Journal of Quantitative Spectroscopy & Radiative Transfer 111 (2010) 2139–2150
Contents lists available at ScienceDirect
Journal of Quantitative Spectroscopy &
Radiative Transfer
journal homepage: www.elsevier.com/locate/jqsrt
HITEMP, the high-temperature molecular spectroscopic database
L.S. Rothman a,, I.E. Gordon a, R.J. Barber b, H. Dothe c, R.R. Gamache d,
A. Goldman e, V.I. Perevalov f, S.A. Tashkun f, J. Tennyson b
a
Harvard-Smithsonian Center for Astrophysics, Atomic and Molecular Physics Division, Cambridge MA 02138, USA
Department of Physics and Astronomy, University College London, London WC1E 6BT, UK
Spectral Sciences, Inc., Burlington MA 01803, USA
d
University of Massachusetts School of Marine Sciences, Department of Environmental, Earth, and Atmospheric Sciences, Lowell MA 01854, USA
e
University of Denver, Department of Physics, Denver CO 80208, USA
f
Institute of Atmospheric Optics, Siberian Branch, Russian Academy of Sciences, Tomsk 634055, Russia
b
c
a r t i c l e in fo
abstract
Keywords:
Spectroscopic database
Molecular spectroscopy
Molecular absorption
Line parameters
High-temperature spectroscopy
HITEMP
A new molecular spectroscopic database for high-temperature modeling of the spectra
of molecules in the gas phase is described. This database, called HITEMP, is analogous to
the HITRAN database but encompasses many more bands and transitions than HITRAN
for the absorbers H2O, CO2, CO, NO, and OH. HITEMP provides users with a powerful tool
for a great many applications: astrophysics, planetary and stellar atmospheres,
industrial processes, surveillance, non-local thermodynamic equilibrium problems,
and investigating molecular interactions, to name a few. The sources and implementation of the spectroscopic parameters incorporated into HITEMP are discussed.
& 2010 Elsevier Ltd. All rights reserved.
1. Introduction
In the second half of the twentieth century, the
simultaneous development of computers, high-resolution
laboratory spectroscopy, and sensitive detectors for field
instruments led to the establishment of a computerreadable archive of spectroscopic parameters applicable
for atmospheric transmission and radiance calculations
[1]. The standard database for this work is the HITRAN
database [2], which is periodically updated and expanded.
HITRAN has its origins in applications for conditions of the
terrestrial atmosphere, particularly temperatures ranging
from the surface of the Earth to the stratosphere. As a
result, there are many molecular bands and line transitions that would be significant at high temperatures that
have not necessarily been considered for the HITRAN
archive. In addition, there are numerous molecular bands,
Corresponding author. Tel.: + 1 617 495 7474; fax: +1 617 496 7519.
E-mail address: LRothman@cfa.harvard.edu (L.S. Rothman).
0022-4073/$ - see front matter & 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jqsrt.2010.05.001
especially in the near-IR and visible spectral regions, that
are still missing from HITRAN due to the lack of either
experimental data or theoretical calculations of these
lines.
The HITRAN database, used as input to various highresolution transmission codes, has been successful for a
vast number of applications. The foremost application is
remote-sensing of the terrestrial atmosphere from spectrometers aboard satellites, balloons, and ground-based
instrumentation. There are also environmental, industrial,
surveillance problems, and numerous other applications.
Industrial, environmental, and surveillance applications
often require a high-temperature spectroscopic database.
However, the use of HITRAN (established at a reference
temperature of 296 K) is usually deficient when applied to
problems where gases are at elevated temperatures. There
is also the obvious requirement in astrophysics to
characterize stellar, brown dwarfs, and planetary atmospheres. For example, the recent detection [3] of water in
extrasolar planet HD189733b relied heavily on the BT2
line list [4], which is used in the present work as
2140
L.S. Rothman et al. / Journal of Quantitative Spectroscopy & Radiative Transfer 111 (2010) 2139–2150
explained below. The analysis would not have been
possible with HITRAN nor reliable with the earlier version
of the high-temperature analog [5] to HITRAN. The need
for a high-temperature database embraces, for example,
some planetary atmospheres such as possessed by Venus
or many exosolar planets. An example of the inability of
HITRAN to adequately characterize the Venus night-side
atmosphere was given in Pollack et al. [6]. In addition,
having a database with excited energy levels satisfies the
requirements of some non-local thermodynamic equilibrium (NLTE) problems in the atmosphere. A typical
example is the Meinel bands [7] of OH, which require
transitions between very high energy levels that would
not be necessary in normal radiative-transfer applications.
Early attempts to produce a high-temperature database simply scaled the HITRAN database to estimate the
absorption and emission spectra at elevated temperatures. In the database, the intensity1 of a line transition Sjf
between lower state i and upper state f as a function of
temperature T is given by
Sif ðTÞ ¼ Sif ðTref Þ
Q ðTref Þ expðc2 Ei =TÞ 1expðc2 nif =TÞ
,
Q ðTÞ expðc2 Ei =Tref Þ 1expðc2 nif =Tref Þ
ð1Þ
where Tref is the reference temperature of the database
(296 K), Q is the total partition sum, Ei is the energy of the
lower state (cm 1), and vif is the energy difference
between the initial and final state (given as vacuum
wavenumber, cm 1, in the database). The constant c2 is
the second radiation constant (c2 = hc/k= 1.43877 cm K).
The quantities in Eq. (1) are all provided in the HITRAN
compilation. Although the scaling from room temperature
to higher temperatures is sufficiently accurate to estimate
the line intensity at higher temperature provided an
adequate high-temperature partition function is available,
the additional spectral lines that are missing from the
simulation often lead to a significant underestimation of
the source radiance in specific spectral regions.
To address these issues, an analogous database to
HITRAN was established and called HITEMP [5] (hereafter
called HITEMP1995 to distinguish it from the current
effort). This first edition included only the gases H2O, CO2,
CO, and OH. The water-vapor line parameters were the
result of a calculation using the direct numerical diagonalization (DND) method [8] and were aimed at being
sufficient for 1000 K. There was also a calculation for
1500 K, but with a limited dynamic range of intensities.
The carbon dioxide parameters were also the result of the
same theoretical methodology [8], applicable at 1000 K.
The carbon monoxide line list was adapted from the work
of Goorvitch [9] that was constructed for a solar atlas.
Finally, the hydroxyl line list [10] was added to
1
The units adopted for the parameters in the molecular spectroscopic databases are the cgs system. The units of intensity are given in
wavenumber per column density, cm 1/(molecule cm 2), which can be
simplified to cm molecule 1 although this form obscures the origin.
Note also that we employ the symbol n throughout for line position in
cm 1, thereby dropping the tilde (n~ ) that is the official designation of
wavenumber.
HITEMP1995 since an extensive list was available in
HITRAN itself for NLTE applications.
One desirable feature in creating the HITEMP database
(hereafter called HITEMP2010 when referring to the new
edition) is to have it consistent with the HITRAN database.
That is, it is preferable to have any transitions in common
be identical since some simulations might use HITEMP for
the source with HITRAN representing the intervening
atmospheric path. Having this correlation of lines was
simple for CO and OH, where the HITRAN and HITEMP line
lists for these gases were generated from the common
sources. For CO2, constraints were applied that heavily
weighted the molecular constants in the fit to the HITRAN
values. However, for H2O the problem was much more
complicated since the data in HITRAN consist of many
different contributions, both experimental and calculated,
and from different sources. Thus common lines found in
HITEMP1995 were superseded by their counterpart in
HITRAN (the edition of the HITRAN database [11] at that
time). This method relies on the quantum identifications
of transitions in both databases being consistent, by no
means assured, especially for the higher polyads. This
method can also introduce discontinuities in the line
positions of bands as one migrates to higher rotational
values.
