Search

With a new generation of observatories coming online this decade, the process of characterizing exoplanet atmospheres will need to be reinvented. Currently mostly on the instrumental side, characterization bottlenecks will soon stand by the models used to translate spectra into atmospheric properties. Limitations stemming from our stellar and atmospheric models have already been highlighted. Here, we show that the current limitations of the opacity models used to decode exoplanet spectra propagate into an accuracy wall at ~0.5-1.0 dex (i.e., 3 to 10x) on the atmospheric properties, which is an order of magnitude above the precision targeted by JWST Cycle 1 programs and needed for, e.g., meaningful C/O-ratio constraints and biosignatures identification. We perform a sensitivity analysis using nine different opacity models and find that most of the retrievals produce harmonious fits owing to compensations in the form of >5$\sigma$ biases on the derived atmospheric parameters translating in the aforementioned accuracy wall. We suggest a two-tier approach to alleviate this problem involving a new retrieval procedure and guided improvements in opacity data, their standardization and optimal dissemination., Comment: 16 Pages, 12 Figures

The amount of water vapor in the terrestrial atmosphere is highly variable both spatially and temporally. In the tropics it sometimes constitutes 4-5% of the atmosphere. At the same time collisional broadening of spectral lines by water vapor is much larger than that by nitrogen and oxygen. Therefore, in order to accurately characterize and model spectra of the atmospheres with significant amounts of water vapor, the line-shape parameters for spectral lines broadened by water vapor are required. In this work, the line-broadening coefficients (and their temperature dependence exponents) due to the pressure of water vapor for lines of CO2, N2O, CO, CH4, O2, NH3, and H2S from both experimental and theoretical studies were collected and carefully reviewed. A set of semi-empirical models based on these collected data was created and then used to estimate water broadening and its temperature dependence for all transitions of selected molecules in the HITRAN2016 database.

This paper presents an overview of the current status of the Virtual Atomic and Molecular Data Centre (VAMDC) e-infrastructure, including the current status of the VAMDC-connected (or to be connected) databases, updates on the latest technological development within the infrastructure and a presentation of some application tools that make use of the VAMDC e-infrastructure. We analyse the past 10 years of VAMDC development and operation, and assess their impact both on the field of atomic and molecular (A&amp, M) physics itself and on heterogeneous data management in international cooperation. The highly sophisticated VAMDC infrastructure and the related databases developed over this long term make them a perfect resource of sustainable data for future applications in many fields of research. However, we also discuss the current limitations that prevent VAMDC from becoming the main publishing platform and the main source of A&amp, M data for user communities, and present possible solutions under investigation by the consortium. Several user application examples are presented, illustrating the benefits of VAMDC in current research applications, which often need the A&amp, M data from more than one database. Finally, we present our vision for the future of VAMDC.

We present an evaluation of airborne intensity‐modulated continuous‐wave (IM‐CW) lidar measurements of atmospheric column CO2 mole fractions during the Atmospheric Carbon and Transport–America (ACT‐America) project. This lidar system transmits online and offline wavelengths simultaneously on the 1.57111‐μm CO2 absorption line, with each modulated wavelength using orthogonal swept frequency waveforms. After the spectral characteristics of this system were calibrated through short‐path measurements, we used the HITRAN spectroscopic database to calculate the average‐column CO2 mole fraction (XCO2) from the lidar‐measured optical depths. Using in situ measurements of meteorological parameters and CO2 concentrations for calibration data, we demonstrate that our lidar CO2 measurements were consistent from season to season and had an absolute calibration error (standard deviation) of 0.80 ppm when compared to XCO2 values calculated from in situ measurements. By using a 10‐s or longer moving average, a precision of 1 ppm or better was obtained. The estimated CO2 measurement precision for 0.1‐, 1‐, 10‐, and 60‐s averages was determined to be 3.4, 1.2, 0.43, and 0.26 ppm, respectively. These correspond to measurement signal‐to‐noise ratios of 120, 330, 950, and 1,600, respectively. The drift in XCO2 over 1‐hr of flight time was found to be below 0.1 ppm. These analyses demonstrate that the measurement stability, precision, and accuracy are all well below the thresholds needed to study synoptic‐scale variations in atmospheric XCO2.

