, Fine: 0.05 < < 0.4, Coarse: > 0.4

Fine Coarse

Fine Coarse

PEARL (Level 1.0), poly. order = 2, = 0.5 m

April 2008

Coarse Mode Event (Nucleation of Smoke & Dust) during ARCTAS*

*Saha et al., GRL, 2010

0.0

0.1

0.2

0.3

0.4

16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

PEARL (Level 1.0), poly. order = 2, = 0.5 m

April 2009

Total Fine Coarse

Suspicious coarse mode peaks possibly due to frost in the Cimeloptics.

Coarse mode event at PEARL (Artifacts)

AHSRL does not show any coarse mode event

Frost Issues

Cimel SunphotometerSept. 9, 2011

Cimel SunphotometerSept. 22, 2009

Star PhotometerWinter 2012

0

0.2

0.4

0.6

0.8

1

14 15 16 17 18 19

fine mode OD coarse mode OD total OD

-10

10

30

50

14 15 16 17 18 19

day in April, 2008

100*[L(sky)-L(ext.)]/L(ext) at 1020, 870, 675 and 440 nm

Almucantar % differences,

Almucantar radiance differences at 6° azimuth shouldagree within 5%.

y = 0.9659x + 0.1163R2 = 0.8315

y = 1.7528x + 0.0893R2 = 0.7687

y = 1.8462x + 0.348R2 = 0.2402

0

0.5

1

1.5

2

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 11 / m

AO

T(50

0 nm

)

April 15April 16April 14Linear (April 16)Linear (April 15)Linear (April 14)

Computed AOT varies suspiciously as 1 / m(m – airmass)

April 2008

Coarse mode event at Thule (Artifacts)

0

10

20

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Win

d Sp

eed

(km

hr-1

)

Climate Normals (1971-2000)

Coarse Mode Retrievals vs Wind speed (Eureka)

0

1

2

3

MAR MAY JUL SEP MAR MAY JUL SEP MAR MAY JUL SEP MAR MAY JUL SEP MAR MAY JUL SEP

r eff,

c (

m)

PEARL

Eureka (OPAL) Wind Speed Climatology2007 2008 2009 2010 2011

Sunphotometry at High Arctic Sites

Barrow-NOAA

Barrow-AERONET PEARL

OPALThule

Resolute Bay

Alert

HornsundNy-Alesund

Kangerlussuaq

Pan-Arctic sites for the AOD Climatology

0.00

0.05

0.10

0.15

0.20

AO

D (5

00 n

m)

Barrow NOAA (2001-2011)Barrow AERONET (1999-2011)Resolute Bay (2004-2011)OPAL (2007-2011)PEARL (2007-2011)Alert (2004-2011)Thule (2007-2011)Kangerlussuaq (2008-2011)Hornsund (2006-2011)Ny-Alesund (1991-2011)

gy

0

100

200

300

MAR APR MAY JUN JUL AUG SEP OCT

Nob

s (D

ays)

Barrow NOAABarrow AERONETResolute BayOPALPEARLAlertThuleKangerlussuaqHornsundNy-Alesund

Pan-Arctic AOD Climatology

AOD monthly climatology for an enhanced number of high Arctic stations on a pan-Arctic scale. It simultaneously shows the spring to summer decrease noted for Eureka as well as a west to east AOD decrease.

Summary & Conclusions

A multi-year AOD and effective radius climatology for the high Arctic showed a number of consistent features:

Spring to summer decrease of fine-mode AOD (probably attributable to biomass burning and/or anthropogenic pollution).

Significant correlation of fine mode AOD with CO concentration (which indicates a predominance of biomass burning aerosols throughout the entire year).

West to east decrease in AOD on a pan-Arctic scale.

Acknowledgements

0.00

0.05

0.10

0.15

0.20

07-May 21-May 04-Jun 18-Jun 02-Jul 16-Jul 30-Jul 13-Aug 27-Aug

AO

D (5

00 n

m)

OPAL

PEARL

2011

Differential Sunphotometry at Eureka