The HITRAN2008molecular spectroscopic databa
L.S.Rothman a,Ã,I.E.Gordon a,A.Barbe b,D.Chris Benner c,P.F.Bernath d,M.Birk e,V.Boudon f, L.R.Brown g,A.Campargue h,J.-P.Champion f,K.Chance a,L.H.Coudert i,V.Dana j,V.M.Devi c, S.Fally k,1,J.-M.Flaud i,R.R.Gamache l,A.Goldman m,D.Jacquemart n,I.Kleiner i,
N.Lacome n,W.J.Lafferty o,J.-Y.Mandin j,S.T.Massie p,S.N.Mikhailenko q,C.E.Miller g,
N.Moazzen-Ahmadi r,O.V.Naumenko q,A.V.Nikitin q,J.Orphal i,V.I.Perevalov q,A.Perrin i, A.Predoi-Cross s,C.P.Rinsland t,M.Rotger b,f,M.Sˇimecˇkova´a,2,M.A.H.Smith t,K.Sung g, S.A.Tashkun q,J.Tennyson u,R.A.Toth g,A.C.Vandaele v,J.Vander Auwera k
a Harvard-Smithsonian Center for Astrophysics,Atomic and Molecular Physics Division,Cambridge,MA02138,USA
b Universite´de Reims-Champagne-Ardenne,Groupe de Spectrome´trie Mole´culaire et Atmosphe´rique,51062Reims,France
c The College of William an
d Mary,Department of Physics,Williamsburg,VA23187,USA
d University of York,Department of Chemistry,York,UK
e DLR–Remote Sensing Technology Institute,Wessling,Germany
f CNRS-Universite´de Bourgogne,Institut Carnot de Bourgogne,21078Dijon,France
g California Institute of Technology,Jet Propulsion Laboratory,Pasadena,CA91109,USA
h Universite´Joph Fourier/CNRS,Laboratoire de Spectrometrie Physique,38402Saint-Martin-d’He`res,France
i CNRS et Universite´s Paris Est et Paris7,Laboratoire Inter-Universitaire des Syste`mes Atmosphe´riques,94010Cre´teil,France
j UPMC Universite´Paris06,UMR7092,Laboratoire de Physique Mole´culaire et Applications,75252Paris,France
k Universite´Libre de Bruxelles,Service de Chimie Quantique et Photophysique,C.P.160/09,B-1050Brusls,Belgium
l University of Mass Lowell,Department of Environmental Earth&Atmospheric Sciences,Lowell,MA01854,USA
m University of Denver,Department of Physics,Denver,CO80208,USA
n UPMC Universite´Paris06,UMR7075,Laboratoire de Dynamique,Interactions et Re´activite´,75252Paris,France
渺的组词o National Institute of Standards and Technology,Gaithersburg,MD20899,USA
p National Center for Atmospheric Rearch,Boulder,CO80307,USA
q Institute of Atmospheric Optics,Tomsk634055,Russia
r University of Calgary,Department of Physics and Astronomy,Calgary,AB,Canada T2N1N4
s University of Lethbridge,Department of Physics and Astronomy,Lethbridge,AB,Canada T1K3M4
t NASA Langley Rearch Center,Science Directorate,Hampton,VA23681,USA
u University College London,Department of Physics and Astronomy,London WC1E6BT,UK
v Institut d’Ae´ronomie Spatiale de Belgique,B-1180Brusls,Belgium
a r t i c l e i n f o
Article history:
Received21December2008
Received in revid form
11February2009
Accepted13February2009
Keywords:
HITRAN
Spectroscopic databa
Molecular spectroscopy
Molecular absorption
a b s t r a c t
This paper describes the status of the2008edition of the HITRAN molecular
免费美女图片卡纸做花朵spectroscopic databa.The new edition is thefirst official public relea since the
2004edition,although a number of crucial updates had been made available online
since2004.The HITRAN compilation consists of veral components that rve as input
for radiative-transfer calculation codes:individual line parameters for the microwave
through visible spectra of molecules in the gas pha;absorption cross-ctions for
molecules having den spectral spectra in which the individual lines are
not resolved;individual line parameters and absorption cross-ctions for bands in the
ultraviolet;refractive indices of aerosols,tables andfiles of general properties associated
with the databa;and databa management software.The line-by-line portion of the
Contents lists available at ScienceDirect
journal homepage:/locate/jqsrt
Journal of Quantitative Spectroscopy&
Radiative Transfer
0022-4073/$-e front matter&2009Elvier Ltd.All rights rerved.
doi:10.1016/j.jqsrt.2009.02.013
ÃCorresponding author.Tel.:+16174957474;fax:+16174967519.