Mass-dependent diagonal Born-Oppenheimer corrections (DBOC) to the ab initio electronic ground state potential energy surface for tseveral isotopologues of the ozone molecule are reported for the first time. The comparison with experimental band centers shows a significant improvement of the accuracy with respect to the best Born-Oppenheimer (BO) ab initio calculations reducing the total root-mean-squares (calculated - observed) deviations by about factor of two. For the set of 16O3 vibrations up to five bending and four stretching quanta, the mean (calculated - observed) deviations drop down from 0.7 cm-1 (BO) to about 0.1 cm-1, with the most pronounced improvement seen for bending states and for mixed bend-stretch polyads. In case of bending band centers directly observed under high spectral resolutions, the errors are reduced by more than order of magnitude from observed levels, approaching nearly experimental accuracy. New sets of ab initio vibrational states can be used for improving spectroscopic effective models for analyses of observed high-resolution spectra, particularly in cases of accidental resonances with ,,dark'' states requiring accurate theoretical predictions.

A methane line list for the HITEMP spectroscopic database, covering 0-13,400 cm$^{-1}$ ($>$746 nm), is presented. To create this compilation, ab initio line lists of $^{12}$CH$_{4}$ from Rey et al. (2017) ApJ, 847, 105 (provided at separate temperatures in the TheoReTS information system), are now combined with HITRAN2016 methane data to produce a single line list suitable for high-temperature line-by-line calculations up to 2000 K. An effective-temperature interpolation model was created in order to represent continuum-like features at any temperature of interest. This model is advantageous to previously-used approaches that employ so-called ``super-lines'', which are suitable only at a given temperature and require separate line lists for different temperatures. The resultant HITEMP line list contains $\sim$32 million lines and is significantly more flexible than alternative line lists of methane, while accuracy required for astrophysical or combustion applications is retained. Comparisons against experimental observations of methane absorption at high temperatures have been used to demonstrate the accuracy of the new work. The line list includes both strong lines and quasi-continuum features and is provided in the common user-friendly HITRAN/HITEMP format, making it the most practical methane line list for radiative transfer modeling at high-temperature conditions., 25 pages, 12 figures, submitted to the Astrophysical Journal Supplement Series

Total internal partition sums (TIPS) are reported for 166 isotopologues of 51 molecules important in planetary atmospheres. Molecules 1 to 50 are taken from the HITRAN2016 list, and, in some cases, additional isotopologues are considered for some of the molecules. Molecules 51–53 are C3H4, CH3, and CS2, respectively. TIPS are not reported for the O atom and CF4; thus, while there are 53 species in the list, data are reported for 51 molecules. The TIPS are determined by various methods from 1 K to a Tmax that ensures the TIPS reported have converged. These data are provided with HITRAN2016 and a new version of the TIPS code is available in both FORTRAN and python languages.

The accuracy of atmospheric trace gas retrievals depends directly on the accuracy of the molecular absorption model used within the retrieval algorithm. For remote sensing of well-mixed gases, such as carbon dioxide (CO 2 ), where the atmospheric variability is small compared to the background, the quality of the molecular absorption model is key. Recent updates to the 1.6 µm and 2.06 µm CO 2 absorption model used within the Orbiting Carbon Observatory (OCO-2) algorithm are described and validated. A set of 164 atmospheric spectra from the Total Carbon Column Observing Network (TCCON) is used to compare three models, both previous and current versions of absorption coefficient tables (largely derived from recent multispectrum fitting analyses targeted specifically at these bands) as well as a recent model constructed to use the HITRAN 2012 compilation. Both spectral residuals and retrieved column-averaged CO 2 mixing ratios (XCO 2 ) are included in the comparison. Absorption coefficients based on the updated multispectrum fitting analyses provide residuals comparable to or smaller than either the previous version of the multispectrum fits or the HITRAN 2012-based model. For the 2.06 µm band the updated model finds noticeably lower residuals for low water content cases. It is found that apart from a scaling factor the prior and updated absorption models result in similar retrieved values of XCO 2 for the 2.06 µm band and a slightly different airmass dependence for the 1.6 µm band.

