Upload
jetta
View
31
Download
0
Tags:
Embed Size (px)
DESCRIPTION
Aerosol Impacts on Arctic Climate During the 20th Century: A GISS Climate Model Study Dorothy Koch Columbia University/ Goddard Institute for Space Studies Surabi Menon, Stephen Warren, Igor Alienov, Reto Ruedy, Tami Bond SPAC Workshop Oslo, Norway November 5, 2007. - PowerPoint PPT Presentation
Citation preview
Aerosol Impacts on Arctic Climate Aerosol Impacts on Arctic Climate During the 20th Century: A GISS During the 20th Century: A GISS
Climate Model StudyClimate Model Study
Dorothy Koch Columbia University/ Goddard Institute for Space Studies
Surabi Menon, Stephen Warren, Igor Alienov, Reto Ruedy, Tami Bond
SPAC Workshop Oslo, Norway November 5, 2007
Aerosol effects in the Arctic1. Direct Effect, warming or cooling?: Sulfates cool, BC
may cool the surface but warm column
2. Indirect Effect, warming or cooling?: Increased CDNC may cause cooling (SW) or warming (LW)
3. BC-albedo Effect, warming: black carbon darkens snow, enhances melting, exposes darker surfaces causing more melting
Which Effects warm/cool?
What are their seasonalities?
How do changing LL GHG influence the aerosol effects?
ModelGISS ModelE GCM (Schmidt et al., 2006) 4x5x20 levelsAtmosphere-ocean equilibrium climate simulations using q-flux ocean for 1995
and 1890 simulationsAerosol species (mass, externally mixed) interact with climate: Black and organic carbon (BC, OC) , sulfate (SO4), sea saltAerosols coupled to GCM: transport, cloud processing, boundary layer, dry
deposition, radiationRemoval: Sulfate and sea-salt are soluble
Non-biomass burning BC and OM become soluble with timeBiomass burning OM and BC and dust have fixed solubility
Aerosol emissions: Sulfate fossil fuel: 1995 EDGAR v3.2; 1890 Van Aardenne et al. (2001) Carbonaceous fossil and bio- fuels:
1996 Bond et al. (2004); 1890 Bond et al. (2007) Biomass burning: GFED v1, average of 1997-2001; for 1890 reduce tropical
burning by 0.5 Natural OM: Terpene emissions Natural sulfate: DMS oceanic biogenic, volcanoes Sea salt source model wind-speed parameterization
Can the model simulate Arctic haze?Model aerosol concentrations compared with surface
measurements
Model
Observed
Is aerosol column amount (AOD), absorption (AAOD) OK?
Model AOD, AAOD compared to AERONET
Model
Observed
Model
AERONET observed
Tests for model BC Tests for model BC deposition, removal:deposition, removal:
Model deposition compared to BC deposition compiled in Flanner et al. (2007).
Percent dry deposition from Davidson et al (1985)
Scavenging ratio from Davidson et al (1985) and Noone and Clarke (1988)
These are sensitive to removal assumptions. Here we assume 12% removal by ice phase (compared to liquid phase)
Major aerosol source regions for the Arctic
N. America
Europe
S.E. Asia
North Asia
At the surface most Arctic aerosol is from Europe and biomass burning Koch and Hansen, JGR (2005); Koch et al., JGR (2007)
Surface Arctic BC is mostly from residential and transport sources
Koch et al., J. Geophys. Res., (2007)
Residential Transport Industry Power
Most Arctic sulfate comes from the industry and power sectors Koch et al., J. Geophys. Res., (2007)
AOD
Emission trends during the past century
Emission histories
We model Arctic pollution in 1890 and 1996.
SulfurVan Aardenne et al., 2001
Black carbonBond et al., 2007
Total
Total
Europe
EuropeN. Am.
N. Am.
S.E. Asia
S.E. Asia
Europe+SE_Asia+N_Am+Russia
Europe+SE_Asia+N_Am+Russia
Arctic Surface BC: was larger in 1920, with more from Europe, N. America, and Northern Asia
McConnell et al. (2007): BC and sulfate in Greenland ice core.
Fires?