E-mail address:lrothman@cfa.harvard.edu(L.S.Rothman).
叙事手法
1Current address:Institut d’Ae´ronomie Spatiale de Belgique,B-1180Brusls,Belgium.
2Current address:J.Heyrovsky Institute of Physical Chemistry,Prague,Czech Republic.
Journal of Quantitative Spectroscopy&Radiative Transfer110(2009)533–572
Spectroscopic line parameters
Absorption cross-ctions
Aerosols databa contains spectroscopic parameters for 42molecules including many of their isotopologues.&2009Elvier Ltd.All rights rerved.
1.Introduction
This article describes the data that have been added,modified,or enhanced in the HITRAN (Hi gh Resolution Tran smission)compilation since the previous update of 2004[1](hereafter called HITRAN2004in the text).The compilation encompass the HITRAN line-transition parameters,infrared cross-ctions,UV (ultraviolet)line-by-line parameters and cross-ctions,aerosol refractive indices,and documentation.The file structure for the compilation remains the same as the previous edition and can be en in Fig.1of Ref.[1].The compilation is available on an anonymous ftp site.Instructions for accessing the databa can be found in the HITRAN web site (www.cfa.harvard.edu/HITRAN ).
The HITRAN databa is the recognized international standard,ud for a vast array of applications including terrestrial and planetary atmospheric remote nsing,transmission simulations,fundamental laboratory spectroscopy studies,industrial process monitoring,and pollution regulatory studies.An international HITRAN advisory committee,compod of a dozen experts in the field of spectroscopy,has been established under the auspices of NASA.This committee reviews and evaluates new data and makes recommendations for updates and replacements in the compilation.
Many recent developments have pushed the requirements of HITRAN in terms of accuracy and degree of completeness.Among the developments one can cite the retrievals that various satellite remote-nsing missions are now capable of due in part to the high signal-to-noi ratio of the spectra and to advances in retrieval algorithms.Notable satellite spectrometer instrumentation includes MLS (Microwave Limb Sounder)[2]and TES (Tropospheric Emission Spectrometer)
[3]on the Aura platform,MIPAS (Michelson Interferometer for Passive Atmospheric Sounding)[4]on ENVISAT,ACE-FTS (Atmospheric Chemistry Experiment)[5]on SCISAT,AIRS (Atmospheric Infrared Sounder)[6]on Aqua,IASI (Infrared Atmospheric Sounding Interferometer)[7]on MetOP-A,OCO (Orbiting Carbon Obrvatory)[8],and GOSAT (Greenhou gas Obrving SATellite)[9].The satellite instruments have put demands on HITRAN that include incread accuracy (by almost an order of magnitude in some cas)for the basic parameters:line position in vacuum wavenumbers,n (in cm À1),intensity of the line,S (in cm À1/(molecule cm À2)),and line-shape parameters.3They also require more species,additional molecular bands,and weak lines throughout the spectral region covered by HITRAN (microwave through UV).In fact,the remote-nsing experiments have demonstrated that the basic Lorentz line-shape parameter for collisional broadening ud in HITRAN ,from which it is possible to calculate the Voigt line profile,is not satisfactory in many cas.To reduce
the residuals between obrvation and simulation,it has often been necessary to invoke more sophisticated non-Voigt line shape functions such as Rautian or Galatry [10]and line mixing.
Section 2of this paper prents the most significant of the improvements featured in this newly updated edition of HITRAN as it relates to the line-by-line parameters.Note that the line lists described here either include or superde intermediate updates that were placed on the HITRAN web site after HITRAN2004.The status of the infrared cross-ctions,ts of UV data,and the aerosol refractive indices of aerosols,are discusd in Sections 3–5.