With these limitations of the older version, HITEMP1995, in mind, and with new calculations and
experiments that have become available, we embarked
on a program [12] to substantially update the database, as
described in the following sections.
2. Structure of the database
The format of HITEMP2010 has been maintained to be
the same as that of HITRAN. Thus, the HITEMP1995
edition [5] had the 100-character length transition record
that was established in 1986 [13]. The current HITRAN
edition has increased the length of each transition to 160
characters. Table 1 gives a description of the current list of
parameters, definition of units, and format. The databases
are in ASCII files and, while following the HITRAN format
makes the HITEMP2010 database quite large, it was
chosen to make it easily compatible to the same
programs that make use of HITRAN.
Similarly, the intensity is given at the HITRAN standard
temperature of 296 K, even though calculations have been
performed at much higher temperatures. Usage of the
database requires converting by Eq. (1) back to the
temperature of the application, which is done routinely
in transmission or radiance codes. The standardization to
296 K means that some of the intensities in the database
may have very low exponents and require double
precision when being used.
One of the most difficult entities to define for the
archival databases is the criterion for the cutoff in
intensities. This lower limit for intensities had some
physical meaning for the uniformly mixed gases in
HITRAN, i.e., to include all lines that contributed at least
a 10% absorption over a maximal path in the terrestrial
atmosphere, namely space to space tangent to the surface.
L.S. Rothman et al. / Journal of Quantitative Spectroscopy & Radiative Transfer 111 (2010) 2139–2150
2141
Table 1
Description of the parameters and format of the current HITRAN and HITEMP databases.
Symbol
Parameter
Field length
Data type
Comments or units
M
I
molecule number
isotopologue number
Vacuum wavenumber
Intensity
Einstein A-coefficient
Air-broadened halfwidth
Self-broadened halfwidth
Lower-state energy
Temperature-dependence coefficient
Air-pressure-induced line shift
Upper-state ‘‘global’’ quanta
Lower-state ‘‘global’’ quanta
Upper-state ‘‘local’’ quanta
Lower-state ‘‘local’’ quanta
Uncertainty indices
Reference indices
Flag
Statistical weight of the upper state
Statistical weight of the lower state
2
1
12
10
10
5
5
10
4
8
15
15
15
15
6
12
1
7
7
Integer
Integer
Real
Real
Real
Real
Real
Real
Real
Real
Text
Text
Text
Text
Integer
Integer
Text
Real
Real
HITRAN chronological assignment
Ordering by terrestrial abundance
cm 1
cm 1/(molecule cm 2) at standard 296 K
s1
HWHM at 296 K (in cm 1 atm 1)
HWHM at 296 K (in cm 1 atm 1)
cm 1
Temperature-dependent exponent for gair
cm 1 atm 1 at 296 K
see Table 3 of Ref. [14]
see Table 3 of Ref. [14]
see Table 4 of Ref. [14]
see Table 4 of Ref. [14]
Uncertainty indices for 6 critical parameters (n, S, gair, gself, n, d)
Reference pointers for 6 critical parameters (n, S, gair, gself, n, d)
Pointer to program and data for the case of line mixing
See details in Ref. [15]
See details in Ref. [15]
n
S
A
gair
gself
E00
n
d
V0
V00
q0
Q00
Ierr
Iref
n
g0
g00
Table 2
High-temperature water-vapor line lists.
Line list
Spectral range (cm 1) Number of transitions
Potential-energy surface Dipole-moment surface
HITEMP1995 Ref. [5] Fitted Ref. [8]
AMES Ref. [16]
Fitted Ref. [16]
SCAN Ref. [18]
Fitted Ref. [19]
Fitted Ref. [21]
BT2 [4]
Fitted Ref. [8]
ab initio Ref. [17]
ab initio Ref. [20]
ab initio adjusted from Ref. [17]
The ability to accurately calculate or measure weak lines
has improved over time, and this criterion has been
relaxed in HITRAN to include weaker lines. However,
numerous groups have applied different constraints on
the construction of line lists. With calculations it is often
simply a maximum rotational quantum number that is
chosen, since at higher values the energy expressions may
be too expensive to calculate using variational methods or
they may be divergent using perturbation theory. The
earlier edition, HITEMP1995, employed a rationale of
scaling the cutoff in intensities by the Boltzmann
distribution at the different temperatures of 1000 and
296 K for water vapor and carbon dioxide, respectively.
Other groups have limited their line lists based on
experimental limitations such as detector response,
optical properties of absorption cells, etc. The cutoff
criteria of the intensities for the molecules in
HITEMP2010 are discussed in the following section.
3. Components of HITEMP
3.1. H2O
For the HITEMP1995 water vapor line list a cutoff in
intensities of 3.7 10 27 cm molecule 1 at 1000 K was
employed. This cutoff has significantly reduced the
0–24,900
0–25,000
450–30,000
0–30,000
1,283,466
3.08 108
3 109 ( 1 108 in reduced format)
5.06 108
capabilities of HITEMP1995 to accurately predict spectra
above 1000 K, although it has been fairly accurate for
temperatures below and around 1000 K. Taking into
account that the potential-energy surface (PES) and dipole
moment surface (DMS) used for the generation of
HITEMP1995 are now considered outdated, there is an
obvious need of creating a new HITEMP database.
The purpose of this work is to provide a compilation of
the best available experimental and theoretical data for
hot water vapor that would provide accurate simulations
of spectra for temperatures up to 4000 K. A number of
new theoretical line lists have emerged over the last
decade. The most extensive theoretical datasets, along
with sources for the PES and DMS used for their
generation, spectral ranges, and number of lines are given
in Table 2.
The BT2 line list has been extensively used, and in a
number of cases has been subjected to comparative
studies, with other available line lists. In particular
Campargue et al. [22] found that it accurately predicted
weak water spectra in the 770 nm transparency window;
Bailey [23] strongly recommended it for use in models of
the Venus atmosphere, and Kranendonk et al. [24] found
that its use allowed them to perform thermometry in
flames up to 2000 K with greatly improved reliability. All
these works give comparisons with model spectra
generated using BT2. We have compared the BT2 dataset
2142
L.S. Rothman et al. / Journal of Quantitative Spectroscopy & Radiative Transfer 111 (2010) 2139–2150
with experimental data from Ref. [25], one of the few
available laboratory emission spectra of very hot water for
which absolute intensities are available.
Fig. 1 shows experimental and synthetic spectra in
1.3–2.3 mm range. The experiment was carried out at
2900 K, and we compared it with the 3000 K synthetic
spectra generated using line lists from Table 2. The
emission measurements [25] were made with an
echellete grating monochromator using InSb or HgCdTe
0.030
Experiment [25]
0.025
BT2 [4]
HITEMP1995 [5]
Emission
0.020
SCAN [18]
AMES [16]
0.015
0.010
0.005
0.000
1.4
1.6
1.8
2.0
Wavelength (µm)
2.2
Fig. 1. Comparison of synthetic emission generated using different
theoretical water-vapor line lists with high-temperature observations
[25]. The reference temperature is 3000 K and the intensity cutoff is
10 27 cm molecule 1.
detectors. The average spectral slit widths were 60 and
180 Å for wavelengths l o5 and l 45 mm, respectively.
