Appl. Phys. B 66, 105–113 (1998) Applied Physics B Lasers and Optics  Springer-Verlag 1998 Raman-shifted laser sources suitable for differential–absorption lidar measurements of ozone in the troposphere M.J.T. Milton 1 , G. Ancellet 2 , A. Apituley 3 , J. Bösenberg 4 , W. Carnuth 5 , F. Castagnoli 6 , T. Trickl 5 , H. Edner 7 , L. Stefanutti 6 , T. Schaberl 4 , A. Sunesson 3 , C. Weitkamp 8 1 Division of Quantum Metrology, National Physical Laboratory, Teddington, Middlesex TW11 0LW, UK (Fax: +44-181/943-6755, E-mail: martin.milton@npl.co.uk) 2 Service d’Aeronomie du CNRS, BP No3, 91375, Verrieres le Buisson, France 3 Rijksinstituut voor Volksgezondheid en Milieuhygiene, PO Box 1, 3720 BA, Bilthoven, The Netherlands 4 Max Planck Institut für Meteorologie, Bundesstrasse 55, D-2000 Hamburg, Germany 5 Fraunhofer Institut für Atmosphärische Umweltforschung (IFU), Kreuzeckbahnstrasse 19, D-8100, Garmisch-Partenkirchen, Germany 6 Consiglio Nazionale delle Ricerche Istituto di Ricerca sulle Onde Elettromagnetiche, via Panciatichi 64, 50127, Firenze, Italy 7 Department of Physics, Lund Institute of Technology, P0 Box 118, S-22100 Lund, Sweden 8 Institut fur Physik, GKSS Forschungszentrum GmbH, Postfach 1160, 2054 Geesthacht, Germany Received: 21 January 1997/Revised version: 20 May 1997 Abstract. Measurements have been performed at eight laboratories on Raman-shifted radiation from frequency- quadrupled Nd:YAG and KrF lasers. Each of these sources is a possible candidate for making differential–absorption lidar measurements of ozone in the free troposphere. The spectral purity, efficiency and divergence of the different sources are compared and explained using a simplified model of the Ra- man interaction. Particular attention has been given to the use of buffer gases to optimize the Raman gain and divergence. The experimental results show that the operating conditions required to optimize the output energy of each of these types of laser source can be quite different. In the case of the KrF laser, it has been shown that high energies can be generated at low active gas pressures, and that increasing the pressure increases the beam quality, but decreases the energy. PACS: 42.65; 42.55 Differential-absorption lidar has been used widely for the measurement of ozone concentrations in the troposphere and the stratosphere [1]. These measurements have been made with fixed and with mobile ground-based systems and also with systems working on airborne platforms [2, 3]. The ob- jective of this work, conducted as part of the TESLAS [4] sub-project of the EUROTRAC [5] collaborative programme, has been to compare different approaches to building accurate and cost-effective differential–absorption lidar systems for measuring ozone in the boundary layer and free troposphere. This paper reviews work carried out to evaluate the perfor- mance of different laser sources based on Raman shifting. The project has also included an intercomparison of meas- urements performed with different systems based on these sources under field conditions [6–9]. The performance achievable with a differential–absorption lidar system is largely determined by the laser source used. This paper describes measurements performed at eight lab- oratories which show how wavelengths in the spectral re- gion used for measuring ozone can be generated by ‘Raman shifting’ the output of either a KrF laser (at 248.5 nm) or a frequency-quadrupled Nd:YAG laser (at 266.2 nm). This approach is an alternative to the use of a continuously tunable source such as a frequency-doubled dye laser. It has several possible advantages including a high conversion efficiency from the primary laser source which is particularly important for applications in mobile and transportable systems where power inputs are often limited. Additionally, the wavelength accuracy, stability and spectral purity of the radiation gen- erated is governed by the characteristics of the pump laser which, under favourable conditions, can be very good. How- ever, there are also disadvantages, for example, the quality of the generated beam can often be difficult to control and the performance achievable with any particular laser source is difficult to predict in detail. In this paper, we discuss how the pressure of the Raman- active medium can be adjusted to optimize the conversion efficiency and beam quality of a particular Raman interac- tion. In particular, the addition of buffer gases to the active gas to produce the optimum conditions has been investigated. Our work shows that the exact performance obtained is crit- ically dependent on the energy, pulse length, spectral purity and beam quality of the pump laser. We have also observed the presence of unwanted Raman scattering from rotation- al transitions, and describe methods for reducing the energy scattered in this way. 1 Measurement of ozone by differential-absorption lidar The most convenient spectral region for monitoring ozone in the troposphere with the differential–absorption technique is between 250 and 300 nm [1, 10], where the Hartley band has a broad and strong absorption cross-section [11]. The structureless nature of the absorption spectrum of ozone in this region has a number of consequences for differential– absorption lidar measurements. The most important of these is that the on- and off-resonant wavelengths are usually spaced a minimum of 2 nm apart in order to achieve sufficient