Laser spectroscopic study of ozone in the 100

Laser spectroscopic study of ozone in the 100← 000 band for the SWIFT instrument M. Guinet, C. Janssen, D. Mondelain, C. Camy-Peyret LPMAA, CNRS- UPMC (France) 18/06/2010

Importance of ozone in terrestrial atmosphere • In the stratosphere: UV filter for solar radiation • Key role in tropospheric chemistry (OH radical precursor) • Green house gas Studied by a large palette of instruments (UV spectrophotometer, Dobson spectrometer, FTIR…) 1% uncertainty for intensity is required for atmospheric applications (reactive gas!) Comparison of published data sets → several % inconsistencies for intensities (10 µm bands)

Stratospheric Wind Interferometer For Transport studies (SWIFT) - O 3: tracer for atmospheric winds (SWIFT instrument) - Stratospheric wind speed (Doppler shift) & ozone concentration measurements - Understand the transport of O 3 in the stratosphere High accuracies required (for the 15 strong 16 O 3 transitions) - Absolute positions: 5 10 -5 cm-1 ( 13 m s-1) - Absolute intensities: 1% Other spectroscopic parameters measured: air, self

A set up at the border of metrology and spectroscopy Stabilization principle Stabilized He. Ne / = 10 -8 Laser Diode 633 nm / < 4. 10 -8 Michelson interferometer 8. 9 µm

A set up at the border of metrology and spectroscopy Stabilization principle Stabilized He. Ne / / = 10 -8 Laser Diode 633 nm / < 4. 10 -8 Michelson interferometer 8. 9 µm

A set up at the border of metrology and spectroscopy Stabilization principle Stabilized He. Ne / / = 10 -8 Laser Diode 633 nm / < 4. 10 -8 Michelson interferometer 8. 9 µm

A set up at the border of metrology and spectroscopy Stabilization principle Stabilized He. Ne / / = 10 -8 Laser Diode 633 nm / < 4. 10 -8 Michelson interferometer 8. 9 µm

A set up at the border of metrology and spectroscopy Stabilization principle Stabilized He. Ne / / = 10 -8 Laser Diode 633 nm / < 4. 10 -8 Michelson interferometer 8. 9 µm Both highly tunable and stabilized system Ultra flexible setup (atmospheric windows accessible) Amplitude modulation scheme good S/N (several thousand)

Stabilized spectrometer N 2 O Laser Diode D (FP) Interferometer locked onto stabilized He. Ne laser D O 3 Generation system I 0 N 2 O O 3 UV PT 100 C UV D D O 3 Generation Step by step mode Step < 10 -4 cm-1 S/N : 3000 Accuracy (2 ): Position < 8. 10 -5 cm-1 Intensity < 2% FP Laser Diode M. Guinet, D. Mondelain, C. Janssen, C. Camy. Peyret, JQSRT, 111, 961 -972 (2010)

Spectra 10 000 points SWIFT Linearization Absolute calibration

Metrological approach Reduction and taking into account of systematic biases : Sample purity, spectrometer, experimental conditions… Traceability : Following the BIPM recommendation (photometer UV). Calibrated tools (PT 100, micrometer, pressure gauge, stabilized He. Ne). Mass spectrometer Expertise : O 3 sample purity test (> 99. 5 % purity sample) : IR spectrometer (CO 2, H 2 O, N 2 O) mass spectrometer (N 2, NOx) pressure (O 2, N 2, non condensable gases), Stable conditions (temperature, very low ozone decomposition 2 - 4 ‰ / hour) Checking : Check BIPM UV recommended cross section (Hearn) with a calibrated pressure gauge. Hearn A. G. , Proc. Phys. Soc. , 78 (1961) 932 -940 C. Janssen and M. Guinet, RSI. , submitted

High accuracy absorption measurement of O 3 cross section at 253. 65 nm International standard (BIPM) : = 1. 147 10 -17 cm² (± 2. 1%) our measures : = 1. 131 10 -17 cm² (± 0. 7%) M. Guinet, C. Camy-Peyret, D. Mondelain and C. Janssen, Refinement of the ozone standard – absolute ozone absorption cross section at the mercury emission line position 253. 65 nm, Metrol. , en préparation

Results on intensities • Estimated uncertainties between 0. 8 and 1. 3% considering: – statistical error over the 11 spectra with pure O 3 – systematic errors (T, offset, UV, LIR…) Comparison with HITRAN 08: -2. 2 1. 1(2 ) % (Our UV cross section) -2. 6 1. 3(2 ) (Mauersberger cross section)

Results on intensities IR UV • Cross cell : UV and IR measured simultaneously 3. 6% inconsistencies between UV (Hearn) and IR (HITRAN 08) recommended values in agreement with Picquet-Varrault et al, J. Phys. Chem A, 109 (2005) 1008 -1014.

Absolute positions of strong O 3 lines • The N 2 O line positions were accurately measured by an heterodyne experiment (Maki and Wells) • Linearization and calibration procedure applied to O 3 and N 2 O spectra N 2 O positions → overall accuracy of 8 10 -5 cm-1 (2 ) Wind speed uncertainty: ~ 20 m s-1 O 3 positions → Mean difference with HITRAN 08: (5 8 (2 )) 10 -5 cm-1 Maki A. G. , Wells J. S. NIST Special publication (1991) 821

