Evaluation of Terra MODIS Aerosol and Water Vapor Measurements Using ARM SGP Data Richard Ferrare, Lorraine Heilman Brasseur, Dave Turner, Dave Whiteman, Lorraine Remer, Bo-Cai Gao MODIS Atmosphere Group Meeting December 17, 2001
Evaluation of Terra MODIS Aerosol and Water VaporMeasurements Using ARM SGP Data
Richard Ferrare, Lorraine Heilman Brasseur,Dave Turner, Dave Whiteman, Lorraine Remer, Bo-Cai Gao
MODIS Atmosphere Group MeetingDecember 17, 2001
Outline
• MODIS aerosol comparisons
• MODIS water vapor comparisons
• Aerosol and water vapor profile diurnal variability
• Average aerosol and water vapor profiles during Terra overpass
• DOE ARM Aerosol IOP experiment
• Summary
DOE ARM SGP Measurements Used for MODIS Validation
• Aerosol optical thickness (AOT)• Cimel Sun photometer (Cimel)• Multi-Filter Rotating Shadowband Radiometer (MFRSR)
• Precipitable Water Vapor (PWV)• Microwave Radiometer (MWR)• CART Raman Lidar (CARL)• Cimel Sun photometer (Cimel)
• Aerosol and Water Vapor Profiles• CART Raman Lidar (CARL)
• Used MODIS, SGP data from March 2000 through September 2001
Example MODIS and SGP Aerosol Measurements
MODIS AOT (470 nm) CARL aerosol extinction profiles
CARL and Cimel AOT
0.01
0.1
1
20012000OctJulAprJanOctJulAprJan
AO
T
AOT (470 nm) MODIS Cimel MFRSR
0.01
0.1
1
20012000OctJulAprJanOctJulAprJan
AO
T
AOT 670 nm MODIS Cimel MFRSR
Aerosol optical thickness over SGP during MODIS measurements
• SGP AOT measured by Cimel, MFRSR• AOT generally low (~between 0.02-0.3)
470 nm 670 nm
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.70.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7 Slope=0.74Intercept=0.11N=29 R=0.58rms diff=0.09 (68%)bias diff=0.05 (38%)
- - - - MODIS AOT uncertainty
MO
DIS
AO
T (
470
nm
)
SGP AOT
Cimel (470 nm) MFRSR (500 nm) linear fit 95% pred. level
0.0 0.1 0.2 0.3 0.40.0
0.1
0.2
0.3
0.4
N=29 R=0.18rms diff=0.067 (67%)bias diff=-0.01 (-10%)
- - - MODIS AOT uncertainty
MO
DIS
AO
T (
660
nm
)
SGP AOT
Cimel (670 nm) MFRSR (674 nm)
MODIS AOT vs. SGP AOT
• (470 nm) • MODIS biased high at low AOT • differences generally within MODIS uncertainty
• (660 nm)• large scatter, no linear trend for low AOT• differences generally within MODIS uncertainty
ARM SGP (March 2000-Sept 2001) clear-sky subset
Cimel AOT (340 nm)
CARL A
OT (355 nm
)
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
cloudf1>1 and car470a2>0 and car470a2<2 and cim470ao>0 and cim470ao<2
Slope=0.94Intercept=0.01R=0.80 N=66rms diff=0.07 (33%)
ARM SGP (March 2000-Sept 2001) clear-sky subset
Cimel AOT (470)
CART Ram
an Lidar AO
T (470)
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40
cloudf1>1 and car470a2>0 and car470a2<2 and cim470ao>0 and cim470ao<2
Slope=0.97Intercept=0.02R=0.79 N=66rms diff=0.05 (40%)
CART Raman lidar AOT (355 nm), Cimel AOT (340 nm) agree within about5% in mean
Using Angstrom exponent (340-500 nm) from CimelAOT to extrapolate CARL AOT from 355 nm to 470 nmgives generally good results
CART Raman lidar AOT vs. Cimel AOT
ARM SGP (March 2000-Sept 2001) MODIS clear-sky subset
Cimel Angstrom Exponent (440 - 670 nm)
MO
DIS A
ngstrom Exponent (470 - 660 nm
)
0
1
2
3
4
5
0 1 2 3 4 5
cloudf1>1 and car470a2>0 and car470a2<2 and cim470ao>0 and cim470ao<2 and mod470ao>0 and mod470ao<2
ARM SGP (March 2000-Sept 2001) clear-sky subset
