ABSTRACT
Lidar measurements of temperature for the upper troposphere and lower stratosphere are commonly derived by the Raman technique. Lidar signals derived from vibrational Raman processes have been subjected to numerous simulation tests to examine their sensitivity to the presence of aerosols and ozone in the atmosphere. The influence of aerosols characteristics (wavelength dependence of aerosol extinction and particle phase function) and of ozone concentration on Raman temperature profiles is estimated. Simulations indicate large temperature deviations for post-volcanic conditions. For a Raman backscatter at 607 nm, bias is below 1 K for a total optical depth less than 9 x 10(-3) in the case of a stratospheric contamination and less than 6 x 10(-3) for a tropospheric contamination. The effect of aerosols depends on phase function and a few parameters such as altitude, optical depth and the shape of the high-altitude cloud. The wavelength dependence of aerosol extinction has some influence only for severe post-volcanic conditions (Scattering Ratio, SR >2). For a Raman backscatter at 387 nm, bias is larger and can be significant even in background aerosol conditions. Changes in the ozone density profile lead to significant Raman temperature deviations only for some specific conditions. Results suggest that both aerosol and ozone corrections are necessary to obtain an accuracy better than the 1 K requested for most atmospheric applications.