Accounting for volatile organic compounds (VOCs) is crucial for many applications, including, for example, climate change monitoring, tracking of pollution and biomass burning, and atmospheric remote sensing. For many such molecules, the line-by-line spectroscopic data are still either unavailable, or incomplete. The spectroscopic databases, however, often give the wavenumber-dependent absorption cross-sections for each temperature and pressure set. In this paper we give a detailed overview for the current state of the HITRAN2016 infrared absorption cross-section part. The updated cross-sections include data for around 320 compounds, with more than 260 compounds being included in the database for the first time. These cross-sections come from high-resolution laboratory observations, predominantly from Fourier transform spectrometers. The update largely relies on spectra from the Pacific Northwest National Laboratory quantitative spectroscopic database and the Hodnebrog et al. (Rev Geophys 2013) compilation, but also on other recently published data. We give a description for the most atmospherically important compounds arranged by chemical categories, and validate the existing data sources for many of them. We also describe the HITRAN infrastructure which gives access to the data for all molecules from HITRAN, and discuss the best practices of searching the data. This update increases the amount of molecules in the HITRAN database almost six-fold compared to the previous edition, HITRAN2012; it is available via the HITRANonline website (www.hitran.org) and HITRAN Application Programming Interface (HAPI).

We describe a new, online interface to the HITRAN database that overcomes the many limitations of the existing, 160-character fixed-width, text based format (the “.par” files mostly distributed through an FTP site until now). The interface, called HITRANonline, accesses a relational database [JQSRT 2013:130, 57-61] in which the spectroscopic data are stored in a flexible, extensible and structured format. This allows an arbitrary number of different parameters for each transition to be stored so that HITRAN can represent, for example, non-Voigt line shape profiles as well as parameters representing broadening by species other than “air” and “self”. The online interface provides many ways of visualizing data as part of querying the database and allows users to create and save their own output formats to suit their own needs. A bibliography file produced with each data file provides citations and notes to the original data sources to make it easier for users to credit data providers. Once registered with the HITRANonline service, users also have (private) access to thier own search history which summarizes and can repeat queries.

The HITRAN Application Programming Interface (HAPI) is presented. HAPI is a free Python library, which extends the capabilities of the HITRANonline interface ( www.hitran.org ) and can be used to filter and process the structured spectroscopic data. HAPI incorporates a set of tools for spectra simulation accounting for the temperature, pressure, optical path length, and instrument properties. HAPI is aimed to facilitate the spectroscopic data analysis and the spectra simulation based on the line-by-line data, such as from the HITRAN database [JQSRT (2013) 130, 4–50], allowing the usage of the non-Voigt line profile parameters, custom temperature and pressure dependences, and partition sums. The HAPI functions allow the user to control the spectra simulation and data filtering process via a set of the function parameters. HAPI can be obtained at its homepage www.hitran.org/hapi .

We demonstrate a new method for populating line-by-line spectroscopic databases with beyond-Voigt line-shape parameters, which is based on ab initio quantum scattering calculations. We report a comprehensive dataset for the benchmark system of He-perturbed H2 (we cover all the rovibrational bands that are present in the HITRAN spectroscopic database). We generate the entire dataset of the line-shape parameters (broadening and shift, their speed dependence, and the complex Dicke parameter) from fully ab initio quantum-scattering calculations. We extend the previous calculations by taking into account the centrifugal distortion for all the bands and by including the hot bands. The results are projected on a simple structure of the quadratic speed-dependent hard-collision profile. We report a simple and compact formula that allows the speed-dependence parameters to be calculated directly from the generalized spectroscopic cross sections. For each line and each line-shape parameter, we provide a full temperature dependence within the double-power-law (DPL) representation, which makes the dataset compatible with the HITRAN database. The temperature dependences cover the range from 20 to 1000 K, which includes the low temperatures relevant for the studies of the atmospheres of giant planets. The final outcome from our dataset is validated on highly accurate experimental spectra collected with cavity ring-down spectrometers. The methodology can be applied to many other molecular species important for atmospheric and planetary studies.

A new ro-vibrational line list for the 12C32S2, 32S12C34S, 32S12C33S, and 13C32S2 isotopologues of carbon disulfide, including line positions, intensities, and linewidths, has been produced as part of the HITRAN2020 project. The carbon disulfide line list covers the range from about 0 to 6470 cm−1 (1.55 µm). The calculations of the line positions and intensities were performed by applying the PGOPHER program using a large set of measured line positions available in the literature, including the most recent high-precision dual-comb laser spectroscopy measurements and transition dipole moments for each measured band. The bands which did not have available intensity measurements were scaled to the experimental cross-sections from the PNNL database. Semi-empirical models for rotational dependence of air- and self-broadened half widths are developed based on the available experimental data. A total of 94992 transitions corresponding to 423 cold and hot bands of the 12C32S2, 32S12C34S, 32S12C33S, and 13C32S2 isotopologues up to a maximum rotational angular momentum J=150 have been calculated. The CS2 line list has been generated in HITRAN format and is provided as supplementary material and will be a part of the next edition of the HITRAN database distributed through https://www.hitran.org . This list will be useful in the fields of atmospheric environmental chemistry, medical diagnostics and planetary atmospheres.