Cimel Angstrom Exponent (440-670 nm)
Cimel A
ngstrom Exponent (3
40-5
00 nm
)
0.0
0.5
1.0
1.5
2.0
2.5
3.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0
Slope=0.51Intercept=0.95R=0.69 N=93rms diff=0.33 (30%)
ARM SGP (March 2000-Sept 2001) clear-sky subset
Cimel AOT (470 nm)
CART Ram
an Lidar AO
T (470 nm
)
0.0
0.1
0.2
0.3
0.4
0.5
0.0 0.1 0.2 0.3 0.4 0.5
Used Cimel (340-500) Angstrom ExpSlope=0.94 Intercept=0.01R=0.80 N=66rms diff=0.07 (33%)
Used MODIS (470-660) Angstrom Exp.Slope=0.32 Intercept=0.06R=0.37 N=26rms diff=0.07 (33%)
Using MODIS(470-660) Ang. ExpoUsing Cimel (340-500) Ang. Expo
Can not use CARL AOT (355 nm) alone toevaluate MODIS AOT (470 nm) because wecan not use MODIS Angstrom exponent δ(470-660 nm) to extrapolate CARL AOTfrom 355 nm to 470 nm
Cimel δ (350-500 nm) ≠ Cimel δ (470-660 nm) MODIS δ (470-660 nm) ≠ Cimel δ (470-660 nm)
CARL AOT at 470 nm
-0.4
-0.3
-0.2
-0.1
0.0
0.1
0.2
0.3
0.4
MODIS lower
MODIS higher
20012000OctJulAprJanOctJulAprJan
AO
T D
iffe
ren
ce
MODIS (470 nm) - Cimel (470 nm) MODIS (470 nm) - MFRSR (500 nm) MODIS (660 nm) - Cimel (670 nm) MODIS (660 nm) - MFRSR (674 nm) -300
-200
-100
0
100
200
300
400
500
MODIS lower
MODIS higher
20012000OctJulAprJanOctJulAprJan
AO
T D
iffe
ren
ce (
%)
MODIS (470) - Cimel (470) MODIS (470) - MFRSR (500) MODIS (660) - Cimel (670) MODIS (660) - MFRSR (674)
(MODIS - SGP) AOT difference vs. date
No obvious trend with time at either 470 or 660 nm
Absolute differences Relative differences
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35-0.2
-0.1
0.0
0.1
0.2
0.3
0.4
MODIS lower
MODIS higher
AO
T d
iffe
ren
ce
SGP AOT
MODIS (470) - Cimel (470) MODIS (470) - MFRSR (500) MODIS (660) - Cimel (670) MODIS (660) - MFRSR (674)
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35-200
-100
0
100
200
300
400
500
MODIS lower
MODIS higher
AO
T d
iffe
ren
ce (
%)
SGP AOT
MODIS 470 - Cimel 470 MODIS 470 - MFRSR 500 MODIS 660 - Cimel 670 MODIS 660 - MFRSR 674
(MODIS - SGP) AOT difference vs. AOT
Absolute differences Relative differences
• Relative differences decrease with increasing AOT• (470 nm) differences < 0-20% (MODIS higher) for AOT > 0.1
MODIS near IR PWVMODIS IR PWV
CARL water vaporprofiles
CARL and MWR PWV
0
1
2
3
4
5
6
20012000OctJulAprOctJulApr JanJan
PW
V (
cm)
MODIS near IR MODIS IR MWR
Precipitable water vapor over SGP during MODIS measurements
PWV varies between 5 – 50 mm
-80
-60
-40
-20
0
20
40
60
80
MODIS near IR drier
MODIS near IR wetter
OctJulAprJanOctJulAprJan20012000
PW
V D
iffe
ren
ce (
%)
Date
MODIS Near IR - SGP MWR MODIS Near IR - SGP Cimel
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
MODIS near IR drier
MODIS near IR wetter
20012000OctJulAprJanOctJulAprJan
PW
V d
iffe
ren
ce (
cm)
MODIS near IR - MWR MODIS near IR - Cimel
(MODIS near IR - SGP) PWV difference vs. date
• Nov. 1 2000• modified water vapor transmittance lookup table to improve “continuum”absorption• changed to side b electronics improved 1.24 µm radiometric calibration
• changes reduced MODIS near IR overestimates of PWV • Jun. 1 2000 regenerated lookup tables using new HITRAN2000 database
• have not yet analyzed sufficient data to evaluate change
ARM SGP (March 2000-Sept 2001) MODIS clear-sky subset
MWR PWV (cm)
MO
DIS near IR PW
V (cm
)
0
1
2
3
4
5
6
7
8
