The feasibility of using the optimal Markovian filtration in sensing of the troposphere at altitudes up to 3 km by the DIAL method is illustrated on the basis of a stochastic model of the altitude behaviour of temperature fluctuations smoothed by lidar pulses. The Calman-Bucy algorithms for optimal estimate of the fluctuating profiles of temperature and its variance are synthesized. Their efficiency is analyzed by a numerical simulation technique as applied to sensing in the line of absorption of the A–band of oxygen at the transition Pp27 centered at wavelength 768.3802 nm.