All calculations were performed using a rectangular
apparatus function with 2 Å ( = 2 cm 1) width. It should
be stressed that all experimental points shown in Fig. 1
were obtained by digitizing Fig. 2 of Ref. [25]; thus, the
accuracy of experimental points is rather low.
A cutoff of 10 27 cm molecule 1 was employed for all
of the lists in the process of generating this figure. It is
clear that the list created by Barber et al. [4], hereafter
referred to as BT2, provided the best match with the
experimental spectrum. It was chosen to be the starting
point for constructing the new HITEMP2010.
The BT2 dataset, along with accompanying software, is
capable of generating over 505 million transitions of H16
2 O
in the 0–30,000 cm 1 region with oscillator strengths
down to 10 36 Debye2. The only restriction applied in
creating the BT2 list was that only rotational levels with
Jr50 were considered. As discussed in the original BT2
paper [4], even at a temperature of 4000 K, the missing
levels above J= 50 contribute less than 0.02% to the total
partition function. Even with this limitation, BT2 cast into
the HITRAN2004 format would yield a file as large as
90 Gigabytes, which is not practical for the general use of
the database. Therefore it was decided to reduce the
number of transitions, but in a way that the database
would still be suitable for the majority of high-temperature applications including the most demanding such as
non-local thermodynamic equilibrium (NLTE) calculations. A particularly good example is the work on
emission spectra from comets where BT2 assigned highly
NLTE transitions that had been previously ignored and
Fig. 2. Flow diagram of water-vapor line list assembly.
L.S. Rothman et al. / Journal of Quantitative Spectroscopy & Radiative Transfer 111 (2010) 2139–2150
which undoubtedly contain new physical insight into the
system [26].
The intensity cutoff taken at a certain temperature (as
was done in HITEMP1995) is a very coarse approach since
the intensity distribution changes with temperature,
while it is desired that the database will be suitable over
the great range of temperatures. Other approaches for
introducing intensity cutoffs to the water-vapor lines
exist. For instance, Sharp and Burrows [27] suggested that
for calculations of opacities for brown dwarfs and
exoplanet atmospheres, lines with SZ10 40 cm
molecule 1 at 296 K be considered. Also recently, Perez
et al. [28] undertook a similar task for data reduction
when they were reducing the amount of lines in the AMES
database [16] for plume signature computations. The
sophisticated line rejection procedure in that work was
performed with a criterion of maximum allowable error
on the radiance (DI) of the heterogeneous path. DI is
proportional to the maximum error, dak, in the absorption
coefficient in each layer k, and the rejection was carried
out in a way that the absorption coefficient due to
rejected lines in the layer k is lower than dak. By applying
this criterion, which was also temperature dependent,
Perez et al. [28] have omitted 95% of the lines from the
AMES database in the 600–6000 cm 1 region, while still
being able to perform their calculations at the desired
level of accuracy. However, for this HITEMP database, we
have decided to retain a much higher percentage of
transitions from the BT2 database since some of the
HITEMP users are working on even more extreme
applications than plume signatures, such as radiative
transfer in stellar atmospheres, for example.
At first, the complete line list at 296 K was generated
using BT2 data and the software that accompanies it. The
partition function used was that from HITRAN,
Q(296)= 174.64. Then a filtering procedure was applied
using the following criteria. The intensity S(T) was
evaluated at temperatures from 296 to 4000 K using
Eq. (1) in 500 K intervals. At each of the probed
temperatures, S(T) is compared with the intensity cutoff
Scut(T), defined in three wavenumber domains.
c n
S n
2
0 o n r 5000cm1 ,
ð2aÞ
Scut ðTÞ ¼ crit tanh
2T
ncrit
Scut ðTÞ ¼ Scrit
5000 o n r7000 cm1 ,
1
Scut ðTÞ ¼ 1 1027 cm molecule
27
n 47000 cm1 ,
1
ð2bÞ
ð2cÞ
where Scrit = 3 10
cm molecule , ncrit = 2000 cm 1.
Eq. (2a) was chosen to account for the radiation field at
very low wavenumber. The different criteria for the other
wavenumber ranges, Eqs. (2b) and (2c), were simply
introduced to permit more lines in the near-IR and visible,
which are of course weak, to be retained.
Only those lines with S(T) 4Scut(T) at any temperature
between 296 and 4000 K were kept for inclusion into
HITEMP2010. On average about 25% of the lines that can
be generated using BT2 end up being retained using this
rejection procedure. The resultant file is then converted to
HITRAN2004 format [14] and serves as a base for
assembling the new HITEMP2010 compilation.
2143
The next step is achieving consistency with the latest
HITRAN water-vapor data, which is the file 01_hit09.par.
For that purpose, whenever there is a transition in the
filtered BT2 file described above that is also available in
HITRAN, it will be removed and replaced with the one in
HITRAN with two important exceptions. The first exception is associated with a problem of correspondence of
quantum assignments of energy levels between the
theoretical BT2 list and the experimental lists. This is
due to the fact that only the rotational quantum number J
and symmetry can be unambiguously identified using the
theoretical potential-energy surface. Different experimental works that contributed to HITRAN have used different
potential-energy surfaces to aid in the assignments of the
observed transitions, which can occasionally result in
ambiguities of quantum assignments. Therefore it was
decided to replace a BT2 line with one from HITRAN only
if the difference between the line positions in the two
datasets is less than 1.5 cm 1 when n r5000 cm 1,
2 cm 1 when 5000o n r15,000, and 2.5 cm 1 when
n 415,000 cm 1. A second exception is when HITRAN
intensities have the reference code 19, which corresponds
to experimental intensities of two blended lines combined
into one. In this case only the line position is taken from
HITRAN, while the BT2 intensity is retained. Since the BT2
list provides data only for the principal isotopologue, the
lines for other isotopologues have been taken from
HITRAN for the current effort. Future updates will
consider high-temperature data for the isotopologues,
such as is available in the recent VTT line list for HDO [29].
Finally, we have created a line list (hereafter referred to
as the list of semi-empirical line positions or SELP) that
contains the transition wavenumbers calculated using the
database of experimental energy levels [30] as updated by
Zobov using recent high-temperature experiments such as
Refs. [31–33]. When possible, the line positions originating
from the BT2 database have been replaced with the ones
from the SELP list, with the same restrictions that were
applied for the introduction of HITRAN line positions. In
the future the HITEMP database will benefit from the
ongoing International Union of Pure and Applied Chemistry (IUPAC) effort, which includes validating and putting
together all available experimentally determined energy
levels. This effort has already proved efficient in the case of
17
the H18
2 O and H2 O isotopologues [34].
The flow diagram in Fig. 2 summarizes the procedure
of creating the new HITEMP2010 water-vapor line list.
The addition of the collision-induced parameters,
namely the air-broadened half-width gair and its temperature-dependence exponent n, the pressure-induced
line shift dair, and the self-broadened half-width gself, was
accomplished using the slightly extended [2] procedure of
Gordon et al. [35], here improved to consider the high
rotation–vibration levels now in the HITEMP database.