0 1 2 3 4 5 6 7 8
After Nov. 1, 2000Slope = 1.15 Intercept = -0.08 cmN = 55 R = 0.99rms diff= 0.25 cm (16%)bias diff (modis nir-mwr)=0.156 cm (10%)
Before Nov. 1, 2000Slope = 1.63 Intercept = -0.623 cmN = 25 R = 0.93rms diff= 1.1 cm (54%)bias diff (modis nir-mwr)=0.67 cm (33%)
MODIS near IR PWV comparison vs. Microwave radiometer (MWR) After Nov. 1, 2000
• bias difference decreases from 0.67 cm (33%) to 0.16 cm (10%)
• rms difference decreases from 1.1 cm (54%) to 0.25 cm (16%)
ARM SGP (March 2000-Sept 2001) MODIS clear-sky subset
MWR PWV (cm)
PWV
Diff (%
) (MO
DIS near IR - M
WR)
-80
-60
-40
-20
0
20
40
60
80
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5
Before Nov. 1, 2000After Nov. 1, 2000
MODIS near IR wetter
MODIS near IR drier
MODIS near IR PWV comparison vs. Microwave radiometer (MWR) After Nov. 1, 2000
• bias difference decreases from 0.67 cm (33%) to 0.16 cm (10%)
• rms difference decreases from 1.1 cm (54%) to 0.25 cm (16%)
ARM SGP (March 2000-Sept 2001) MODIS clear-sky subset
MWR PWV (cm)
Cimel PW
V (cm
)
0
1
2
3
4
5
0 1 2 3 4 5
Slope = 1.07Intercept = -0.029 cmN = 44R = 0.996rms diff= 0.156 cm (9%)bias diff (cimel-mwr)=0.089 cm (5%)
MWR vs. Cimel PWV comparison
If you use Cimel PWV to validate MODIS…
PWV derived using Cimel processing using LOWTRAN 7 database is about 5-8% higherthan ARM MWR
[If Cimel PWV processing usedGiver corrected HITRANdatabase (which increased 940 nmline strengths by ~14%), thenCimel PWV decreases by about14%.]
See Schmid et al. Comparison ofcolumnar water-vapormeasurements from solartransmittance methods. AppliedOptics, Vol. 40, No. 12, 1886-1896(2001).
ARM SGP (March 2000-Sept 2001) MODIS clear-sky subset
Cimel PWV (cm)
MO
DIS near IR (PW
V)
0
1
2
3
4
5
6
7
8
0 1 2 3 4 5 6 7 8
After Nov. 1, 2000Slope = 1.05 Intercept = -0.02 cmN = 32 R = 0.997rms diff= 0.11 cm (7%)bias diff (modis nir-mwr)=0.051 cm (3%)
Before Nov. 1, 2000Slope = 1.69 Intercept = -0.794 cmN = 12 R = 0.95rms diff= 1.31 cm (56%)bias diff (modis nir-mwr)=0.82 cm (35%)
After Nov. 1, 2000• bias difference decreases from 0.82 cm (35%) to 0.05 cm (3%)
• rms difference decreases from 1.3 cm (54%) to 0.11 cm (3%)
MODIS near IR PWV comparison vs. Cimel
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5-1.0
-0.8
-0.6
-0.4
-0.2
0.0
0.2
0.4
0.6
0.8
1.0After Nov. 1, 2000
mean diff = 0.11 cmrms diff = 0.18 cm
MODIS near IR drier
MODIS near IR wetter
PW
V d
iffe
ren
ce (
cm)
SGP PWV (cm)
MODIS near IR - MWR MODIS near IR - Cimel
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5-30
-20
-10
0
10
20
30
MODIS near IR wetter
MODIS near IR drier
mean diff = 6.3%rms diff = 8.2%
After Nov. 1, 2000
PW
V d
iffer
ence
(%
)
SGP MWR (cm)
MODIS near IR - MWR MODIS near IR - Cimel
(MODIS near IR - SGP) PWV difference vs. PWV
• MODIS near IR PWV about 5-15% higher than SGP MWR and Cimel PWV• No systematic variations with PWV
ARM SGP (March 2000-Sept 2001) MODIS night clear-sky subset
MWR PWV (cm)
CARL PW
V (cm
)
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
Slope = 1.002Intercept = -0.002 cmN = 42R = 0.999rms diff= 0.045 cm (3%)bias diff (carl-mwr)=0.0012 cm (0.1%)
CART Raman Lidar water vapor measurements• Nighttime profiles extend through troposphere (0.06 – 10 km) permitting PWV retrievals• Daytime profiles limited by background skylight (0.06 - ~ 3.5 km) – no PWV retrieved• CARL calibrated so that CARL PWV matches MWR PWV• CARL PWV has excellent long-term stability when compared with MWR
