Analysis of the Terrestrial Thermospheric Dayglow
of the N2 c’1Σu+ (0) and b’1Σu+ (1) - X 1Σg + Emission
and Implications for Cassini UVIS Titan Data Analysis
Xianming Liu and Donald Shemansky
Space Environment Technologies
Alan, N. Heays and Brenton R. Lewis
Australian National University
Paul D. Feldman
Johns Hopkins University
Status of N2 parameters required for modeling
Cassini UVIS observations (I)
• Photoabsorption cross sections
– High-resolution measurements by Stark et al. for strong bands
– Other low & medium resolution measurements (>10,700 cm-1 or <
935 Å)- saturation!
– Couple-channel Schrödinger Equation (CSE) model of Lewis’ group
• reproduces Stark et al’s measurements within 10%, eg. the b(3)-X(0)
and photodissociation by solar Lyman-γ lines.
• predicts line oscillator strengths, predissociation rates, photodissociation
cross section for both strong and weak bands for levels below 10,700
cm-1 (>935Å) over wide temperature range
• needs to be examined and refined by emission measurement
• requires additional high-resolution experimental measurements and
information for F and G 3Πu for higher levels
Status of N2 radiative parameters required for
modeling Cassini UVIS observations (II)
• Emission cross section – very limited
– measurements by Walter and Cosby on predissociation yields for a
few isolated bands with laser spectroscopy
– e+N2 measurements by Ajello et al. (1989), James et al. (1990) and
Liu et al. (2005) – errors up to 25%, and Liu et al. (2007)
• Problems for accurate modeling
– Strong variation of oscillator strength and predissociation rates with
ro-vibrational quantum numbers.
• Cross section strong temperature dependence
• Experimental emission cross section inapplicable unless
temperatures are close to each other
– Reliable absorption cross section not available until recently
– Strongly coupled system requiring systematic experimental
measurement and joint theoretical investigation
Status of N2 radiative parameters required for
modeling Cassini UVIS observations (III)
• Estimated time for accurate N2 radiative parameters (≤12%
error
– Levels below 10,700 cm-1 (>935Å)
•
•
•
•
Some available now
1-3 years for all strong or moderately strong bands
2-5 years for all VUV bands 935-1800Å
Requires additional funding from NSF and NASA
– Levels above 10,700 cm-1 (<935Å)
• Requires high-resolution photoabsorption from other groups (eg.
Collaborator Stark et al)
Potential Energy Curves of N2 Valence-Rydberg
States
Classification of N2 Singlet-ungerade states
• 2 valence states: b’1Σu+ and b 1Πu states.
• 3 Rydberg series: npσ, npπ and nsσ.
– npσ and npπ series converge to the N2+ X 2Σg +
state and are designated as c’n+11Σu+ and cn 1Πu.
– nsσ series converges to the N2+ A 2 Πu state and
is designated as on 1Πu.
Adiabatic and Diabatic Curves of Singlet-ungerade States
Stahel et al, J. Chem. Phys, 79, 2541 (1983).
FUSE-Earth vs UVIS-Titan Dayglow
• FUSE higher resolution
– 0.1-0.4 Å vs 2.5-3 Å for UVIS
– Easier separation of atomic and ionic emission features
from the N2 features
– A very important step towards accurate analysis of the
Titan dayglow spectrum
• Limited spectral coverage
– 904 to 1188 Å with 11 Å gap from 1083 to 1094 Å
• Stronger atomic and ionic oxygen emission
features in FUSE-Earth observation
Brightness (R/Å)
10
0
1047
5
1048
1049
1050
Wavelength (Å)
15
1051
1005
1052
1006
1053
0
1024
0
1072
1025
1073
1026
1074
1027
10
1075
1028
Wavelength (Å)
Wavelength (Å)
c4'(0) & b'(1) - X(4)
c4'(0) & b'(1) - X(5)
1076
Wavelength (Å)
O I (1027.431)
N I (1027.15, 1027.24)
N I (1026.69, 1026.78)
O I (1028.157)
N I (1028.357, 1028.449)
c4'(1,4)
1077
