Journal oJA~mospher~c and zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Terrestrial Physics, Vol. 49, No. 3, pp. 287-298, 1987. Printed in Great Britain. OX-9169/87 $3.00+ .OO 0 1987 Pergamon Journals Ltd. Upper stratospheric and mesospheric temperatures derived from lidar observations at Aberystwyth D. B. JENKINS, D. P. WAREING, L. THOMASand G. VAUGHAN Department of Physics, University College of Wales, Aberystwyth, Dyfed SY23 3BZ, U.K. (Received in final form I August 1986) Abstract-From lidar observations of relative atmospheric density above Aberystwyth (52.4”N, 4.1”W) upper stratospheric and mesospheric temperatures have been derived for a total of 93 nights between December 1982 and February 1985. Excellent agreement was found between radiances synthesised from these temperatures and those measured by satellite-borne instruments. Summer temperatures showed a smooth and regular variation with altitude and reasonably good agreement with the CIRA (1972) model atmosphere. By contrast, winter temperatures showed a much greater variability with altitude and greater changes from night to night, with the frequent occurrence of a large amplitude wave-like perturbation in the mesosphere with about 15 km vertical wavelength and amplitude about 20K between 60 and 80 km. Pronounced warmings of the stratosphere were observed during the three winters of observation. During the warming event occurring in early February 1983 the stratopause temperature increased to 303K at 43 km, while the major warming event of late December 1984/early January 1985 produced a stratospheric temperature gradient of 16K km-’ between 34 and 36 km. During the latter event a distinct local temperature minimum at 32.6 km was observed on New Year’s Eve, this descending to 29 km by the following night and being accompanied by a lowering of the stratopause from 43 to 38.5 km in the same period. These results demonstrate the ability of the present technique to resolve the high stratopause temperatures and steep stratospheric temperature gradients which occur during stratospheric warmings, in marked contrast to the limited resolution achieved by satellite experiments. 1. INTRODUCTION make the majority of these measurements, with the Lidar (laser radar) systems have been used to derive frequency-tripled wavelength at 355 nm employed on atmospheric temperature by means of a number of some nights to check for the presence of Mie scattering different scattering mechanisms : Rayleigh scattering by aerosols. by the atmosphere above about 30 km (KENT and WRIGHT, 1970; HAUCHECORNE and CHANIN, 1980, 1982, 1983; CHANIN and HAUCHECORNE, 1981); Raman scattering by nitrogen in the 1 l-25 km altitude region (PETTIFER, 1975) ; resonance scattering by sodium atoms between about 80 and 100 km (BLA- MONT et al., 1972; MEGIE et al., 1978; GIBSON et al., 1979). At Aberystwyth (52.4”N, 4.1”W) a lidar system has been used to measure upper stratospheric and meso- spheric temperatures from December 1982 to Feb- ruary 1985. Rayleigh scattering of the laser beam has been employed to measure relative density above about 34 km and a method similar to that described by HAUCHECORNE and CHANIN (1980) used to infer the altitude variation of temperature. Observations are restricted to night-time by the unacceptably high background of scattered sunlight present during the day. The frequency-doubled output at 532 nm of a Q- switched Neodymium-YAG laser has been used to 2. EXPERIMENTAL DETAILS The lidar sysem used has been described by THOMAS et al. (1983). The Neodymium-YAG laser has a pulse repetition frequency of 10 Hz and a mean output of 3 W at 532 nm and 1 W at 355 nm. The transmitter and receiver beams had widths less than 0.8 and about 1 mrad, respectively, and were virtually coaxial with both directed vertically. The detector was an EM1 98 13 QKBM photomultiplier operated in the photon counting mode. The signal was recorded in 2047 chan- nels each of 500 ns duration corresponding to a range resolution of 75 m up to 153 km, the counts from successive pulses being added to reduce the statistical uncertainty. The duration of observations shown for individual nights represents the effective integration period. The derivation of atmospheric temperature from the recorded photon counts is described in Appendix 287