ARM SGP (March 2000-Sept 2001) MODIS clear-sky subset
MWR PWV (cm)
MO
DIS IR PW
V (cm
)
0
1
2
3
4
5
6
0 1 2 3 4 5 6
irpwv>0 and irpwv<7 and mwrpwv>0 and mwrpwv<7 and radius<30 and cldflag1>1 and mwrliq<0.01
Slope = 0.73 Intercept = 0.66 cmN = 150 R = 0.77rms diff= 0.684 cm (38%)bias diff (modis ir-mwr)=0.18 cm (21%)
MODIS IR PWV vs. MWR PWV
• (daytime+nightime) offset (intercept) = 6.8 mm • bias difference ~ 2 mm (~20%) (MODIS IR wetter), rms difference 7 mm (~40%)
ARM SGP (March 2000-Sept 2001) MODIS clear-sky subset
CART Raman Lidar PWV (cm)
MO
DIS IR PW
V (cm
)
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
irpwv>0 and irpwv<7 and carlpwv>0 and carlpwv<7 and radius<30 and cldflag1>1 and mwrliq<0.01 and daynight=0
Slope = 0.56 Intercept = 0.72 cmN = 41 R = 0.78rms diff= 0.524 cm (34%)bias diff (modis ir-carl)=0.28 cm (18%)
• (nightime) offset (intercept) = 7 mm • bias difference ~ 3 mm (~20%) (MODIS IR wetter), rms difference 5 mm (~30%)
MODIS IR PWV vs. CARL PWV
ARM SGP (March 2000-Sept 2001) MODIS clear-sky subset
Cimel PWV (cm)
MO
DIS IR PW
V (cm
)
0
1
2
3
4
5
6
0 1 2 3 4 5 6
irpwv>0 and irpwv<7 and cimpwv>0 and cimpwv<7 and radius<30 and cldflag1>1 and mwrliq<0.01 and daynight=1
Slope = 0.92 Intercept = 0.53 cmN = 44 R = 0.875rms diff= 0.694 cm (39%)bias diff (modis ir-cimel)=0.385 cm (21%)
MODIS IR PWV vs. Cimel PWV
• (daytime) offset (intercept) = 5 mm • bias difference ~ 4 mm (~20%) (MODIS IR wetter), rms difference 7 mm (~40%)
-100
-50
0
50
100
150
200
MODIS IR drier
MODIS IR wetter
20012000OctJulAprJanOctJulAprJan
PW
V d
iffe
ren
ce (
%)
Date
MODIS IR - MWR MODIS IR - CARL MODIS IR - Cimel
-2.5
-2.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
MODIS IR drier
MODIS IR wetter
20012000OctJulAprJanOctJulAprJan
PW
V d
iffe
ren
ce (
cm)
Date
MODIS IR - MWR MODIS IR - CARL MODIS IR - Cimel
(MODIS IR - SGP) PWV difference vs. Date
• relative error increases in winter due to low PWV
Absolute differences Relative differences
0 1 2 3 4 5-2.5
-2.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
MODIS IR drier
MODIS IR wetter
mean diff = 0.19 cmrms diff = 0.63 cm
PW
V d
iffe
ren
ce (
cm)
(MO
DIS
IR -
SG
P)
SGP PWV (cm)
MODIS IR - MWR MODIS IR - CARL MODIS IR - Cimel
0 1 2 3 4 5-100
-50
0
50
100
150
200
mean diff = 28%rms diff = 52%
MODIS IR drier
MODIS IR wetter
PW
V D
iffe
ren
ce (
%)
(MO
DIS
IR -
SG
P)
SGP PWV (cm)
MODIS IR - MWR MODIS IR - CARL MODIS IR - Cimel
(MODIS IR - SGP) PWV difference vs. PWV
• large relative error at low PWV due to MODIS IR PWV offset (floor around 5-7 mm)• mean difference ~ 2 mm (~25%) (MODIS wetter), rms difference 6 mm (~50%)
0 1 2 3 4 50
1
2
3
4
5
NightSlope=0.55Intercept=0.75 cmR=0.77 N=113rms diff=0.45 cm
MO
DIS
IR P
WV
(cm
)
SGP PWV (cm)
MWR CARL linear fit 95% prediction level
0 1 2 3 4 50
1
2
3
4
5
DaySlope=0.99Intercept=0.46 cmR=0.87 N=78rms diff=0.54 cm
MO
DIS
IR P
WV
(cm
)
SGP PWV (cm)
MWR Cimel linear fit 95% prediction level
(MODIS IR - SGP) PWV daytime vs. nighttime performance
MODIS IR PWV has better agreement with SGP PWV for daytime measurements (smaller offset, increase in slope closer to unity, higher linear correlation)