Some problems encountered in this new line list are that
the vibrational and rotational quantum numbers are often
not fully known, or only partially known. Only a small
percentage of transitions taken from the BT2 database has
a full description of the quantum numbers (in normal
0
mode representation), i.e. v10 , v20 , v30 , v100 , v200 , v300 , J0 , Ka,
0 00
00
00
Kc , J , Ka , and Kc . Here, v1, v2, and v3 are the quanta of
2144
L.S. Rothman et al. / Journal of Quantitative Spectroscopy & Radiative Transfer 111 (2010) 2139–2150
vibration for the symmetric stretch, the bending mode,
and the antisymmetric stretch, respectively. Single primes
denote the upper state, while double primes the lower
state. The total angular momentum quanta are given by J,
and Ka and Kc are the standard quanta of projection for an
asymmetric rotor molecule. Some transitions have either
only the rotational quantum numbers, or only J0 and
symmetry for the upper level and J00 , Ka00 , and Kc00 for the
lower level, or finally J and symmetry for the upper and
lower levels.
Thus the algorithm of Gordon et al. [35] must be
modified to work correctly for all situations. For Jr25 and
the complete set of quantum assignments, the previous
extended algorithm [35] works fine. When the vibrational
quantum numbers are not given, the change in vibrational
quanta can be estimated by the wavenumber of the
transition. For cases with full rotational quantum
numbers, approximate values for gair and dair can be
obtained using the work of Jacquemart et al. [36], which
is incorporated into the extended algorithm. Likewise
when only the lower state J00 , Ka00 , Kc00 or only the J00 and
symmetry are given, the work of Ref. [36] can be used.
For transitions with J425 or missing quantum
information, we used half-widths averaged for each J00 as
discussed below. This option currently exists in the
extended algorithm for Jr25. Because of the nature of
the HITEMP database, the method must be extended to
J= 50. We have been able to accomplish this by noting that
for water vapor as one goes up in J and Ka the collision
process responsible for the broadening goes more and
more off-resonance and the half-width goes to a limiting
value. For air broadening, both the measurement database
[37] and the theoretical calculations for the pure rotation
band [38] were investigated. From this analysis, a limiting
value of gair (J00 = 50)=0.00839 cm 1 atm 1 was determined. The rotation band data were then taken and
averaged as a function of J00 . Because the data at J00 equal to
19 and 20 do not have a complete sampling of transitions,
these averages were adjusted to 0.0270 and 0.0240 cm 1
atm 1, respectively. The points at J00 equal to 49 and 50
were set to the limiting value. The data were then fit to a
third order polynomial in J00 . The data, averaged as a
function of J00 , and the polynomial fit are presented in the
top panel of Fig. 3. A similar figure was made with the
measurement database results, but was not used due to
scatter in the measured values.
The above procedure was repeated for self-broadening
of water vapor. The measurement database and the
calculations made for the 3.2–17.76 mm region [39] were
studied giving the self-broadened limiting value of gself
(J00 = 50)= 0.0400 cm 1 atm 1. The averages were calculated using the calculated data for the 3.2–17.76 mm
region and the value at J00 = 18 was adjusted to
0.1800 cm 1 atm 1. The points at J00 equal to 49 and 50
were set to the limiting value and these data fit to a third
order polynomial in J00 . In the lower panel of Fig. 3 the
data, the average as a function of J, and the polynomial fit
are presented. Again, a similar figure using the measurement database was made, which displayed even more
scatter than the air-broadened half-width measurements,
so it was not used for the fit.
Fig. 3. Extrapolation of average values of (a) air- [38] and (b) selfbroadened [39] calculated half-width values as a function of J00 .
Table 3
H2O air- and self-broadened half-widths in cm 1 atm 1 determined
from the polynomial as a function of J00 .
J00
gair
gself
J00
gair
gself
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
0.10691
0.10035
0.09409
0.08812
0.08244
0.07703
0.07190
0.06703
0.06242
0.05806
0.05395
0.05007
0.04642
0.04299
0.03978
0.03677
0.03397
0.03136
0.02894
0.02670
0.02463
0.02273
0.02098
0.01939
0.01795
0.01664
0.50361
0.47957
0.45633
0.43388
0.41221
0.39129
0.37113
0.35170
0.33300
0.31501
0.29773
0.28113
0.26521
0.24996
0.23536
0.22140
0.20806
0.19534
0.18323
0.17170
0.16076
0.15038
0.14056
0.13128
0.12252
0.11429
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
0.01546
0.01441
0.01347
0.01265
0.01193
0.01130
0.01076
0.01031
0.00992
0.00961
0.00936
0.00916
0.00901
0.00890
0.00882
0.00876
0.00873
0.00870
0.00868
0.00866
0.00863
0.00858
0.00851
0.00841
0.00827
0.10656
0.09933
0.09257
0.08629
0.08046
0.07507
0.07012
0.06559
0.06146
0.05773
0.05438
0.05140
0.04878
0.04651
0.04457
0.04295
0.04165
0.04064
0.03991
0.03946
0.03927
0.03932
0.03961
0.04013
0.04085
L.S. Rothman et al. / Journal of Quantitative Spectroscopy & Radiative Transfer 111 (2010) 2139–2150
The final half-widths determined from the polynomial
fits of the J00 -averaged data are given in Table 3.
The user has to be aware that at high temperatures
some of the parameters may be found to be inaccurate.
For example, the exponential temperature dependence
model used for the air-broadened half-widths given in
HITRAN has been shown to be inadequate for many
water-vapor transitions and for large temperature ranges
[38,40,41].
A sample extract of 41 HITEMP2010 water lines is
shown in Fig. 4. The definition of the parameter columns
can be ascertained from Table 1. The line in bold font is a
transition for which the line position was taken from SELP.
The Transition in bold italic is taken from HITRAN2008. All
the other lines have wavenumbers and intensities taken
from BT2. Whenever a rotational quantum number could
not be determined unambiguously, the index of symmetry
(1–4 as defined in Ref. [4]) accompanied with a negative
sign was used. Note that 1 and 2 indicate para states,
whereas 3 and 4 indicate ortho states. For the case of
unassigned vibrational quanta, a ‘‘ 2’’ label was used.
It is worth noting that if one uses the JavaHAWKS
software accompanying the HITRAN compilation to generate
the list corresponding to the desired temperature, the
partition sums from HITRAN will be automatically employed.
These partition sums are available only up to 3000 K, and if
one wants to perform calculations for water vapor at higher
temperatures, the partitions sums from Vidler and Tennyson
[42] that are available up to 6000 K are recommended.
3.2. CO2
The parameters for carbon dioxide in the HITEMP1995
database were based on the direct numerical diagonalization
2145
(DND) [8] calculation. For the line positions, a power series
expansion in the product of rotational quantum number,
J(J+1), was employed. In the case of well-observed bands, the
experimental line positions were heavily weighted and
artificial high-J lines were added based on the DND
calculation, but with very low weight. In this manner, it
was expected to avoid the usual divergence while at the
same time maintaining a consistency with HITRAN for
common lines. The DND calculations relied ultimately on
the experimental dataset being used for the development of
the PES and DMS. Comparisons with new observations and
with the more recent carbon dioxide spectroscopic database
(CDSD-296) [43] have determined that the CDSD database is
a clear improvement over the CO2 data in HITEMP1995.
In the intervening years since the first edition of
HITEMP, a high-temperature version of the Carbon
Dioxide Spectroscopic Database was developed for
1000 K [44]. This databank included more than 3.6 106
lines of four major isotopologues, 12CO2, 13CO2, 16O12C18O,
and 16O12C17O and covered the spectral range from 263 to
9648 cm 1. For CDSD, a cutoff in intensities of 1.0 10 27
cm molecule 1 at 1000 K was employed. Terrestrial
isotopic abundances were used. The databank was
extensively tested against measured medium-resolution,
high-temperature spectra of pure CO2 [45] for the 15, 4.3,
2.7, and 2.0 mm bands at temperatures of 1000, 1300, and
1550 K. Good agreement between measured and calculated medium-resolution spectra was found. CDSD was
used as a source databank to create high-accuracy,
compact databases of narrow-band k-distributions for
pure CO2 [46] and for mixtures of water vapor and carbon
dioxide [47]. The databank was also used to verify a
multiscale Malkmus model for treatment of inhomogeneous gas paths [48].