2.Line-by-line parameters
This edition of HITRAN contains three new entries,methyl bromide (CH 3Br),methyl cyanide (CH 3CN),and tetrafluoromethane (CF 4).It is worth repeating that the number of transitions included in the databa is limited by:(1)a reasonable minimum cutoff in absorption intensity (bad on the nsitivity of instruments that obrve absorption over extreme terrestrial atmospheric path lengths),(2)lack of sufficient experimental data,or (3)lack of calculated transitions.
The format for the line-by-line portion of the compilation remains the same as in the previous edition (e Table 1of Ref.[1]),except that the lf-broadened half-width parameter has now been written in
a Fortran format of F5.3rather than F5.4.The latter distinction is not significant unless the ur employs the Fortran write function.
The molecules for which data are included in the line-by-line portion of HITRAN are mostly compod of small numbers of atoms and have low molecular weights.Large polyatomic molecules have many normal modes of vibration and ‘‘heavy’’species have fundamentals at very low wavenumbers.For three of the molecules in this edition of HITRAN ,SF 6,ClONO 2,and CF 4,we have kept the parameters for this edition in a supplemental folder (e Fig.1of Ref.[1]).The rationale for this is that the line-by-line parameters reprent only a few bands,and neglect many significant hot bands for the ‘‘heavy’’species.For most applications,the IR cross-ctions of the molecules in the HITRAN compilation provide a better simulation.
3
The HITRAN databa does not adhere to SI units for both historical and application-specific reasons.We also employ the symbol n throughout for
line position in cm À1,thereby dropping the tilde (~n
)
that is the official designation of wavenumber.We normally express the HITRAN unit for intensity as cm À1/(molecule cm À2)rather than simplifying to the equivalent cm/molecule.In this way we emphasize the quantity as wavenumber per column density,which is consistent with the viewpoint of atmospheric radiative-transfer codes.L.S.Rothman et al./Journal of Quantitative Spectroscopy &Radiative Transfer 110(2009)533–572
534
L.S.Rothman et al./Journal of Quantitative Spectroscopy&Radiative Transfer110(2009)533–572535 The ur of the HITRAN line-by-line data and the cross-ction data is encouraged to consult and cite the original sources of the data.In the ca of the line-by-line parameters,there are indices pointing to the sources of six parameters: the transition wavenumber,n;the intensity,S;the air-and lf-broadened half-width parameters,g air and g lf;the exponent for the temperature dependence of the air-broadened half-width parameter,n;and the air-pressure shift parameter,d.The sources are contained in a paratefile in the compilation.
The following subctions cover all molecules who parameters have been updated since the last edition of HITRAN[1]. The descriptions are generally ordered by increasing wavenumber region,and
we have attempted to describe the improvements in the line positions and intensities prior to tho in the other parameters,when feasible.Future improvements are also mentioned where necessary.
2.1.H2O(molecule1)
Water vapor spectroscopy is of paramount importance to many applications.Not only are the spectroscopic parameters needed for studies of the climate and energy budget of the Earth,but also for the atmospheres of stars(e for example Ref.[11])and now even exoplanets[12].The recommended line list for water remains in a state of continued evolution. Substantial changes to the half-width parameters for the main isotopologue H216O and the addition of new data for isotopically substituted species are among the prominent recent modifications.
The2004edition of HITRAN[1]featured a major update in line positions and line intensities for all HITRAN water-vapor isotopologues between500and8000cmÀ1bad on the work of Toth[13],with the exception of the principal isotopologue which had calculated values from Coudert[14]up to800cmÀ1.However,recently reported measurements of transitions in the n2band in the1000–2000cmÀ1range[15]suggest that Toth’s data systematically underestimated the intensities of the strongest transitions in this region by between5%and10%.This conclusion is supported by independ
ent ab initio calculations[16].The intensities of the unblended strong lines have therefore been replaced using the new measurements; for four blended strong lines,tho located at1512.30732,1539.05857,1539.06079,and1684.83515cmÀ1,the theoretical results are from variational calculations using an ab initio dipole surface[17].There have been other recent measurements at shorter infrared wavelengths[15,18,19]as well as a comprehensive ab initio analysis of the line intensities[20].The issue of whether or not adjustments are also needed for the line intensities at the wavelengths is currently being studied with a view to coming up with recommendations for a future edition of the databa.