Column SO4
Column SO41995: Europe, SE Asia, Russia1920: Europe, N. America
6 pairs of Equilibrium Climate Experiments
Aerosols (LL) GHG Aerosol Effects
1995/1880 1995 Direct
1995/1880 1995 Direct+Indirect
1995/1880 1995 Direct+Indirect+BC-albedo
1995/1880 1995/1880 Direct
1995/1880 1995/1880 Direct+Indirect
1995/1880 1995/1880 Direct+Indirect+BC-albedo
1. Distinguish relative aerosol impacts on climate2. How are aerosol effects modified by GHG changes
Parameterization of black carbon effect on snow albedo
1. Model albedo depends upon model BC snow concentration as in Warren and Wiscombe (1985)
2. Snow grain size as a function of snow age and surface air temperature is calculated from Marshall (1989)
New: grains = 0.1 mm
Old: grains = 1 mm
Parameterization of aerosol indirect effects
Model cloud droplet number concentration depends upon aerosol number concentration as in Menon and Rotstayn (2006) for warm clouds only.
Surface Air Temperature Changes: 1995-1880
C
Direct Also Indirect also BC-albedo Indirect only BC-albedo only
G
HG
N
o G
HG
Surface Air Temperature Changes: 1995-1880
C
Direct Also Indirect also BC-albedo Indirect only BC-albedo only
G
HG
N
o G
HG
Surface Air Temperature Changes: 1995-1880
C
Direct Also Indirect also BC-albedo Indirect only BC-albedo only
G
HG
N
o G
HG
Surface Air Temperature Changes: 1995-1880
C
Direct Also Indirect also BC-albedo Indirect only BC-albedo only
G
HG
N
o G
HG
b) - a)
Surface Air Temperature Changes: 1995-1880
C
Direct Also Indirect also BC-albedo Indirect only BC-albedo only
G
HG
N
o G
HG
b) - a) c) - b)
Surface Air Temperature Changes: 1995-1880
C
Direct Also Indirect also BC-albedo Indirect only BC-albedo only
G
HG
N
o G
HG
Arctic [Global]
Surface Air Temperature Changes:
1995-1880Seasonality
C
GHG
No GHG
Direct
+Indirect
+BC-albedo
BC-albedo
Indirect
Surface Air Temperature Changes:
1995-1880Seasonality
GHG
No GHG
Direct
BC-albedo
IndirectEffect Peak season
Direct - spring/fall
Indirect - winter
BC-albedo + spring/fall
Indirect effectSeasonality
GHG
No GHG
Direct
BC-albedo
IndirectEffect Peak season
Indirect - winter
Low Cloud Changes: 1995-1880
%
Direct Also Indirect also BC-albedo Indirect only BC-albedo only
G
HG
N
o G
HG
GISS model: cool more low level cloudswarm fewer low level clouds
High Cloud Changes: 1995-1880
%
Direct Also Indirect also BC-albedo Indirect only BC-albedo only
G
HG
N
o G
HG
GISS model: cool more low level clouds, less high cloudswarm fewer low level clouds, more high clouds
Cloud Changes: 1995-1880 Seasonality
GHG
No GHGIndirect
Indirect effect: More low clouds, fewer high clouds during summer
Indirect
Indirect
Indirect
Radiative forcing changes: 1995-1880 Seasonality
GHG
No GHG
IndirectIndirect
Indirect
Indirect effect: LW forcing is maximum and negative in winter
Snow/Ice Changes: 1995-1880
%
Direct Also Indirect also BC-albedo Indirect only BC-albedo only
G
HG
N
o G
HG
Indirect effect: Colder temperatures, increased snow cover, brighter surfaces reduce LW radiation especially during winter when cloud cover is less. INDIRECT INDIRECT EFFECT!INDIRECT INDIRECT EFFECT!
BC-albedo effectSurface Air Temperature Changes: 1995-1880
C
Direct Also Indirect also BC-albedo Indirect only BC-albedo only
G
HG
N
o G
HG
Arctic [Global]
Snow/Ice Changes: 1995-1880
%
Direct Also Indirect also BC-albedo Indirect only BC-albedo only
G
HG
N
o G
HG
BC-albedo effect: Is larger when GHG is fixed, since increasing GHG already reduces snow/ice cover. (Also, BC deposition increase is smaller…)
BC-albedo effect: Spatial correlation between change in surface air temperature and snow/ice
Change in temperature snow/ice
G
HG
N
o G
HG
warmer temperature less snow/ice
BC-albedo effect: Cloud changes also correlate.