Night Day
ARM SGP (March 2000-Sept 2001) MODIS clear-sky subset
MWR PWV (cm)
MO
DIS IR PW
V (cm
)
0
1
2
3
4
5
0 1 2 3 4 5
NightDay
DaySlope = 0.94 Intercept = 0.51 cmN = 78 R = 0.87rms diff= 0.68 cm (41%)bias diff (modis ir-mwr)=0.65 cm (39%
NightSlope = 0.54 Intercept = 0.76 cmN = 72 R = 0.76rms diff= 0.68 cm (36%)bias diff (modis ir-mwr)=0.24 cm (12%
(MODIS IR - SGP) PWV daytime vs. nighttime performance
MODIS IR PWV has better agreement with SGP MWR PWV for daytime measurements (smaller offset, increase in slope closer to unity, higher linear correlation)
(This time used single instrument, MWR, to look at diurnal changes)
Winter Spring
Summer
Fall
Aer
osol
Ext
inct
ion
(km
-1)
Average Diurnal Variation of Aerosol Extinction Profiles
• CARL aerosol extinction profiles averaged over 837 days between March 98 - October 01• Higher extinction concentrated over smaller vertical extent at night
Sunrise
Sunrise
Sunrise
Sunrise
Sunset Sunset
Sunset Sunset
Ter
ra T
erra
Aqu
a
Aqu
a
Winter
SummerA
eros
ol E
xtin
ctio
n (k
m-1
)
Average Diurnal Variation of Aerosol Extinction Profiles and AOT• Large changes in vertical profile; smaller changes in AOT (st. dev ~ 10%)
Sunrise
Sunrise
Sunset
Sunset
Ter
ra
0 2 4 6 8 10 12 14 16 18 20 22 240.0
0.1
0.2
0.3
0.4
0.5
0.6
Aq
ua
Aq
ua
Aq
ua
Te
rra
Fall (day) Fall (night) Winter (day) Winter (night)
Aero
so
l O
pti
cal T
hic
kn
ess (
355 n
m)
Time (UT)
Summer (day) Summer (night) Spring (day) Spring (night)
Aqu
a AOT
Wat
er V
apor
Mix
ing
Rat
io (
g/kg
)
Winter Spring
Summer Fall
Sunrise
Sunrise
Sunrise
Sunrise
Sunset Sunset
Sunset Sunset
Ter
ra T
erra
Aqu
a
Aqu
a
Average Diurnal Variation of Water Vapor Profiles
• generally smaller diurnal changes than aerosol extinction near the surface• larger diurnal changes in spring and summer near top of mixed layer
0 2 4 6 8 10 12 14 16 18 20 22 240
10
20
30
40
50
60
Aq
ua
Terr
a
Fall (day) Fall (night) Winter (day) Winter (night)
Inte
gra
ted
Wa
ter
Va
po
r (a
rb.
un
its
)
Time (UT)
Summer (day) Summer (night) Spring (day) Spring (night)
Wat
er V
apor
Mix
ing
Rat
io (
g/kg
)
Winter
Summer
Sunrise
Sunrise
Sunset
Sunset
Ter
ra
Average Diurnal Variation of Water Vapor Profiles
• smaller diurnal changes in profiles and integrated water vapor (st. dev ~3-5%)
Aqu
a
Integrated water vapor
Rel
ativ
e H
umid
ity
(%)
Winter Spring
Summer
Fall
Sunrise
Sunrise
Sunrise
Sunrise
Sunset Sunset
Sunset Sunset
Ter
ra T
erra
Aqu
a
Aqu
a
Average Diurnal Variation of Relative Humidity Profiles• RH computed using CARL water vapor, AERI+model temperatures• Increase in aerosol extinction near surface at night correlated to RH• Terra, Aqua measurements occur when RH (and aerosol size, composition?) vary with z
0
1
2
3
4
5
6
7
30 40 50 60 70
Relative Humidity (%)
Alti
tude
(km
) Summer Spring Fall Winter
0
1
2
3
4
5
6
7
0 2 4 6 8 10 12 14
Water Vapor Mixing Ratio (g/kg
Alt
itu
de
(km
)
Summer Spring Fall Winter
0
1
2
3
4
5
6
7
0.00 0.05 0.10 0.15 0.20 0.25
Aerosol Extinction (355 nm) (km-1)
Alt
itu
de
(km
)
Summer Spring Fall Winter
Average Profiles at time of Terra Overpass
• Aerosol extinction:• profile shape varies with season• scale height varies with season and AOT
• Water Vapor• profile shape and scale height are constant
See Turner, Ferrare, Brasseur, GRL, 28, 4441-4444, 2001.