A number of recent experimental measurements were employed to update the self- and air- broadening coefficients of line transitions of H2S for HITRAN2016. A semi-empirical method based on the fit to the available experimental data was used to compute the self- and air-broadening coefficients of all of the transitions of H2S. The line positions and intensities of the transitions in the (010)-(000) band were improved. In addition intensities and line positions of some of the lines in the triad bands and in the NIR region were found to be inaccurate in the previous version in HITRAN, and they were updated and fixed in HITRAN2016.

Spectroscopic analysis of hydrocarbons, halocarbons and related species is required in atmospheric applications such as climate-change research, tracking of pollution and biomass burning, and remote sensing. Critically evaluated line-by-line spectroscopic data are needed for such analyses. To address this need we report the large addition of spectra to the HITRAN2016 database. This extension increases the number of molecules in the database by almost a factor of 6 from that in HITRAN2012. Spectra from the Pacific Northwest National Laboratory database (Appl. Spectrosc. 2004;58:1452–61) and the Hodnebrog et al. compilation (Rev Geophys 2013;51:300–78) have been integrated into HITRAN2016. Spectra from high-resolution laboratory observations, predominantly from Fourier transform spectrometers, are now provided for 328 compounds in the IR spectral range. Spectra for atmospherically-important compounds organized by chemical category are presented. The HITRAN infrastructure ( www.hitran.org ), HITRAN Application Programming Interface (HAPI), and best practices for searching and downloading data are discussed.

Ab initio dipole moment surfaces (DMSs) of the ozone molecule are computed using the MRCI-SD method with AVQZ, AV5Z, and VQZ-F12 basis sets on a dense grid of about 1950 geometrical configurations. The analytical DMS representation used for the fit of ab initio points provides better behavior for large nuclear displacements than that of previous studies. Various DMS models were derived and tested. Vibration-rotation line intensities of 16O3 were calculated from these ab initio surfaces by the variational method using two different potential functions determined in our previous works. For the first time, a very good agreement of first principle calculations with the experiment was obtained for the line-by-line intensities in rotationally resolved ozone spectra in a large far- and mid-infrared range. This includes high overtone and combination bands up to ΔV = 6. A particular challenge was a correct description of the B-type bands (even ΔV3 values) that represented major difficulties for the previous ab initio investigations and for the empirical spectroscopic models. The major patterns of various B-type bands were correctly described without empirically adjusted dipole moment parameters. For the 10 μm range, which is of key importance for the atmospheric ozone retrievals, our ab initio intensity results are within the experimental error margins. The theoretical values for the strongest lines of the ν3 band lie in general between two successive versions of HITRAN (HIgh-resolution molecular TRANsmission) empirical database that corresponded to most extended available sets of observations. The overall qualitative agreement in a large wavenumber range for rotationally resolved cold and hot ozone bands up to about 6000 cm−1 is achieved here for the first time. These calculations reveal that several weak bands are yet missing from available spectroscopic databases.

Recent experimental measurements and calculations were employed to update the self- and air- broadening coefficients of line transitions of SO2. A semi-empirical method based on the fit to the available experimental data was used to compute the air-broadening coefficients of SO2. The self-broadening coefficients of SO2 were updated using scaled semi-classical calculations from Tasinato et al, J. Quant. Spectrosc. Radiat. Transf. 130 (2013) 233–248. .