Fig. 4. A sample extract from the new HITEMP water file. The parameters and columns are defined in Table 1. The transition in bold font is the line for
which the frequency is taken from SELP. The transition in bold italic font is taken from HITRAN2008.
2146
L.S. Rothman et al. / Journal of Quantitative Spectroscopy & Radiative Transfer 111 (2010) 2139–2150
Since publication of the paper by Tashkun et al. [44], a
number of new measurements of CO2 line positions and
intensities have been published. In 2008 an updated
version of the CDSD databank called CDSD-1000 [49] was
created. This version is considered as an improved and
extended list with respect to the 2003 version. The
Fig. 5. Graphical overview of the CDSD-1000 databank [49]. All spectra are calculated at Tref = 1000 K. The X axis is in wavenumber (cm 1) and the Y axis
in intensity in HITRAN units (cm molecule 1).
L.S. Rothman et al. / Journal of Quantitative Spectroscopy & Radiative Transfer 111 (2010) 2139–2150
principal improvement is inclusion of 13CO2 data in the
range 3080–8100 cm 1. In addition, data of three less
abundant isotopologues, 16O13C18O, 16O13C17O, and
12 18
C O2, were added. This version follows the HITRAN2004 [14] format and is freely available via the
Internet site ftp.iao.ru/pub/CDSD-2008/1000. A new version of CDSD, having Tref = 4000 K and an intensity cutoff of
10 27 cm molecule 1, is under development.
The HITEMP2010 CO2 parameters were calculated
from CDSD-1000 parameters by rescaling intensities from
T= 1000 to T= 296 K. The TIPS-2003 code supplied with the
HITRAN compilation was used to calculate total internal
partition sums. A graphical overview of the HITEMP2010
CO2 is presented in Fig. 5.
In order to illustrate the quality of CDSD-1000, we
performed simulations of a high-temperature emission
spectrum given in Fig. 3 of Ludwig et al. [50]. The
experimental conditions were T= 1950 K, pressure= 1 atm,
geometrical path= 3.12 cm, and slit width= 5 cm 1. The
experimental curve of Ref. [50] was digitized. Results of
the simulation using HITEMP1995 and CDSD-1000 are
given in Fig. 6. It is seen that in the 500–700 cm 1
spectral interval both databanks give practically the same
emissivity values. In the 700–850 cm 1 interval the
CDSD-1000 simulation is closer to the observed curve
than the HITEMP1995 one.
Fig. 6. Comparison of synthetic emission generated using CDSD-1000
[49] (dashed line) and HITEMP1995 [5] (dotted line) with hightemperature observations (solid line) given in Fig. 3 of Ref. [50]. The
temperature is 1950 K, the total pressure is 1 atm, the geometrical path
is 3.12 cm, and the slit width is 5 cm 1.
3.3. CO
The line list for carbon monoxide that has been
available in HITEMP1995 was retained for this new
edition of HITEMP2010. The data were assembled from
the solar atlas of Goorvitch [9]. The only exception is
where there have been updates to the line parameters in
HITRAN for carbon monoxide after 1995; in these cases
the improved values were adopted for the HITEMP2010
compilation.
2147
3.4. NO
The nitric oxide X2P’X2P line parameters included in
HITEMP2010 are described in Goldman et al. [51,52].
These consist of the 14N16O transitions for Dn = 0, y, 5
with n0 =0, y, 14, and Jmax =125.5. The Hamiltonian constants of Coudert et al. [53] were used for the rovibrational energy levels of n =0,1,2 taking into account
L-doubling but not hyperfine structure. For the remaining
bands, the Hamiltonian constants of Amiot [54] were
used. The calculations for high v, J involved significant
extrapolation beyond the available spectral measurements [52].
The line intensities were derived from wave functions
and electric dipole-moment functions (EDMF). The wavefunction calculations were developed by Goorvitch and
Galant [55,56] and by Chackerian et al. [57]. A combination of experimental EDMF by Spencer et al. [58] and
theoretical EDMF by Langhoff et al. [59] and Holtzclaw
et al. [60] was used.
For future work, it is important to address errors in
terms of values identified for high v, J [51,52], and add
various isotopologues of NO to the compilation.
3.5. OH
The 16O1H line parameters in HITEMP1995 were those
generated by Goldman et al. [10], for the X2P’X2P
transitions of Dn = 0, y, 6 with n0 =0, y, 10 and Jmax = 49.5.
The line positions were calculated by combining and
extending the studies of Abrams et al. [61], Mélen et al.
[62], Coxon [63], and Coxon and Foster [64]. The line
intensities were calculated from a composite of Nelson
et al. [65] and Chackerian et al. [57] EDMF.
In subsequent studies, by Cosby et al. [66] and by Colin
et al. [67], it was found that some of the spectroscopic
constants used by Goldman et al. [10] led to errors at
higher vibrational (n 43) and rotational (J419.5) levels,
producing differences with experiment up to 0.14 cm 1.
Recently, Bernath and Colin [68] reanalyzed all the
published experimental data for the electronic ground
state of the hydroxyl radical, to which they added
transitions determined from their solar spectrum [68].
They produced a new set of term values for v = 0, y, 10,
extrapolated to five J values above the last observed one.
These results were used to revise all the OH transitions
(where hyperfine structure was not resolved) in the
HITRAN database with updated positions and groundstate energy values, and the same was performed for the
new HITEMP database. Note that line positions in HITEMP
of the transitions beyond those kept for HITRAN are not
validated.
The intensities have not been updated in this current
effort. Recently, van der Loo and Groenenboom [69,70]
used ab initio calculations to generate a new potential
energy curve, electric dipole-moment function, and spinorbit coupling function for the OH X2P state. These
functions were used to compute a new set of Einstein
A-coefficients. In a future update of the 16O1H line
parameters for HITEMP (and HITRAN) the new Einstein
2148
L.S. Rothman et al. / Journal of Quantitative Spectroscopy & Radiative Transfer 111 (2010) 2139–2150
A-coefficients [70] may be utilized after validation. Also,
more data on the isotopologues of OH should be added to
the compilation.
4. Computation of highly excited molecular spectra
HITEMP continues to be an extremely useful database
for many applications requiring the prediction of spectral
signatures from highly excited vibrational levels, for
example at high altitudes where NLTE conditions prevail,
in auroral events, or in rocket plume simulations. Under
these conditions, vibrational bands are described by their
own vibrational temperatures, which deviate from the
local temperature. Current NLTE models, such as SAMM2
[71] predict line strengths by calculating the population
ratio and combining it with the square of the transition
moments, which can be easily derived from the HITEMP
data, as described in Gamache and Rothman [72].