DOE ARM Proposed Aerosol Experiment• Aerosol IOP (Intensive Operations Period) (~May, 2003 at SGP)
Objectives:•Diffuse Flux closure
Use new and additional measurements of aerosol absorption and extinction toaccurately constrain aerosol absorption to resolve differences between measuredand modeled diffuse radiation
• CCNInvestigate relationship between CCN number concentrations at the surface andcloud base, and determine whether profiles of aerosol extinction and RH can beused to determine cloud nucleating properties just below cloud base
• AOT closureCharacterize routine (Raman, MPL) lidar and aircraft in situ profilingmeasurements of aerosol scattering and extinction and how aerosol humidificationfactor varies with altitude
Measurements:• Use one (possibly two) instrumented aircraft• Additional surface aerosol and radiation measurements• Possible coordination with DOE Tropospheric Aerosol Program (TAP) (i.e. chemistry)
Potential MODIS Terra, Aqua validation/science opportunity
Summary
• Aerosol Optical Thickness (AOT)- low range of AOT (0-0.3) hampers full evaluation and gives large rms differences- MODIS AOT (470 nm) higher by 30-40% for all AOT, 10-20% for AOT>0.1- MODIS AOT (660 nm) not well correlated to SGP AOT for low AOT- comparisons show results generally fall within MODIS AOT uncertainties
• Precipitable Water Vapor (PWV)- before Nov. 1, 2000 MODIS near-IR biased high by about 35-40%- after Nov. 1, 2000 MODIS near-IR biased high by ~10%- relatively small (10-20%) rms differences- MODIS IR has apparent offset (PWV floor around 5-7 mm)- MODIS IR biased high 2-4 mm(~10-20%) due to offset, rms diff ~ 6 mm (50%)- MODIS IR daytime retrievals in better agreement with SGP than nighttime
• Vertical Variability of Aerosols- Raman lidar profiles show diurnal variability in aerosol, water vapor profiles
- diurnal variability of AOT ~ 10% (st. dev), PWV ~ 3-5% (st. dev)- average aerosol extinction profiles vary with season and AOT- average water vapor profiles have constant shape and scale with PWV
• Proposed DOE ARM Aerosol IOP (SGP, ~ May 2003)- Diffuse radiation (aerosol absorption) and AOT closure, CCN- Potential for joint Terra/Aqua validation/science
Recent Publications
Turner, D.D., W.F. Feltz, and R.A. Ferrare, Continuous water vapor profiles from operational ground-based active and passiveremote sensors, Bull. Amer. Meteor. Soc., 81, 1301-1317, 2000.
Kato, S., M.H. Bergin, T.P. Ackerman, T.P. Charlock, E.E. Clothiaux, R.A. Ferrare, R.N. Halthore, N. Laulainen, G.G. Mace, J.Michalsky, and D.D. Turner, A comparison of the aerosol optical thickness derived from ground-based and airbornemeasurements, J. Geophys. Res., 105, No. D11, 14701-14717, 2000.
Peppler, R.A., C.P. Bahrmann, J.C. Barnard, J.R. Campbell, M.-D. Cheng, R.A. Ferrare, R.N. Halthore, L.A. Heilman, D.L.Hlavka, N.S. Laulainen, C.-J., Lin, J.A. Ogren, M.R. Poellot, L.A. Remer, K. Sassen, J.D. Spinhirne, M.E. Splitt, D.D. Turner,ARM Southern Great Plains Site Observations of the Smoke Pall Associated with the 1998 Central American Fires, Bull. Amer.Meteor. Soc., 81, 2563-2592, 2000.