Experimental capabilities of molecular spectroscopy and its applications nowadays require a sub-percent or even sub-per mille accuracy of the representation of the shapes of molecular transitions. This implies the necessity of using more advanced line-shape models which are characterized by many more parameters than a simple Voigt profile. It is a great challenge for modern molecular spectral databases to store and maintain the extended set of line-shape parameters as well as their temperature dependences. It is even more challenging to reliably retrieve these parameters from experimental spectra over a large range of pressures and temperatures. In this paper we address this problem starting from the case of the H2 molecule for which the non-Voigt line-shape effects are exceptionally pronounced. For this purpose we reanalyzed the experimental data reported in the literature. In particular, we performed detailed line-shape analysis of high-quality spectra obtained with cavity-enhanced techniques. We also report the first high-quality cavity-enhanced measurement of the H2 fundamental vibrational mode. We develop a correction to the Hartmann–Tran profile (HTP) which adjusts the HTP to the particular model of the velocity-changing collisions. This allows the measured spectra to be better represented over a wide range of pressures. The problem of storing the HTP parameters in the HITRAN database together with their temperature dependences is also discussed.

Recent experimental measurements and calculations were used to update the self- and air- broadening coefficients of CH3Cl. The semi-empirical method was used to compute the self-broadening coefficients of CH3Cl. The air-broadening coefficients of CH3Cl were calculated by a semi-classical approach. The1900-2600 cm-1 spectral region was completely with line positions and intensitiescalculated based on the recently analyzed Fourier transform spectra. The algorithm we used to update the broadening and temperature dependence parameters is described in this report and the corresponding CH3Cl file is provided which containsthe traditional 160 characters.

Energies and lifetimes (widths) of vibrational states above the lowest dissociation limit of $^{16}\mathrm{O}_{3}$ were determined using a previously developed efficient approach, which combines hyperspherical coordinates and a complex absorbing potential. The calculations are based on a recently computed potential energy surface of ozone determined with a spectroscopic accuracy [Tyuterev et al., J. Chem. Phys. 139, 134307 (2013)]. The effect of permutational symmetry on rovibrational dynamics and the density of resonance states in ${\mathrm{O}}_{3}$ is discussed in detail. Correspondence between quantum numbers appropriate for short- and long-range parts of wave functions of the rovibrational continuum is established. It is shown, by symmetry arguments, that the allowed purely vibrational ($J=0$) levels of $^{16}\mathrm{O}_{3}$ and $^{18}\mathrm{O}_{3}$, both made of bosons with zero nuclear spin, cannot dissociate on the ground-state potential energy surface. Energies and wave functions of bound states of the ozone isotopologue $^{16}\mathrm{O}_{3}$ with rotational angular momentum $J=0$ and 1 up to the dissociation threshold were also computed. For bound levels, good agreement with experimental energies is found: The rms deviation between observed and calculated vibrational energies is 1 ${\mathrm{cm}}^{\ensuremath{-}1}$. Rotational constants were determined and used for a simple identification of vibrational modes of calculated levels.

To increase the potential for use of the HITRAN database in astronomy, experimental and theoretical line-broadening coefficients, line shifts and temperature-dependence exponents of molecules of planetary interest broadened by H2, He, and CO2 have been assembled from available peer-reviewed sources. The collected data were used to create semi-empirical models so that every HITRAN line of the studied molecules has corresponding parameters. Since H2 and He are major constituents in the atmospheres of gas giants, and CO2 predominates in atmospheres of some rocky planets with volcanic activity, these spectroscopic data are important for remote sensing studies of planetary atmospheres. In this paper we make the first step in assembling complete sets of these parameters, thereby creating datasets for SO2, NH3, HF, HCl, OCS and C2H2.

The air-broadening coefficients for HO2 in HITRAN were updated sincedifferent measurements have recently been made in infraredandsubmillimeter-wave spectral regions. The semi-empiricalprocedure we developed is described in this report andcorresponding HO2 file in the traditional 160 character format isprovided.

In the recent article by Byrne and Goldblatt, "Radiative forcing for 28 potential Archean greenhouse gases", Clim. Past. 10, 1779–1801 (2014), the authors employ the HITRAN2012 spectroscopic database to evaluate the radiative forcing of 28 Archean gases. As part of the evaluation of the status of the spectroscopy of these gases in the selected spectral region (50–1800 cm−1), the cross sections generated from the HITRAN line-by-line parameters were compared with those of the PNNL database of experimental cross sections recorded at moderate resolution. The authors claimed that for NO2, HNO3, H2CO, H2O2, HCOOH, C2H4, CH3OH and CH3Br there exist large or sometimes severe disagreements between the databases. In this work we show that for only three of these eight gases a modest discrepancy does exist between the two databases and we explain the origin of the differences. For the other five gases, the disagreements are not nearly at the scale suggested by the authors, while we explain some of the differences that do exist. In summary, the agreement between the HITRAN and PNNL databases is very good, although not perfect. Typically differences do not exceed 10 %, provided that HITRAN data exist for the bands/wavelengths of interest. It appears that a molecule-dependent combination of errors has affected the conclusions of the authors. In at least one case it appears that they did not take the correct file from PNNL (N2O4 (dimer)+ NO2 was used in place of the monomer). Finally, cross sections of HO2 from HITRAN (which do not have a PNNL counterpart) were not calculated correctly in BG, while in the case of HF misleading discussion was presented there based on the confusion by foreign or noise features in the experimental PNNL spectra.