The region9500–14500cmÀ1for the main isotopologue has been updated using the new analysis by Tolchenov and Tennyson[21]who employed a novelfitting technique to reanalyze a ries of Fourier transform absorption spectra of pure water vapor recorded by Schermaul et al.[22,23].However,any data attributed to Brown et al.[24]that were in HITRAN2004have been retained.Analogously,the14500-to26000-cmÀ1region has been updated using the work of Tolchenov et al.[25]replacing the data from Coheur et al.[26]in HITRAN2004.Comparisons with previous studies on water-vapor absorption in this region suggest that the new parameters give a more consistent reprentation of the spectrum.
An update has also been made for the parameters of H217O and H218O isotopologues in the near-IR and visible region bad on the work of Tanaka et al.[27].This work is a reanalysis of long-path length Fourier transform spectra originally recorded at Kitt Peak by Chevillard et al.[28]and analyzed initially by Tanaka et al.[29].The lines listed previously in this region for both isotopologues have been removed and replaced by1087lines of H218O spanning the range 12400–14520cmÀ1and891lines of H217O in the range11365–14475cmÀ1.In addition,some misidentified lines that have now been attributed to oxygen,have been removed from the water-vapor line list.
A major addition has been made with3528monodeuterated water-vapor(HDO)transitions in the near infrared and visible, specifically11600–23000cmÀ1.Previous editions of the databa did not contain any HDO transitions in this region.The data are due to a re-analysis by Voronin et al.[30]of the long-path Fourier transform spectrum recorded by Bach et al.[31].
The pressure-broadened half-width parameters for the three most abundant isotopologues of water,H216O,H218O,and H217O,have been completely updated.Air-broadened half-widths were updated in2006(an interim update)using an algorithm bad on physical principles and statistics developed by Gordon et al.[32],which t a new criterion for the best available air-broadened half-width parameters using a mixture of measurements,calculated,and mi-empirical data. The new
parameters have been tested for different remote-nsing applications and were found to give improved profiles for atmospheric constituents.The algorithm has been improved for the current relea of HITRAN:additional measurements of g air and d[18,33–39]and g lf[18,38–45]have been added to the measurement databas.Additional data[46,47]have been added to the theoretical databa of g air,n,and d.The databa of calculations of g lf for water vapor now contains the data of Antony et al.[48,49]and Cazzoli et al.[44].
The temperature dependence of the air-broadened half-widths has now been added to all water-vapor transitions via an algorithm thatfirst eks values from CRB(Complex Robert–Bonamy)calculations[46,47,50].If a CRB value for a transition is not found,the n values as a function of rotational quantum numbers from Table7of Ref.[1]are ud.
怎样美白皮肤2.2.CO2(molecule2)
High-resolution spectroscopic monitoring of the evolution of carbon dioxide in the terrestrial atmosphere is obviously one of great importance for policy makers.Carbon dioxide is also prevalent in the atmospheres of some rocky planets,such
as Venus and Mars.With its many bands of very different intensity throughout the spectrum,carbon di
oxide is also an excellent tool for probing atmospheres to different depths.
Since the last edition of the HITRAN databa [1],there have been a large number of experimental and theoretical investigations of carbon dioxide spectra.A notable effort is the t of extensive Fourier transform spectroscopy (FTS)experiments carried out by the Jet Propulsion Laboratory (JPL)[51–58]in order to support the OCO mission [8].The results of the efforts for the 4300–7000cm À1region have been compiled into a HITRAN-like databa [57]with parameters for nine different isotopologues (including 13C 18O 2which was not previously tabulated in HITRAN ).The parameters listed in Ref.[57]cover a wide dynamic range (4Â10À30–1.29Â10À21cm À1/(molecule cm À2)at 296K)which is substantially larger than the FTS experimental detection parameters for some high-J lines as well as for lines of weak unobrved bands were theoretically extrapolated.Parallel experiments featuring the cavity ring down spectroscopy (CRDS)technique
[59–64]in the 5851–7045cm À1region have shown that theoretical extrapolations of the FTS data in Ref.[57]deviate riously from the CRDS line positions and line intensities for some of the higher-J lines,while some of the weaker bands,obrved to be above 4Â10À30cm À1/(molecule cm À2)are missing completely from the predicted line list (e discussion in Refs.[62,65,66]).The discrepancies are thought to have a reliable basis becau the CRDS technique allows the detection
of lines with much weaker intensities than tho with the FTS,although CRDS spectra are inferior to FTS spectra in terms of overall accuracy of determining line positions.