Recently a semiclassical approach has been employed
by Dothe et al. [73] to compute high-temperature
absorption cross-sections that in the future can supplement HITEMP data where line lists are not readily
available. The approach, dubbed CHEMS (Computation of
Highly Excited Molecular Spectra), exploits the fact that in
the high-temperature limit, rotational lines blend together to form a smooth cross-section envelope that can
be described at a lower spectral resolution. The method is
based on evaluating the Fourier Transform of the time
correlation function of the dipole moment of the vibrating–rotating molecule for normal mode trajectories that
satisfy semiclassical quantization conditions as described
by Schatz [74]. The CHEMS approach was developed to
obtain high-temperature absorption cross-sections required to model plume signatures, when existing databases of such cross-sections are incomplete, inaccurate, or
not available at sufficiently high temperatures. It is
currently being tested in application to CO2, H2O, and
NH3 molecules that are important in hot plume signatures.
comparison to the standard atmospheric database,
HITRAN, in terms of size. The last column presents the
dissociation energies of the molecules.
The structure of the database remains the same as the
HITRAN database, but it is expected that this structure
will be modernized in the near future, most likely
following relational database schemes. One such scheme
that is being investigated is part of the information
system being developed by the water-vapor task group
of IUPAC [34].
There are many applications for a high-temperature
molecular spectroscopic database. It is obviously necessary to include more molecular species. We anticipate
adding NO + , hydrogen halides, N2O, CH4, NH3, and C2H2,
for example. However, the challenges are myriad, both
from the experimental and theoretical sides.
One of the most important additions to HITEMP would
be methane. There exists a substantial effort at the
University of Burgundy called spherical top data system
(STDS) [80] to generate high-resolution spectra of lower
polyads at high temperature (2000 K). Perrin and Soufiani
[81] used the STDS software and used a statistical
approach for hot bands, i.e., create low resolution data
for them.
Ongoing efforts for obtaining absorption parameters of
methane at elevated temperatures, include the work of
Boudon [82] to generate better spectra of lower polyads at
high temperature (2000 K) by updating the STD software
[83] and obtaining the best available input parameters for
it. Meanwhile, a new ab initio potential-energy surface
has appeared [84] and will be studied towards its
potential application to HITEMP.
In the future we hope to add a high-temperature line
list for the ammonia molecule. Recently Yurchenko et al.
[85] computed a low-temperature line list; a hot line list
with improved accuracy is currently being produced.
Instructions for access to the databases can be found
on the web site at http://www.cfa.harvard.edu/HITRAN.
5. Conclusion
Acknowledgments
A new edition of the high-temperature molecular
absorption database, HITEMP2010, has been constructed.
This edition includes molecular transitions for five
species: H2O, CO2, CO, NO, and OH. Table 4 summarizes
the spectral coverage of the new edition, and shows a
We acknowledge the support of NASA through the
Earth Observing System (EOS) program under the Grant
no. NAG5-13534 and the Planetary Atmospheres program
under Grant no. NNX10AB94G. We also acknowledge
the CHEMS (Computation of Highly Excited Molecular
Table 4
Line list comparison between HITEMP2010 and HITRAN [2].
Molecule
Spectral coverage
(cm 1)
Number of isotopologuesa
(HITEMP2010)
Number of transitions
(HITEMP2010)
Number of transitionsb
(HITRAN)
Dissociation energy
(cm 1)
H2O
CO2
CO
NO
OH
0–30,000
258–9648
0–8465
0–9274
0–19,268
6
7
6
3
3
111,377,777
11,167,618
115,218
105 633
40,055
69,201
312,479
4,477
105,079
31,976
41,145.94 70.15 [75]
44,360 [76]
90,674 7 15 [77]
52,265 [78]
35,593 7 25 [79]
a
For H2O, NO, and OH, only the principal isotopologue has been created for high temperature at this time. The other lesser abundant isotopologues
have been transcribed from HITRAN (296 K).
b
The number of transitions listed in this column are for the equivalent number of isotopologues and spectral range consistent with HITEMP2010.
L.S. Rothman et al. / Journal of Quantitative Spectroscopy & Radiative Transfer 111 (2010) 2139–2150
Spectra) SBIR project through Spectral Sciences, Inc. RRG
acknowledges support of this research by the National
Science Foundation through Grant no. ATM-0803135. VIP
and SAT acknowledge support by the Russian Fund of
Basic Research under the Grant nos. 06-05-39016!^EH_a and 09-05-93105-HUHNJ_a.
[23]
[24]
[25]
References
[26]
[1] McClatchey RA, Benedict WS, Clough SA, Burch DE, Calfee RF, Fox K,
et al. AFCRL atmospheric absorption line parameters compilation.
AFCRL-technical report-0096, 1973.
[2] Rothman LS, Gordon IE, Barbe A, Benner DC, Bernath PF, Birk M,
et al. The HITRAN 2008 molecular spectroscopic database. J Quant
Spectrosc Radiat Transfer 2009;110:533–72.
[3] Tinetti G, Vidal-Madjar A, Liang M-C, Beaulieu J-P, Yung Y, Carey S,
et al. Water vapour in the atmosphere of a transiting extrasolar
planet. Nature 2007;448:169–71.
[4] Barber RJ, Tennyson J, Harris GJ, Tolchenov RN. A high-accuracy
computed water line list. Mon Not R Astron Soc 2006;368:
1087–94.
[5] Rothman LS, Wattson RB, Gamache RR, Schroeder J, McCann A.
HITRAN, HAWKS and HITEMP high-temperature molecular database. Proc Soc Photo-Opt Instrum Eng 1995;2471:105–11.
[6] Pollack JB, Dalton JB, Grinspoon D, Wattson RB, Freedman R, Crisp
D, et al. Near-infrared light from Venus’ nightside: a spectroscopic
analysis. Icarus 1993;103:1–42.
[7] Meinel IAB. OH emission bands in the spectrum of the night sky.
Astrophys J 1950;111:555–64.
[8] Wattson RB, Rothman LS. Direct numerical diagonalization: wave of
the future. J Quant Spectrosc Radiat Transfer 1992;48:763–80.
[9] Goorvitch D. Infrared CO line list for the X1S + state. Astrophys J
Suppl Ser 1994;95:535–52.
[10] Goldman A, Schoenfeld WG, Goorvitch D, Chackerian Jr. C, Dothe H,
Mélen F, et al. Updated line parameters for OH X2P—X2P (v00 ,v0 )
transitions. J Quant Spectrosc Radiat Transfer 1998;59:
453–69.
[11] Rothman LS, Gamache RR, Tipping RH, Rinsland CP, Smith MAH,
Benner DC, et al. The HITRAN molecular database—editions of
1991 and 1992. J Quant Spectrosc Radiat Transfer 1992;48:
469–507.
[12] Gordon IE, Dothe H, Rothman LS. The resurrection of the HITEMP
database and its application to the study of stellar and planetary
atmospheres. European Geosciences Union General Assembly,
Vienna, Austria, 2007.
[13] Rothman LS, Gamache RR, Goldman A, Brown LR, Toth RA, Pickett
HM, et al. The HITRAN database—1986 edition. Appl Opt 1987;26:
4058–97.
[14] Rothman LS, Jacquemart D, Barbe A, Chris Benner D, Birk M, Brown
LR, et al. The HITRAN 2004 molecular spectroscopic database.
J Quant Spectrosc Radiat Transfer 2005;96:139–204.
[15] Simecková M, Jacquemart D, Rothman LS, Gamache RR, Goldman A.
Einstein A-coefficients and statistical weights for molecular
absorption transitions in the HITRAN database. J Quant Spectrosc
Radiat Transfer 2006;98:130–55.
[16] Partridge H, Schwenke DW. The determination of an accurate
isotope dependent potential energy surface for water from
extensive ab initio calculations and experimental data. J Chem
Phys 1997;106:4618–39.