Change in high clouds temperature snow/ice
GH
G
No
GH
G
More high clouds, warmer temperature less snow/ice
Surface Air Temperature Changes:
1995-1880Seasonality
GHG
No GHG
Direct
BC-albedo
Effect Peak season
Direct - spring/fall
Indirect - winter
BC-albedo + spring/fall
Snow/Ice Changes: 1995-1880 Seasonality
No GHG
Direct
BC-albedo
GHG
BC-albedo
%
Direct Also Indirect also BC-albedo Indirect only BC-albedo only
G
HG
N
o G
HG
Western Arctic sea-ice changes: Arctic Climate Impact Assessment (2004)
According to the model, either an increase BC-albedo effect, or recent decreases in (sulfate) indirect effect might explain loss of sea ice in western Arctic
During the past decade:1. Most Arctic pollution aerosol is
sulfate. Most sulfate pollution is from Europe. European sulfate is decreasing. Therefore the indirect effect should be decreasing warming
2. Surface BC is also mostly from Europe is probably also decreasing, especially to the north of Europe.
3. Black carbon pollution from Asia and boreal fires may be increasing (large sources of surface and high level BC in western Arctic) and may be contributing to melting in those regions.
McConnell et al. (2007) Greenland data
Black Carbon (BC) Measurements at Alert.
-Measurements conducted by Environment Canada since 1989.
-Higher BC in winter due to “Arctic Haze”.
-Decline in trends of BC measurements since 1989 by 55% (Sharma et al., 2004).
Environment Environnement Canada Canada
-Change in BC measurementsare proportional to changes in emissions and Atmospheric transport to the Arctic.
Alert Measurement Station – 82oN, 62.5oW
Year
0.0
0.5
1.0
1.5
2.0
1988 1993 1998 2003 2008
Date
light absorption (Mm
-1)
0.00
0.05
0.10
0.15
0.20
estimated black carbon (ug/m
3)Barrow light absorption
(Ogren)
Alert BC(Sharma)
Alert Measurement Station – 82oN, 62.5oW
Year
Black Carbon (BC) Measurements at Alert.
-Measurements conducted by Environment Canada since 1989.
-Higher BC in winter due to “Arctic Haze”.
-Decline in trends of BC measurements since 1989 by 55% (Sharma et al., 2004).
Environment Environnement Canada Canada
-Change in BC measurementsare proportional to changes in emissions and Atmospheric transport to the Arctic.
0.0
0.5
1.0
1.5
2.0
1988 1993 1998 2003 2008
Date
light absorption (Mm
-1)
0.00
0.05
0.10
0.15
0.20
estimated black carbon (ug/m
3)Barrow light absorption
(Ogren)
Alert BC(Sharma)
SE Asia
Europe Russia
N. America
Biomass burning
Model surface BC from regions
Arctic haze captured by Calipso retrieval, typically seen in discrete pollution plumes
Courtesy of David Winker
Northeast coast of Greenland
WE
…Seems to originate in Northern Russia
Northeast coast of Greenland
WE
From David Winker
CALIOP (CALIPSO) 16-day orbit pattern
(from Dave Winker)
Caveats
1. Analysis is preliminary. Significance?
2. Indirect effect: warm clouds only
3. Indirect effect: very uncertain
4. BC albedo effect: how to validate?
5. Equilibrium (not transient) experiments
Next steps1. Perform 20th century transient experiments
2. Indirect effect: include ice phase parameterization, treatment of Arctic clouds
3. Perform ongoing comparison of model clouds/aerosols with CALIPSO, IPY campaign measurements
Conclusions1.1. Direct effect cools Arctic surface Direct effect cools Arctic surface
temperaturestemperatures
2.2. Indirect effect has powerful effect Indirect effect has powerful effect cooler temperatures cooler temperatures increased increased snow/ice and albedosnow/ice and albedo
3.3. BC-albedo effect slight warming BC-albedo effect slight warming (about same amount as direct effect (about same amount as direct effect cooling). Note cooling). Note BC in Arctic from BC in Arctic from 1890 to 1990 not very large. But 1890 to 1990 not very large. But there are regional differences.there are regional differences.
4.4. GHG reduces aerosol effects (due GHG reduces aerosol effects (due to GHG large effects on cryosphere).to GHG large effects on cryosphere).
5.5. Eastern Arctic sea ice loss may be Eastern Arctic sea ice loss may be from a) increased BC-albedo effect in from a) increased BC-albedo effect in that region and/or b) decrease in that region and/or b) decrease in indirect effectindirect effect
Effect SAT snow/ice
GHG +2.3 -1.6
Direct -0.3 +0.2
Indirect
+ GHG
-2.2 +3.8
-1.5 +1.6
BC-alb
+GHG
+0.5 -1.2
+0.2 -0.2
Effect SAT snow/ice
GHG +2.3 -1.6
Direct -0.3 +0.2
Indirect
+ GHG
-2.2 +3.8
-1.5 +1.6
BC-alb
+GHG
+0.5 -1.2
+0.2 -0.2
1. Biggest uncertainties:
Cloud effects in the Arctic
Indirect effect (positive or negative)
BC-albedo effect
Transport to the Arctic
CALIPSO, CloudSat, IPY campaigns will help!