Simultaneously,great progress has been made in the global effective Hamiltonian (EH)model developed at the
Universite
´Pierre et Marie Curie (Paris,France)and the Institute of Atmospheric Optics (Tomsk,Russia)[67–70],which was ud in the calculation of the theoretical Carbon Dioxide Spectroscopic Databank (CDSD)[71],significantly improving and extending the previous version [72]and achieving a pronounced agreement with the CRDS experiments.The improvement and extension of the CDSD databank have been achieved due to incorporating new measurements performed during the last five years into the global modeling.The above mentioned CRDS measurements in Grenoble and FTS measurements at JPL have had an especially strong impact on the quality of the modeling.
好听的短句The prent atmospheric version of CDSD consists of 419610lines belonging to 12C 16O 2,13C 16O 2,16O 12C 18O,16O 12C 17O,16O 13C 18O,16O 13C 17O,and 12C 18O 2covering a wavenumber range of 5–12784cm À1.The intensity cutoff of CDSD was t to 10À30cm À1/(molecul
e cm À2).On average,the residuals between CDSD calculated line positions and tho obrved are two times larger than measurement uncertainties.CDSD calculated line intensities are almost always within their measurement uncertainties.
The current atmospheric version of the databank is available via an anonymous ftp site ftp.iao.ru in the folder /pub/CDSD-2008/296.The same site also contains two other dedicated versions of the databank:a version for high-temperature applications (/pub/CDSD-2008/1000)and a version for studying the atmospheres of Venus and Mars (/pub/CDSD-2008/Venus).
The need for a nsible mixing of the experimental and theoretical data is obviously required in the 4300to 7000cm À1region in order to support atmospheric remote nsing of the earth-like planets (Earth,Mars and Venus).In order to do that one has to consider the following caveats:
1.The databa [57](hereafter referred to as the OCO data t)is bad on FTS measurements that are very accurate and,
besides line positions and intensities,allow measurements of collision broadening parameters.However,theoretical extrapolations applied in the OCO data t for transitions weaker than 10À26cm À1/(molecule cm À2)for the principal isotopologue and 10À27cm À1/(molecule cm À2)
for the other isotopologues have led to some very large deviations from subquent obrvations in predicting line positions and especially intensities.
2.The data collected in the cavity ring down lar experiments (hereafter referred as CRDS data)is nearly complete
for the lines stronger than 5Â10À29cm À1/(molecule cm À2).However,the typical accuracy of the line positions (1Â10À3cm À1)is inferior to that of FTS experiments (4Â10À5cm À1).Finally,CRDS measurements do not provide data below 5851cm À1and do not provide pressure-induced parameters.Note that the data t for 13C 16O 2,16O 13C 18O,16O 13C 17O,13C 18O 2and 18O 13C 17O compiled in Ref.[65]provide experimental line positions supplemented with intensities calculated using the EH model and effective dipole moment parameters for completeness (13C 18O 2and 18O 13C 17O isotopologues have not been tabulated in HITRAN before).For 12C 16O 2,16O 12C 17O and 16O 12C 18O [62]only line positions are provided,although parameters for 12C 16O 2are also tabulated in Ref.[61]where the experimental line positions and intensities are supplemented with the CDSD intensities.