Since the discovery of anomalies in ozone isotope enrichment, several fundamental issues in the dynamics linked to the shape of the potential energy surface in the transition state region have been raised. The role of the reeflike structure on the minimum energy path is an intricate question previously discussed in the context of chemical experiments. In this Letter, we bring strong arguments in favor of the absence of a submerged barrier from ultrasensitive laser spectroscopy experiments combined with accurate predictions of highly excited vibrations up to nearly 95% of the dissociation threshold.

A new spectroscopic model is developed for theoretical predictions of vibration-rotation line positions and line intensities of the methane molecule. Resonance coupling parameters of the effective polyad Hamiltionians were obtained via high-order contact transformations (CT) from ab initio potential energy surface. This allows converging vibrational and rotational levels to the accuracy of best variational calculations. Average discrepancy with centers of 100 reliably assigned experimental bands up to the triacontad range was 0.74 cm(-1) and 0.001 cm(-1) for GS rotational levels up to J = 17 in direct CT calculations without adjustable parameters. A subsequent "fine tuning" of the diagonal parameters allows achieving experimental accuracy for about 5600 Dyad and Pentad line positions, whereas all resonance coupling parameters were held fixed to ab initio values. Dipole transition moment parameters were determined from selected ab initio line strengths previously computed from a dipole moment surface by variational method. New polyad model allows generating a spectral line list for the Dyad and Pentad bands with the accuracy ~10(-3) cm(-1) for line positions combined with ab initio predictions for line intensities. The overall integrated intensity agreement with Hitran-2008 empirical database is of 4.4% for the Dyad and of 1.8% for the Pentad range.

International audience; The state of the art analyses of high-resolution ozone spectra and a critical discussion of line parameters updated since the previous review by Rinsland et al. [J Quant Spectrosc Radiat Transfer 2003;82:207–18] are presented. We focus on recent works revisiting improved measurements and analyses of Fourier transform spectra below 5600 cm−1 and review Cavity Ring-Down Spectroscopy measurements and analyses in the 5850– 7920 cm−1 spectral region which have been performed for 16O3 and all 18O enriched ozone isotopologues. Various issues related to resonance perturbations due to dark states are discussed as well as the issues concerning the accuracy of the experimental parameters, the data modelling and the data reduction. Comparisons of vibration levels determined from spectra analyses with predictions from the molecular potential functions are given. The procedures of empirical corrections for calculated line lists are described and the “intricate” question of absolute intensity determination is overviewed. Also, a review of deficiencies, challenge for improvements, and related works in progress are reported.

An accurate description of the complicated shape of the potential energy surface (PES) and that of the highly excited vibration states is of crucial importance for various unsolved issues in the spectroscopy and dynamics of ozone and remains a challenge for the theory. In this work a new analytical representation is proposed for the PES of the ground electronic state of the ozone molecule in the range covering the main potential well and the transition state towards the dissociation. This model accounts for particular features specific to the ozone PES for large variations of nuclear displacements along the minimum energy path. The impact of the shape of the PES near the transition state (existence of the "reef structure") on vibration energy levels was studied for the first time. The major purpose of this work was to provide accurate theoretical predictions for ozone vibrational band centres at the energy range near the dissociation threshold, which would be helpful for understanding the very complicated high-resolution spectra and its analyses currently in progress. Extended ab initio electronic structure calculations were carried out enabling the determination of the parameters of a minimum energy path PES model resulting in a new set of theoretical vibrational levels of ozone. A comparison with recent high-resolution spectroscopic data on the vibrational levels gives the root-mean-square deviations below 1 cm(-1) for ozone band centres up to 90% of the dissociation energy. New ab initio vibrational predictions represent a significant improvement with respect to all previously available calculations.