2. Usefulness of “forcing” in the Arctic???
Maybe a matrix of SAT, snow/ice, sea level pressure
3. Local vs extrapolar forcing:
GHG, warm-cloud indirect effect: both important
BC-albedo, ice-cloud indirect effect: local
4. Delay the onset of spring melt?
Analyze an ice core in Siberia to determine whether Asia is a big source, this is where BC is increasing
Answers to Questions
Support from Clean Air Task Force, NASA RS and MAP
Indirect Effect
SW, LW cloud forcing: -0.4, -0.2 W/m2, no GHG
SW, LW cloud forcing: -0.04, -0.4 W/m2, GHG
SW, LW forcing: -2.1, -1.9 W/m2 > cloud forcing due to surface albedo increase
reff = -1.1m warm stratiform (no GHG)
reff = -0.01 warm stratiform (GHG; LWC, LWP increase)
CDNC = 30 cm-3
Direct Effect
SW clear sky forcing (no GHG) = -0.7 W/m2
SW all-sky forcing (no GHG) = -0.4 W/m2
SW absorption (TOA-surface) about 0.8 W/m2 (no GHG)
or about 1.5 W/m2 (GHG; due to loss of clouds?)
Instantaneous SW forcing from tracer run: -0.2 W/m2
Instantaneous SW absorption from tracer run: 1.3 W/m2
BC-albedo Effect
Radiative forcing (no GHG):
0.3 W/m2 SW
0.2 W/m2 LW
These would be larger than ‘instantaneous’ due to increased clouds.
With GHG little effect: albedo is already decreasing
Flanner et al (2007): 0.1 W/m2
albedo (no GHG)
-0.12 %
SW forcing Changes: 1995-1880
Direct Also Indirect also BC-albedo Indirect only BC-albedo only
G
HG
N
o G
HG
W/m2
LW forcing Changes: 1995-1880
Direct Also Indirect also BC-albedo Indirect only BC-albedo only
G
HG
N
o G
HG
W/m2
Radiative forcing changes: 1995-1880 Seasonality
GHG
No GHG
Direct
Indirect
BC-albedo
Indirect
BC-albedo
Sea level pressure Changes: 1995-1880
mb
Direct Also Indirect also BC-albedo Indirect only BC-albedo only
G
HG
N
o G
HG
Sea level pressure changes: 1995-1880
Direct Also Indirect also BC-albedo Indirect only BC-albedo only
G
HG
N
o G
HG
mb
Sea level pressure changes: 1995-1880 Seasonality
GHG
No GHG
Indirect
BC-albedo
Sulfur trend check: model vs sulfur in Greenland ice core (Fischer et al., 1998)
Core observations
Model S in snow
Model x 6
Change in aerosol optical depths between 1880 and 1995
Cloud Changes: 1995-1880 Seasonality
GHG
No GHG
Direct+Indirect
+BC-albedo
BC-albedo
Indirect
Low Cloud Changes: 1995-1880
%
Direct Also Indirect also BC-albedo Indirect only BC-albedo only
G
HG
N
o G
HG
High Cloud Changes: 1995-1880
%
Direct Also Indirect also BC-albedo Indirect only BC-albedo only
G
HG
N
o G
HG
Snow/ice Changes: 1995-1880
%
Direct Also Indirect also BC-albedo Indirect only BC-albedo only
G
HG
N
o G
HG
SW forcing Changes: 1995-1880
%
Direct Also Indirect also BC-albedo Indirect only BC-albedo only
G
HG
N
o G
HG
W/m2
LW forcing Changes: 1995-1880
W/m2
Direct Also Indirect also BC-albedo Indirect only BC-albedo only
G
HG
N
o G
HG
Cloud Changes: 1995-1880 Seasonality
GHG
No GHG
Direct
+BC-albedo
BC-albedo
Indirect
Effect SAT low cloud high cloud
Direct - spring/fall warm summer (w/o GHG) Cool spring/fall
Indirect - winter Cool summer (w/o GHG) Cool summer
BC-alb + spring/fall Warm spring/fall (w/o GHG) ?
In the column (AOD) most Arctic aerosol (sulfate) is from Europe, most absorbing aerosol (BC) is from SE Asia
Koch et al., J. Geophys. Res. (2007); Koch and Hansen, J. Geophys. Res. (2005)
AODx10
BC in Column (AOD)
Column BC: More BC from Europe and N. America
Evolution of Arctic BC:1920 - 1996 Pollution transport (from SE Asia) is shifted to higher altitudes. Implications for stability, cloud heights..