3.The theoretical CDSD databank is quite complete,with intensities down to 1Â10À30cm À1/(molecule cm À2),at least for
读后感作文600the majority of the HITRAN isotopologues.It has excellent predictive capabilities for line positions and intensities,although it is,of cour,not as good as the accuracy achieved by experiment.In addition,a minor limitation of the EH method occurs when there are interpolyad anharmonic couplings.Four such occurrences have been obrved for the asymmetric isotopologues,namely 16O 12C 18O [62],16O 13C 17O [65]and 16O 13C 18O [64,65].Although the resonance interactions are not common for carbon dioxide,small deviations in the values of predicted line positions and line intensities values from their real values cannot be ruled out completely.
With this information in mind,a procedure,shown in a schematic diagram in Fig.1,was developed in order to keep only the best parameters from the OCO,CRDS and CDSD data ts for compiling the HITRAN 2008CO 2line list in the 4300to
L.S.Rothman et al./Journal of Quantitative Spectroscopy &Radiative Transfer 110(2009)533–572
536
7000cm À1region,which completely replaces HITRAN2004data in this wavenumber range.In this procedure,the CO 2transitions that are critical for the OCO mission are always assumed to have superior quality within the FTS detection limit (lines stronger than 10À26cm À1/(molecule cm À2)for t
he principal isotopologue and 10À27cm À1/(molecule cm À2)for the other isotopologues.For the weaker lines in the 5851–7045cm À1region,the CRDS line positions are taken,wherever available,and supplemented with CDSD intensities.For the weak lines not prent in the CRDS data t (this especially concerns lines below 5851cm À1and blended lines unobrved by CRDS due to overlapping with stronger lines),the CDSD line parameters were taken.The line positions and intensities for two rare isotopologues,17O 12C 18O and 18O 13C 18O,which are abnt in CDSD,have been taken from the CO 2list generated for OCO.
Finally,the g air ,g lf ,n ,and d parameters,available in the OCO data t,have been included in this combined line list.Note that the parameters are slightly different from tho listed in the supplementary file of Ref.[57],due to improvements accomplished through the newer work of Predoi-Cross et al.[73].
All combined (mixed)data ts are relatively new.The procedure suggested above is a temporary but necessary solution that has to be tested against atmospheric retrievals.As new,highly accurate measurement data ts become available,this procedure will have to be refined for future updates of the HITRAN databa.
冬奥会开幕
For the spectral regions below 4300cm À1and above 7000cm À1the following improvements have been made to the HITRAN databa:(1)The four bands above 9650cm À1that were added to HITRAN2004were found to have an error associated with an incorrect account of nuclear spin statistics.The bands have now been replaced with the lines from the CDSD databank above 9650cm À1,which includes veral other additional bands.The data are important for the studies of the Venus atmosphere [74].(2)In HITRAN2004some of the bands of the principal isotopologue in the 2.8-m m region were bad on extrapolations of limited experimental data.For example,the 23301–02201band (centered at 3555cm À1)contained 188lines which were extrapolated from 16measured lines and the interaction between the vibrational levels 23301and 12212was not well accounted for at higher-J values.An analogous problem occurs in the 40002–11102and the 30001–01101bands (centered at 3543and 3557cm À1,respectively).Thus,the line positions and intensities for the bands were replaced with the ones from the CDSD databank.(3)Recent FTS measurements [75]of the line intensities for the 11112–01101band of the 13C 16O 2isotopologue (centered at 3499cm À1)have shown differences up to 100%compared to HITRAN2004(the error code in the former HITRAN indeed indicated problems for this band).The intensities of this band were previously calculated by the DND method of Wattson and Rothman [76],which did not fully account for perturbations.Therefore,the parameters for this band have been replaced with the ones f
rom CDSD.(4)As was noted by Wang et al.[77],the HITRAN2004line positions of the 30003–00001band (at 3857cm À1)for 16O 12C 18O differ from new experimental ones by À0.1to 0.1cm À1.The line positions have now been replaced with line positions calculated using the EH method.(5)Although at this point the high-quality experimental data from Toth et al.[58]have not yet been
Fig.1.Flow diagram outlining the asmbly of the CO 2HITRAN2008line list in the 4300–7000cm À1region.
L.S.Rothman et al./Journal of Quantitative Spectroscopy &Radiative Transfer 110(2009)533–572537