Determination of the self pressure broadening Pressure broadening at 2 % accuracy level (2 ) Agrees with HITRAN by 0. 6% with Smith by 1. 7% (Voigt profile) HITRAN [cm-1] • Ultra high resolution spectra (200 – 1000 points / line) • Voigt and Rautian-Sobel'man (hard) line profile • Multi-fit procedure 1132. 59927 1132. 60340 1132. 65698 1132. 78600 1132. 81144 1133. 43351 1133. 58691 1133. 63170 1133. 67120 1133. 72456 1133. 97863 1134. 02880 1134. 25147 1134. 45383 1134. 50970 HITRAN This work [cm-1 atm-1] (Voigt) 0. 1031 0. 1035 0. 0964 0. 0963 0. 1023 0. 1033 0. 0948 0. 0958 0. 1017 0. 0954 0. 1032 0. 0952 0. 1010 0. 10372(15) 0. 1042 0. 09594(8) 0. 0969 0. 10160(11) 0. 10706(6) 0. 10415(15) 0. 09630(9) 0. 09331(11) 0. 10289(12) 0. 0923 0. 09736(8) 0. 10675(12) 0. 09272(9) 0. 10169(8) This work - HITRAN [cm-1 atm-1] [%] (Hard) 0. 60 0. 10540(17) -0. 48 0. 09660(7) -0. 68 3. 64 0. 82 1. 58 -2. 60 1. 17 0. 10222(12) 0. 10753(6) 0. 10488(15) 0. 09678(8) 0. 09362(11) 0. 10370(12) 2. 06 3. 44 -2. 61 0. 68 0. 09777(9)

Air-pressure broadening coefficients 8 absorption spectra recorded with the crossed UV-IR cell and a 50 m astigmatic cell (O 3 decomposition <1% / hour) HITRAN [cm-1] air [cm-1 atm-1] n air 1132. 59927 0. 07902(4) 1132. 60340 0. 0853 1132. 65698 0. 07171(1) 0. 71 1132. 78600 0. 06844(1) 0. 72 1132. 81144 0. 07703(2) 0. 48 1133. 43351 0. 08432(3) 0. 51 1133. 58691 0. 0752 1133. 63170 0. 06994(1) 0. 92 1133. 67120 0. 06808(1) 0. 88 1133. 72456 0. 07493(1) 0. 68 1133. 97863 0. 07136(2) 0. 77 1134. 02880 0. 07034(1) 0. 76 1134. 25147 0. 08347(5) 1134. 45383 0. 06720(1) 1134. 50970 0. 07422(1) Astigmatic cell

Air Pressure Shift Temperature dependence: Determination of air and air at 240 K → n air and n air HITRAN [cm-1] air [cm-1 atm-1] (This work) air ’ 105 ’ 105 [cm-1 atm-1] [cm-1 atm-1 K (Smith) -1] Temperature regulated cell [cm-1 atm-1 K-1] (Smith) (This work) 1132. 59927 Voigt profile 1132. 60340 1132. 65698 -0. 0017(1) 1132. 78600 -0. 0010(1) 1132. 81144 -0. 0004(1) 1133. 43351 -0. 0014(1) 1133. 58691 1. 8 1. 2 -0. 0004(7) -2. 8(10) 0. 5(3) 1133. 63170 -0. 0002(1) 0. 0007(2) 1133. 67120 -0. 0011(1) -0. 0012(1) 1133. 72456 -0. 0005(1) -0. 0012(2) -0. 6(3) 1133. 97863 -0. 0004(1) -0. 0002(3) 0. 1(6) 1134. 02880 -0. 0016(1) -0. 0009(1) 0. 6(2) 1134. 25147 -0. 0018(2) -0. 0009(3) -0. 4(4) 1134. 45383 -0. 0004(1) -0. 0008(1) 1. 7(2) 0. 0005(2) 3. 3(4) 1134. 50970 1. 4 0. 4(2) Smith M. A. H. , Malati Devi V. , Benner D. C. , Rinsland C. P. , J. Mol. Spectrosc. 182 (1997) 239 -259.

Conclusion 253 nm UV cross section determination Our IR measurement are 2. 2 % higher than HITRAN 08 3. 6% inconsistencies between UV (Hearn) and IR (HITRAN 08) recommended values Position in agreement with HITRAN 08 Measurement of temperature dependence of air broadening and air shifting.

Laser spectroscopic study of ozone in the 100← 000 band for the SWIFT instrument M. Guinet, C. Janssen, D. Mondelain, C. Camy-Peyret LPMAA, CNRS- UPMC (France) 18/06/2010

3. 7 10 -5 cm-1 Jitter of the Laser diode The line is used like a frequency/amplitude noise converter

Stratospheric Wind Interferometer For Transport studies (SWIFT) Limb view Stratospheric wind velocity Doppler effect Intensity O 3 à 1133, 4 cm-l 5. 10 -5 cm-1 Wave number 13 m. s-1

Repeatability of the determination of the laser line shape We fit the instrument’s apparatus function like a Voigt profile are fit on the spectrum Parameter SD intensity 9‰ lorentz ± 9‰ The intensity, lorentz (2 10 -4 cm-1) , gaussian (4 10 -4 cm-1) Intensity Set of 32 (on two day) measure in a sealed N 2 O cell 4% 5% gaussian Source of bias ‰ Effect on the intensity mean value ( ± 1. 6 ‰) lorentz fix to it’s mean value 0. 7‰ gaussian fix to it’s mean value Exemple of TDL Line Shape : 0. 3 ‰ 4% on lorentz ± 4% lorentz 3‰ 5% on gaussian 1. 5‰ J. Reid, D. T. Cassidy, and R. T. Menzies. Linewidth measurements of tunable diode lasers using heterodyne and etalon techniques November 1982 / Vol. 21, No. 21 / APPLIED OPTICS

Stabilized spectrometer UV O 3 Generation system I 0 N 2 O FP Accuracy (1 ): Position < 4. 10 -5 cm-1 Intensity < 1% Laser Diode