Whiteman, D.N., G. Schwemmer, D. O'C. Starr, K.D. Evans, B. Demoz, T. Berkoff, S.H. Melfi, M. Cadirola, and G. Jedlovec,"The use of Raman Lidar in Cloud Studies", in Advances in Laser Remote Sensing, Selected Papers Presented at the 20thInternational Laser Radar Conference (ILRC), Vichy, France, 10-14 July 2000, A. Dabas, C. Loth, and J. Pelon, eds., Ecolepolytechnique, France, pp. 271-274, 2001.
Turner, D.D., R.A. Ferrare, L.A. Heilman, T. T. Tooman, A Two Year Climatology of Water Vapor and Aerosols in the LowerTroposphere Measured by a Raman Lidar, in Advances in Laser Remote Sensing, Selected Papers Presented at the 20thInternational Laser Radar Conference (ILRC), Vichy, France, 10-14 July 2000, A. Dabas, C. Loth, and J. Pelon, eds., Ecolepolytechnique, France, pp. 309-312, 2001.
Whiteman, D. N., K. D. Evans, B. Demoz, D. O'C. Starr, D. Tobin, W. Feltz, G. J. Jedlovec, S. I. Gutman, G. K. Schwemmer, M.Cadirola, S. H. Melfi, F. J. Schmidlin, 2001: Raman lidar measurements of water vapor and cirrus clouds during the passage ofhurricane Bonnie, J. of Geophys. Res., 106, No. D6, 5211-5225.
Ferrare, R.A., D.D. Turner, L.A. Heilman, O. Dubovik, and W. Feltz, Raman Lidar Measurements of the Aerosol Extinction-to-Backscatter Ratio Over the Southern Great Plains, J. Geophys. Res., 106, 20333-20347, 2001.
Turner, D.D., R.A. Ferrare, and L.A. Brasseur, Average aerosol extinction and water vapor profiles over the Southern GreatPlains, Geophys. Res. Letters, 28, 4441-4444, 2001.
Turner, D.D., R.A. Ferrare, L.A. Heilman, W.F. Feltz, and T. Tooman, Automated Retrievals of Water Vapor and AerosolProfiles over Oklahoma from an Operational Raman Lidar, J. Atmos. Oceanic Tech., in press, July, 2001.
ARM SGP (March 2000-Sept 2001) MODIS clear-sky subset
Sensor zenith Angle (deg)
PWV
difference (%) (M
OD
IS IR - MW
R)
-100
-50
0
50
100
150
200
0 10 20 30 40 50 60 70
NightDay
(MODIS IR - SGP) PWV difference vs. sensor zenith angle
• Nearly Continuous Operation• Nd:YAG (355 nm) (day/night)
– 12 W• 61 cm telescope• Wavelengths
– Rayleigh/Aerosol (355 nm)– Depolarization (355 nm)– Raman water vapor (408 nm)– Raman nitrogen (387 nm)
• 39 meter range resolution• low, high sensitivity channels• measures
• water vapor and aerosolprofiles
• precipitable water vapor andaerosol optical thickness
• aerosol and clouddepolarization
Additional information: http://www.arm.gov/docs/instruments/static/rl.html
Southern Great Plains (SGP) CART Raman Lidar (CARL)
Raman lidar observations of the aerosol vertical variability
17.0 17.5 18.0 18.5 19.0 19.5 20.0
7
6
5
4
3
2
1
Date (May, 1998)
Aerosol Extinction (km-1)
0
1
2
3
4
5
6
7
0.0 0.1 0.2 0.3 0.4 0.5
Aerosol Extinction (km-1)
Alt
itu
de
(km
)
30 40 50 60 70 80 90 100 110 120
May 18, 1998 19:50-20:20 UTExtinction/Backscatter Ratio
Extinction profile derived from CARL backscatter profileand mean S
a = 59 sr
CARL Extinction Profile
Aerosol Extinction/Backscatter Ratio Sa (sr)
• measures vertical variability in aerosol extinction/backscatter (Sa)• variability due to changes in size, composition of aerosols
Atmospheric st. deviation of Sa (sr)
Pe
rce
nta
ge
of O
bs
erv
atio
ns
Nu
mb
er o
f Ob
se
rva
tion
s
48.4%
31.0%
14.1%
5.0%
1.5%
0
52
104
156
208
260
312
364
416
468
520
572
624
676
728
780
0%
10%
20%
30%
40%
50%
<= 5 5-10 10-15 15-20 > 20