N I (1029.500, 1029.592)
c4'(3,6)
c4'(2,5)
5
c4'(3,8)
c4'(1,6), c4'(2,7)
40
H I (1025.722)
O I (1025.762)
15
CO 3p E(0)-X(0)
1004
b'(4,6)
1003
10
b'(7,5)
50
b'(7,7)
30
Brightness (R/Å)
c4'(1,3)
10
Brightness (R/Å)
1002
N I (1051.867, 1051.964)
N I (1052.050,1052.149,1052.22
1001
b'(4,5)
b'(7,4)
20
b' (7,6)
N*
0
1000
Ar I (1048.220)
b(5,4)
Brightness (R/Å)
c4'(0) & b'(1) - X(2)
c4'(0) & b'(1) - X(3)
1029
1030
15
5
1078
Brightness (R/Å)
10
5
0
1153
1154
1101
1155
1102
1103
1156
Wavelength (Å)
1157
1104
1158
1105
15
1159
Brightness (R/Å)
5
0
1181
5
1126
1182
1127
15
10
1183
1128
1184
1129
Wavelength (Å)
Wavelength (Å)
c4'(0) & b'(1) - X(8)
c4'(0) & b'(1) - X(9)
1185
Wavelength (Å)
O II (1130.147)
O II (1129.251)
P(1) (1128.507)
O II (1128.081)
10
O III (1185.961)
N I (1101.291)
15
P(1) (1185.048)
1100
0
1125
O III (1184.174)
0
1099
P(1) (1101.687)
4
O III (1183.150)
6
O III (1182.770)
8
N I (1100.360; 1000.465)
20
b'(4,10)
O III (1181.748)
c4'(3,12)
10
Brightness (R/Å)
12
P(1) (1156.28)
14
O II (1154.096)
16
N I (1098.954, 1099.152)
18
c4'(4,12)
O II (1153.357)
b'(4,9)
Brightness (R/Å)
c4'(0) & b'(1) - X(6)
c4'(0) & b'(1) - X(7)
2
1130
1186
1131
1187
Summary I. Model Results
• Inferred thermospheric temperature 500 ±50 K
• Model reproduces observed brightness of v”=2-9 within
FUSE observation error (±15% )
• Reliable estimate for v”>9 level
• Radiation loss
– 98% for v”=0 level (2115 R expected for optically thin vs 24±7 R
observed)
• Multiple scattering, predissociation, radiative escaping to v”>0
• Reduction of emission rate and distortion of band shape
– 68% for v”=1 level (315 R expected vs 100 R observed)
• By predissociative b(2)-X(0) absorption
• By self-absorption of c’(0)-X(1) itself (require significant
population at v”=1 level -> non-LTE N2(v”) population!
Analysis of b(1)-X(v”) will provide definitive answer!)
c4'(0) & b'(1) - X(0)
c4'(0) & b'(1) - X(1)
80
100
FUSE
Model (opt. thin) x 1/30
70
60
Brightness (R/Å)
50
Brightness (R/Å)
80
P(1) (958.602)
60
FUSE
Model (opt. thin) x 2/7
P(1) (980.502)
70
90
40
30
20
50
40
30
20
10
10
-0
-0
-10
957
-10
958
959
Wavelength (Å)
960
961
979
980
981
982
Wavelength (Å)
983
Summary II. Neutral N2 Excitation mechanisms in
thermospheres of Earth and Titan
• Excitation by Photoelectrons
– Principal mechanism for every singlet-ungerade levels except where
resonant solar photoexcitation is important
• Resonant photoexcitation by solar radiation
–
–
–
–
–
b(3) by H Ly-γ (complete predissociation)
b(6) by H Ly-δ, b(10) by H Ly (n=9)
b’(4) by H Ly-ε, b’(6) by H Ly (n=8), b’(7) by H Ly (n=12)
b’(8), o(2), and b(12) by H Ly (n>15)
b(12) also by H ionization continuum
• Resonant photoexcitation by strong N2 emission
– b(2) from v”=0 by c’1Σu+ (0)-X 1Σg +(1) transition
– b’(4) from v”=1 by c’1Σu+ (0)-X 1Σg +(0) transition [non-LTE
X(v”)]
Summary III. Prominent N2 dayglow emission features
from the FUSE and implication for UVIS-Titan modeling
•
c’1Σu+ - X 1Σg + band system (v’=0, 1, 3, 4, 5, and 6)
– v’ = 0 can be accurately modeled
– v’ = 1-4 have good oscillator strengths but unreliable predissociation rates
– v’ = 5 & 6 require both oscillator strengths and predissociation rate
•
b’1Σu+ - X 1Σg + band system (v’= 1, 4, 6, 7, 9, 12, and 14)
– v’ = 1 can be accurately modeled
– Others require accurate predissociation rates
•
b 1Πu - X 1Σg + band system (v’ = 1, 6, 7, and 10)
– Most levels require accurate predissoiciation rates
•
c 1Πu - X 1Σg + band system (v’ = 0 and 1) (relatively weak)
– Requires accurate diabatic transition moment
Example of strong temperature dependence and the
importance of accurate predissociation rates at rotational
levels
I.
