AEROSOL COMPONENT
MODELSan initial comparison
PARISJune 2003
AEROSOL in global models
modeling of aerosol climatic impacts is done
- at coarse resolution (ca 30x30)
- in many individual steps- individually by aerosol type
many processespossibilities for errors
modeling of aerosol climatic impacts is done- at coarse resolution (ca 3*3deg)- in many individual steps- individually by aerosol type
many processespossibilities for errors
AOT data-sets
• satellites:
– MODIS– TOMS– POLDER– AV. 2Ch– AV. 1Ch
• ground:
– AERONET
yearly avg.
testing models - PAST
• monthly statistics• 8 models !
0
0.05
0.1
0.15
0.2
0.25
component combined aot (0.55um)
EC
GO
MI
GI
CC
GR
UL
NC
MO
TO
A,n
A,g
PO
Aero
---Models---- Data
best satellite AERONET
AOT
testing models - PAST • Consistency? … just look at global averages for mass (turquoise background) and
opt.depth (brown background) … and differences in mass aot !
00.5
11.5
22.5
33.5
4
org. carbon mass (mg/m2)
ECGOMIGICCGRULHANC
0
0.01
0.02
0.03
0.04
0.05
0.06
dust aot (0.55um)
ECGOMIGICCGRULHANC
01020304050607080
dust mass (mg/m2)
ECGOMIGICCGRULHANC
0
0.005
0.01
0.015
0.02
0.025
0.03
org. carbon aot (0.55um)
ECGOMIGICCGRULHANC
0
0.1
0.2
0.3
0.4
0.5
0.6
blk carbon mass (mg/m2)
ECGOMIGICCGRULHANC
0
0.001
0.002
0.003
0.004
0.005
0.006
0.007
blk carbon aot (0.55um)
ECGOMIGICCGRULHANC
01020304050607080
sea-salt mass (mg/m2)
ECGOMIGICCGRULHANC
0
0.01
0.02
0.03
0.04
0.05
sea-salt aot (0.55um)
ECGOMIGICCGRULHANC
0
2
4
6
8
10
sulfate mass (mg/m2)
ECGOMIGICCGRULHANC
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
sulfate aot (0.55um)
ECGOMIGICCGRULHANC
Dust Sulfate Sea-Salt org. Carbon black Carbon
first Impressions
• Models agree on – high carbon in central Africa (~ 60%)– high sulfate for Europe and E.Asia ( ~ 45%)– dominant sea-salt in mid-latitudes of the SH– dominant dust over N.Africa and central Asia
• Models disagree on– source strength for dust and for biomass burning– carbon contrib. for tropics and over urban regions– transport (contributions in off-source regions)– sea-salt contributions over oceans
Relative Model Tendencies
• ECHAM4 strong du- seasonality, low ss and bc MEE, rh-sensitivity
• GOCART strong transport, strong du, strong oc-, bc- seasonality
• MIRAGE strong su (+oc), weak on trop.sources, high lat. bias
• GISS low on mass (except for su), strong du MEE
• CCSR strong oc-, bc-, du- seasonality (+ sources), weak transport• Grantour lowest oc/bc -mass ratios, strong ss MEE and opt.depth• ULAQ strongest su-, oc- urban sources, weak transport (bc)
• NCAR weak bc, low oc und bc MEE, high ss MEE
• HadHam strong su- seasonality, weak on bc
su -sulfate, du -dust, ss -seasalt, oc -org.carbon, bc -black carbon
Quantification- local -
• % hits vs AERONET -aot• Yearly average: +/- 20%• Season-phase: +/- 1 month• Season-strength: +/- 50%
0
10
20
30
40
50
60
70
80
s-phase s-strength yr.avg s-p,s-s s-p,s-s,av[%]
ECGOMIGICCGRULMO
20 AERONET sites
MODIS aots are usually larger than aots fromAERONET (..mainly due to snow in winter)
Quantification- regional -
• % hits vs MODIS/TOMS- aot• Yearly average: +/- 20%• Seasonality phase : +/- 1
month (for 3mo-avg max)
01020
30405060
708090
sea.-phase average av+ph [%]
ECGOMIGICCGRUL
evaluated regions
• most models predict lower aots than MODIS/TOMS
• discrepancies are largest in remote regions (sea-salt size?)
• but larger simulated aots for Europe (old outdated sources?)
TOPICS
• Compare size assumptions for aerosol types• reduce differences in the MASS to AOT conversion
• Remove size ambiguity from rel.hum fields?• prescribe relative humidity fields (ECWMF) in aot conversions
• Aerosol vertical distribution and lifetime?• compare for simulations with identical sources / sizes
• What is known about removal processes?• compare for simulations with identical sources / sizes
• Is the chemistry (gas to particle) correct?• compare results with and without chemistry