• Average values show increase at top of BL, then a slight decrease with altitude• Large (>20% or >10 sr) variations in Sa vertical profile occur about 20% of time
Raman lidar measurements of aerosol extinction/backscatter ratio Sa
0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.020
30
4050
60
7080
90
100110
120Slope = 38Intercept = 16.5R = 0.72N = 19
CA
RL
Sa (355 n
m)
Angstrom Exponent (443/872 nm)
Comparison of Raman lidar Sa with Cimel aerosol measurements
• Sa increases with accumulation mode aerosol• limited to cases with small Sa vertical variability and large AOT (>0.4)
20
30
40
50
60
70
80
90
100
110
120
0.2 0.3 0.4 0.5 0.6 0.7
Slope=108.6Intercept=16.2R=0.90N=19
Volume Ratio = Volume (fine mode)/Volume (total)
Volume Ratio
CA
RL
Sa (sr)
How does Sa profile correlate with IAP measurements of • Angstrom exponent profile• single scattering albedo profile
DOE ARM In-Situ Aerosol Profiling
Sample Inlet Port view of rack
Objective: Obtain a statistically-significant data set of verticaldistribution of aerosol properties
Measurements: aerosol scatteringand absorption, plus chemicalcomposition, above a similarlyinstrumented surface site
2-3 profiles/week for 1 year
0
1000
2000
3000
4000
0 10 20 30 40scattering (1/Mm)
alt
itu
de
(m
ms
l)
0.6 0.7 0.8 0.9 1.0
single-scattering albedo
SGPMay 4,2000
John Ogren, Betsy AndrewsNOAA/CMDL
Summary of IAP Flight Hours
96
75
93
# complete profiles
-over CART site
-with SGP data
2.1 hoursAverage flight duration
215 hoursTotal flight time
104 flightsTotal flights
98 flight days/
280 day period
Total days
March 25 – December31, 2000
Flight period
John Ogren
Betsy Andrews
NOAA/CMDL
Aerosol Extinction (km-1 )
Alt
itude (
km
) <= .4
(.4,.75]
(.75,1.1]
(1.1,1.35]
(1.35,1.75]
(1.75,2.5]
(2.5,3.1]
> 3.1
0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16
95%5%
75%25%Median
Angstrom Exponent (450/550 nm)
Alt
itude (
km
) <= .4
(.4,.75]
(.75,1.1]
(1.1,1.35]
(1.35,1.75]
(1.75,2.5]
(2.5,3.1]
> 3.1
-1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 2.5
95%5%
75%25%
Median
SGP IAP measurements
wo (550 nm) (ambient)
Alt
itude (
km
) <= .4
(.4,.75]
(.75,1.1]
(1.1,1.35]
(1.35,1.75]
(1.75,2.5]
(2.5,3.1]
> 3.1
0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 1.00 1.05
95%5%75%25%Median
wo (550 nm) (dry)
Alt
itude (
km
) <= .4
(.4,.75]
(.75,1.1]
(1.1,1.35]
(1.35,1.75]
(1.75,2.5]
(2.5,3.1]
> 3.1
0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 1.00 1.05
95%5%75%25%Median
SGP IAP measurements
St. dev of Angstrom Exponent (45/550 nm)
Percent of observations
19.2%
16.7%
17.9%
11.5%
7.7%
12.8%
3.8% 3.8%
1.3%
5.1%
0%
5%
10%
15%
20%
<= .1(.1,.15]
(.15,.2](.2,.25]
(.25,.3](.3,.4]
(.4,.5](.5,.6]
(.6,.7]> .7
(~45% with st. dev > 0.2
St. deviation of Angstrom Exponent (450/550 nm)
Percent of observations
33.3%
38.5%
12.8%
3.8%6.4%
3.8%1.3%
0%
5%
10%
15%
20%
25%
30%
35%
40%
<= .1(.1,.15]
(.15,.2](.2,.25]
(.25,.3](.3,.4]
(.4,.5](.5,.6]
(.6,.7]> .7
(weighted by aerosol extinction at 550 n m
(~15% with st. dev > 0.2
St. dev of wo (550 nm)
No of observations
2.6%
24.4%
17.9%
9.0%7.7%
38.5%
0
3
6
9
12
15
18
21
24
27
30
33
<= .01 (.01,.02] (.02,.03] (.03,.04] (.04,.05] > .05
St. dev of wo (550 nm)
No of observations
9.0%
32.1%
25.6%
11.5%
6.4%
15.4%
0
2
4
6
8
10
12
14
16
18
20
22
24
26
<= .01 (.01,.02] (.02,.03] (.03,.04] (.04,.05] > .05
(weighted by aerosol extinction (550 nm )