Tropospheric temperature measurements using a rotational raman lidar

Abstract

Using the Hampton University (HU) Mie and Raman lidar, tropospheric temperature profiles were inferred from lidar measurements of anti-Stokes rotational Raman (RR) backscattered laser light from atmospheric nitrogen and oxygen molecules. The molecules were excited by 354.7 nanometer (nm) laser light emitted by the HU lidar. Averaged over 60-minute intervals, RR backscattered signals were detected in narrow 353.35 nm and 354.20 nm spectral bands with full-widths-at-half-maxima (FWHM) of 0.3 nm. During the special April 19-30, 2012, Ground-Based Remote Atmospheric Sounding Program (GRASP) campaign, the lidar temperature calibration coefficients were empirically derived using linear least squares and second order polynomial analyses of the lidar backscattered RR signals and of reference temperature profiles, obtained from radiosondes. The GRASP radiosondes were launched within 400 meters of the HU lidar site. Lidar derived temperature profiles were obtained at altitudes from the surface to over 18 kilometers (km) at night, and up to 5 km during the day. Using coefficients generated from least squares analyses, nighttime profiles were found to agree with profiles from reference radiosonde measurements within 3 K, at altitudes between 4 km and 9 km. Coefficients generated from the second order analyses yielded profiles which agreed with the reference profiles within 1 K uncertainty level in the 4 km to 10 km altitude region. Using profiles from GRASP radiosondes, the spatial and temporal homogeneities of the atmosphere, over HU, were estimated at the 1.5 K level within a 10 km radius of HU, and for observational periods approaching 3 hours. Theoretical calibration coefficients were derived from the optical and physical properties of the HU RR lidar and from the spectroscopic properties of atmospheric molecular nitrogen and oxygen. The theoretical coefficients along with lidar measurements of sky background radiances were used to evaluate the temporal stability of the empirically derived temperature profiles from the RR lidar measurements. The evaluations revealed systematic drifts in the coefficients. Frequent reference radiosonde temperature profiles should be used to correct for the drifts in the coefficients.

For the first time, the cause of the coefficient drifts has been identified as the differences in the aging of the spectral responses of the HU lidar detector pairs. For the first time, the use of lidar sky background measurements was demonstrated as a useful technique to correct for the coefficient drift. This research should advance the derivations of lidar temperature calibration coefficients which can be used for long observational periods of temperature fields without the need for frequent lidar calibrations using radiosondes.