[17] Schwenke DW, Partridge H. Convergence testing of the analytic
representation of an ab initio dipole moment function for water:
improved fitting yields improved intensities. J Chem Phys
2000;113:6592–7.
[18] Jørgensen UG, Jensen P, Sørensen GO, Aringer B. H2O in stellar
atmospheres. Astron Astrophys 2001;372:249–59.
[19] Jensen P. The potential energy surface for the electronic ground
state of the water molecule determined from experimental data
using a variational approach. J Mol Spectrosc 1989;133:438–60.
[20] Jørgensen UG, Jensen P. The dipole moment surface and the
vibrational transition moments of H2O. J Mol Spectrosc 1993;161:
219–42.
[21] Shirin SV, Polyansky OL, Zobov NF, Barletta P, Tennyson J. Spectroscopically determined potential energy surface of H16
2 O up to
25,000 cm 1. J Chem Phys 2003;118:2124–9.
[22] Campargue A, Mikhailenko SN, Liu AW. ICLAS of water in the
770 nm transparency window (12,746–13,558 cm 1). Comparison
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
[35]
[36]
[37]
[38]
[39]
[40]
[41]
[42]
[43]
[44]
[45]
2149
with current experimental and calculated databases. J Quant
Spectrosc Radiat Transfer 2008;109:2832–45.
Bailey J. A comparison of water vapor line parameters for modeling
the Venus deep atmosphere. Icarus 2009;201:444–53.
Kranendonk LA, An X, Caswell AW, Herold RE, Sanders ST, Huber R,
et al. High speed engine gas thermometry by Fourier-domain
mode-locked laser absorption spectroscopy. Opt Express 2007;15:
15115–28.
Coppalle A, Vervisch P. Spectral emissivities of H2O vapor at 2900 K
in the 1–9-mm region. J Quant Spectrosc Radiat Transfer 1986;35:
121–5.
Barber RJ, Miller S, Dello Russo N, Mumma MJ, Tennyson J, Guio P.
Water in the near-infrared spectrum of comet 8P/Tuttle. Mon Not R
Astron Soc 2009;398:1593–600.
Sharp CM, Burrows A. Atomic and molecular opacities for brown
dwarf and giant planet atmospheres. Astrophys J Suppl Ser 2007;
168:140–66.
Perez P, Boischot A, Ibgui L, Roblin A. A spectroscopic database for
water vapor adapted to spectral properties at high temperature,
and moderate resolution. J Quant Spectrosc Radiat Transfer
2007;103:231–44.
Voronin BA, Tennyson J, Tolchenov RN, Lugovskoy AA, Yurchenko
SN. A high accuracy computed line list for the HDO molecule. Mon
Not R Astron Soc 2010;402:492–6.
Tennyson J, Zobov NF, Williamson R, Polyansky OL, Bernath PF.
Experimental energy levels of the water molecule. J Phys Chem Ref
Data 2001;30:735–831.
Zobov NF, Polyansky OL, Tennyson J, Shirin SV, Nassar R, Hirao T,
et al. Using laboratory spectroscopy to identify lines in the K- and Lband spectrum of water in a sunspot. Astrophys J 2000;530:
994–8.
Coheur P-F, Bernath PF, Carleer M, Colin R, Polyansky OL, Zobov NF,
et al. A 3000 K laboratory emission spectrum of water. J Chem Phys
2005;122:4307.
Zobov NF, Shirin SV, Polyansky OL, Barber RJ, Tennyson J, Coheur PFo, et al. Spectrum of hot water in the 2000–4750 cm 1 frequency
range. J Mol Spectrosc 2006;237:115–22.
Tennyson J, Bernath PF, Brown LR, Campargue A, Carleer MR,
Császár AG, et al. IUPAC critical evaluation of the rotational–
vibrational spectra of water vapor. Part I—energy levels and
18
transition wavenumbers for H17
2 O and H2 O. J Quant Spectrosc
Radiat Transfer 2009;110:573–96.
Gordon IE, Rothman LS, Gamache RR, Jacquemart D, Boone C,
Bernath PF, et al. Current updates of the water-vapor line list in
HITRAN: a new ‘‘diet’’ for air-broadened half-widths. J Quant
Spectrosc Radiat Transfer 2007;108:389–402.
Jacquemart D, Gamache R, Rothman LS. Semi-empirical calculation
of air-broadened half-widths and air pressure-induced frequency
shifts of water-vapor absorption lines. J Quant Spectrosc Radiat
Transfer 2005;96:205–39.
Gamache RR, Hartmann JM. An intercomparison of measured
pressure-broadening and pressure-shifting parameters of water
vapor. Can J Chem 2004;82:1013–27.
Gamache RR, Laraia AL. N2-, O2-, and air-broadened half-widths,
their temperature dependence, and line shifts for the rotation band
of H16
2 O. J Mol Spectrosc 2009;257:116–27.
Gamache RR. Lineshape parameters for water vapor in the 3.2–
17.76 mm region for atmospheric applications. J Mol Spectrosc
2005;229:9–18.
Wagner G, Birk M, Gamache RR, Hartmann J-M. Collisional
parameters of H2O lines: effect of temperature. J Quant Spectrosc
Radiat Transfer 2005;92:211–30.
Toth RA, Brown LR, Smith MAH, Malathy Devi V, Chris Benner D,
Dulick M. Air-broadening of H2O as a function of temperature:
696–2163 cm 1. J Quant Spectrosc Radiat Transfer 2006;101:
339–66.
Vidler M, Tennyson J. Accurate partition function and thermodynamic data for water. J Chem Phys 2000;113:9766–71.
Tashkun SA, Perevalov VI, Teffo JL, Bykov AD, Lavrentieva NN.
CDSD-296, the carbon dioxide spectroscopic databank: version for
atmospheric applications. In: XIVth symposium on high resolution
molecular spectroscopy. Krasnoyarsk, Russia, July 6–11; 2003. See
also /ftp://ftp.iao.ru/pub/CDSD-296/S.
Tashkun SA, Perevalov VI, Teffo JL, Bykov AD, Lavrentieva NN.
CDSD-1000, the high-temperature carbon dioxide spectroscopic
databank. J Quant Spectrosc Radiat Transfer 2003;82:165–96.
Bharadwaj SP, Modest MF. Medium resolution transmission
measurements of CO2 at high temperature—an update. J Quant
Spectrosc Radiat Transfer 2007;103:146–55.
2150
L.S. Rothman et al. / Journal of Quantitative Spectroscopy & Radiative Transfer 111 (2010) 2139–2150
[46] Modest MF, Mehta RS. Full spectrum k-distribution correlations for
CO2 from the CDSD-1000 spectroscopic databank. Int J Heat Mass
Transfer 2004;47:2487–91.
[47] Wang A, Modest MF. High-accuracy, compact database of narrowband k-distributions for water vapor and carbon dioxide. J Quant
Spectrosc Radiat Transfer 2007;93:245–61.
[48] Bharadwaj SP, Modest MF. A multiscale Malkmus model for
treatment of inhomogeneous gas paths. Int J Therm Sci 2007;46:
479–90.
[49] Tashkun SA, Perevalov VI. CDSD (carbon dioxide spectroscopic
databank): updated and enlarged version for atmospheric applications. Paper T2.3, tenth HITRAN conference. Cambridge MA; 2008.
See also /ftp://ftp.iao.ru/pub/CDSD-2008/S.