II.
III.
IV.
V.
Fig 1. Comparison of 150 K laboratory measurement and model
spectrum using CSE calculation
Fig 2. Comparison of 150 K lab measurement and adjusted CSE
model. The adjustment factors are listed at top right corner
Fig 3. Comparison of 300 K lab measurement with model using
adjustment factors obtained at 150K
Fig 4. Comparison of 300 K lab spectrum with model using a new
set of adjustment factors
Fig 5. Comparison of FUSE spectrum (~500K) with model using
adjustment factors obtained at 300 K
e+N2 (150K, 100 eV)
Model based on CSE calculation with no adjustment
2.5
Calibrated Intensities (arb unit)
100eV, 4.6E-4 Torr
model (no self-abs)
2.0
1.5
0.9
0.4
-0.1
1000.0
1001.0
1002.0
1003.0
1004.0
1005.0
1006.0
Wavelength(A)
1007.0
1008.0
1009.0
1010.0
e+N2 (150K, 100 eV)
Model based on CSE cal. with adjusted transition moments
2.2
2.0
c'(1) x 0.35; c'(3) x 0.70; c'(4) x 0.47; b(6) x 2.5;
100eV, 4.6E-4 Torr
model (no self-abs)
b(7) x 2.5; b'(7)x 2.5; b'(9) x 0.8; b'(12)x0.75, b'(14)x0.3
Calibrated Intensities (arb unit)
1.8
1.6
1.4
1.2
0.9
0.7
0.5
0.3
0.1
-0.1
1000.0
1001.0
1002.0
1003.0
1004.0
1005.0
1006.0
Wavelength(A)
1007.0
1008.0
1009.0
1010.0
e+N2 (100 eV, 300K, Δλ=0.3A)
Mar. 29, 2007 3:47:09 PM
1500
Aug-08-2006
MOD_300K
Aug-21-2006
c'(1) x 0.35; c'(3) x 0.70; c'(4) x 0.47; b(6) x 2.5;
Calibrated Intensities (counts)
1300
b(7) x 2.5; b'(7)x 2.5; b'(9) x 0.8; b'(12)x0.75, b'(14)x0.3
1100
900
700
500
300
100
-100
1000.0
1001.0
1002.0
1003.0
1004.0
1005.0
1006.0
Wavelength(A)
1007.0
1008.0
1009.0
1010.0
e+N2 (100 eV, 300K, Δλ=0.3A)
Mar. 29, 2007 6:49:33 PM
1500
1300
Aug-08-2006
MOD_300K
Aug-21-2006
c'(1) x 0.45; c'(4) x 0.65; b(1)x1.5; b(6) x 3;
Calibrated Intensities (counts)
b(7) x 2.5; b'(7)x 2.5; b'(12)x0.75, b'(14)x0.3
1100
900
700
500
300
100
-100
1000.0
1001.0
1002.0
1003.0
1004.0
1005.0
1006.0
Wavelength(A)
1007.0
1008.0
1009.0
1010.0
FUSE Thermospheric N2 Dayglow Emission Spectrum
Based on scaling factors for 300 K May 15, 2007 12:05:26 PM
70
c'(1)x0.45; c'(4)x 0.65;b'(2)x3; b(1)x1.5; b(6)x3;
60
FUSE_COMPOSITE
Model (100eV, 500K)
b(7)x2.5; b'(7)x2.5; b'(12)x0.75, b'(14)x0.3
Brightness (R/Å)
50
40
30
20
10
0
1000.0
1001.0
1002.0
1003.0
1004.0
1005.0
1006.0
Wavelength (Å)
1007.0
1008.0
1009.0
1010.0
Conclusion
• Laboratory emission cross section is not directly
applicable unless the measurement and
thermospheric temperatures are close to each other
• Accurate modeling requires reliable
predissociation rates at rotational level.
• CSE predissociation rates need to be examined
and refined by high-resolution lab measurement at
300 K.