Abstract

Atmospheric temperature determination in the upper stratosphere and lower mesosphere (30 - 70 km), as derived from the single wavelength lidar measured atmospheric density profile, is now conducted on a routine basis using a ground based system at the Haute Provence Observatory (44°N, 5°E) (ref. ¹). To extend these measurements to the lower altitude range and down to the tropopause level, the aerosol scattering contribution to the return signal should be subtracted. This implies the use of a multiwavelength lidar in order to take advantage of the differential spectral variation of the Rayleigh: and Mie backscattering coefficients. The main difficulty of the experiment lies in the achievable accuracy on the neutral density profile required to obtain a temperature measurement of geophysical interest. Whereas the statistical error on the backscattered signal can be reduced below 0,1 %, the data reduction and density to temperature conversion will not allow a determination of the temperature profile with an accuracy better than ~ 1K which will prevent the study of short term fluctuations in this temperature distribution induced by the existence of organized or turbulent motions in the atmosphere. In conducting the temperature measurement described hereafter the main objective was to determine its feasibility using a ground based system. However the principal interest of such a measurement is related to the advent of a space borne system as it will provide temperature determination in such areas where the global rawinsonde network does not allow any measurement (oceans, deserts…) (ref. ²).