[50] Ludwig CB, Ferriso CC, Acton L. High-temperatures spectral
emissivities and total intensities of the 15-mm band system of
CO2. J Opt Soc Am 1966;56:1685–92.
[51] Goldman A Updated line parameters for NO X 2P—X 2P (v0 0 , v0 ).
Progress report. University of Denver, December 27; 1996.
[52] Goldman A, Brown LR, Schoenfeld WG, Spencer MN, Chackerian Jr.
C, Giver LP, et al. Nitric oxide line parameters: review of 1996
HITRAN update and new results. J Quant Spectrosc Radiat Transfer
1998;60:825–38.
[53] Coudert LH, Dana V, Mandin JY, Morillon-Chapey M, Farrenq R. The
spectrum of nitric oxide between 1700 and 2100 cm 1. J Mol
Spectrosc 1995;172:435–48.
[54] Amiot C. The infrared emission spectrum of NO: analysis of the
Dn =0 sequence up to v =22. J Mol Spectrosc 1982;94:150–72.
[55] Goorvitch D, Galant DC. Schrödinger’s radial equation: solution by
extrapolation. J Quant Spectrosc Radiat Transfer 1992;47:391–9.
[56] Goorvitch D, Galant DC. The solution of coupled Schrödinger
equations using an extrapolation method. J Quant Spectrosc Radiat
Transfer 1992;47:505–13.
[57] Chackerian C, Goorvitch D, Benidar A, Farreno R, Guelachvili G,
Martin PM, et al. Rovibrational intensities and electric dipole
moment function of the X2Pi hydroxyl radical. J Quant Spectrosc
Radiat Transfer 1992;48:667–73.
[58] Spencer MN, Chackerian C, Giver LP, Brown LR. The Nitric oxide
fundamental band: frequency and shape parameters for rovibrational lines. J Mol Spectrosc 1994;165:506–24.
[59] Langhoff Jr. SR, W BC, Partidge H. Theoretical dipole moment for the
X2P state of NO. Chem Phys Lett 1994;223:416–22.
[60] Holtzclaw KW, Rawlins WT, Green BD. The effects of centrifugal
distortion on the infrared radiative transition probabilities of
NO(X2P). J Quant Spectrosc Radiat Transfer 1996;55:481–92.
[61] Abrams MC, Davis SP, Rao MLP, Engleman RJ, Brault JW. Highresolution Fourier transform spectroscopy of the Meinel system of
OH. Astrophys J Suppl Ser 1994;93:351–95.
[62] Mélen F, Sauval AJ, Grevesse N, Farmer CB, Servais C, Delbouille L,
et al. A new analysis of the OH radical spectrum from solar infrared
observations. J Mol Spectrosc 1995;174:490–509.
[63] Coxon JA. Optimum molecular constants and term values for the
X2P (n r 5) and A2S + (n r 3) states of OH. Can J Phys 1980;58:
933–49.
[64] Coxon JA, Foster AC. Rotational analysis of hydroxyl vibration–
rotation emission bands: molecular constants for OH X2P,
6 r n r 10. Can J Phys 1982;60:41–8.
[65] Nelson Jr. DD, Schiffman A, Nesbitt DJ, Orlando JJ, Burkholder JBH.
O3 Fourier transform infrared emission and laser absorption of OH
(X2P) radical: an experimental dipole moment function and stateto-state Einstein A coefficients. J Chem Phys 1990;93:7003–19.
[66] Cosby PC, Slanger TG, Huestis DL, Ostenbrock DE. Term energies,
line positions, and spectroscopic constants for the OH Meinel band
[67]
[68]
[69]
[70]
[71]
[72]
[73]
[74]
[75]
[76]
[77]
[78]
[79]
[80]
[81]
[82]
[83]
[84]
[85]
system. In: The 55th Ohio State University International Symposium on Molecular Spectroscopy, 2000. /http://spectroscopy.mps.
ohio-state.edu/symposium_55/symposium/Program/TA.html#TA-01S.
Colin R, Coheur P-F, Kiseleva M, Vandaele AC, Bernath PF. Spectroscopic constants and term values for the X2Pi state of OH (v= 0–10).
J Mol Spectrosc 2002;214:225–6.
Bernath PF, Colin R. Revised molecular constants and term values
for the X2Pi and B2S+ states of OH. J Mol Spectrosc 2009;257:20–3.
van der Loo MPJ, Groenenboom GC. Theoretical transition
probabilities for the OH Meinel system. J Chem Phys 2007;126:
4314.
van der Loo MPJ, Groenenboom GC. Erratum: ‘‘theoretical transition
probabilities for the OH Meinel system’’ [J Chem Phys 126, 114314
(2007)]. J Chem Phys 2008;128:9902.
Dothe H, Duff JW, Gruninger JH, Acharya PK, Berk A, Brown JH.
Users’ manual for SAMM2, SHARC-4 and MODTRAN-4 merged.
Report: AFRL-VS-HA-TR-2004-1001. 2004.
Gamache RR, Rothman LS. Extension of the HITRAN database to
non-LTE applications. J Quant Spectrosc Radiat Transfer 1992;48:
519–25.
Dothe H, Duff JW, Rothman LS, Gordon I. Computation of highly
excited molecular spectra (CHEMS). In: The 19th international
conference on high resolution infrared spectroscopy. Prague, Czech
Republic; 2006.
Schatz GCA. Program for determining primitive semiclassical
eigenvalues for vibrating/rotating nonlinear triatomic molecules.
Comput Phys Commun 1988;51:135.
Matsyutenko P, Rizzo TR, Boyarkin OV. A direct measurement
of the dissociation energy of water. J Chem Phys 2006;125:
181101.
Davies WO. Carbon dioxide dissociation at 60001 to 11,000 1K.
J Chem Phys 1965;43:2809–18.
Eidelsberg M, Roncin J-Y, Le Floch A, Launay F, Letzelter C, Rostas J.
Reinvestigation of the vacuum ultraviolet spectrum of CO and
isotopic species: the B1S + 2X1S + transition. J Mol Spectrosc 1987;
121:309–36.
Brook M, Kaplan J. Dissociation energy of NO and N2. Phys Rev
1954;96:1540–2.
Ruscic B, Wagner AF, Harding LB, Asher RL, Feller D, Dixon DA, et al.
On the enthalpy of formation of hydroxyl radical and gas-phase
bond dissociation energies of water and hydroxyl. J Phys Chem A
2002;106:2727–47.
Wenger C, Champion JP, Boudon V. The partition sum of methane at
high temperature. J Quant Spectrosc Radiat Transfer 2008;109:
2697–706.
Perrin M-Y, Soufiani A. Approximate radiative properties of
methane at high temperature. J Quant Spectrosc Radiat Transfer
2007;103:3–13.
Boudon V. University of Burgundy, Private Communication, 2008.
Wenger Ch, Boudon V, Champion JP, Pierre G. Highly-spherical top
data system (HTDS) software for the spectrum simulation of
octahedral XY6 molecules. J Quant Spectrosc Radiat Transfer
2000;66(1):1–16 See also: /http://icb.u-bourgogne.fr/OMR/SMA/
SHTDSS.
Warmbier R, Schneider R, Sharma AR, Braams BJ, Bowman JM,
Hauschildt PH. Ab initio modeling of molecular IR spectra of
astrophysical interest: application to CH4. Astron Astrophys
2009;495:655–61.
Yurchenko SN, Barber RJ, Yachmenev A, Theil W, Jensen P,
Tennyson J. A variationally computed T= 300 K line list for NH3. J
Phys Chem A 2009;113:11845–55.