This paper and corresponding seminar given on 20 September 2010 at the 16th International School for Quantum
Electronics in Nesebar, Bulgaria, will describe the key hardware aspects of the Raman-shifted Eye-safe Aerosol
Lidar (REAL) and recent advances in extracting two-component wind vector fields from the images it produces.
The REAL is an eye-safe, ground-based, scanning, elastic aerosol backscatter lidar operating at 1.54 microns
wavelength. Operation at this wavelength offers several advantages compared to other laser wavelengths including:
(1) maximum eye-safety, (2) invisible beam, (3) superior performance photodetectors compared with those
used at longer wavelengths, (4) low atmospheric molecular scattering when compared with operation at shorter
wavelengths, (5) good aerosol backscattering, (6) atmospheric transparency, and (7) availability of optical and
photonic components used in the modern telecommunations industry. A key issue for creating a high-performance
direct-detection lidar at 1.5 microns is the use of InGaAs avalanche photodetectors that have active areas of at
most 200 microns in diameter. The small active area imposes a maximum limitation on the field-of-view of the
receiver (about 0.54 mrad full-angle for REAL). As a result, a key requirement is a transmitter that can produce
a pulsed (>10 Hz) beam with low divergence (<0.25 mrad full-angle), high pulse-energy (>150 mJ), and short
pulse-duration (<10 ns). The REAL achieves this by use of a commercially-available flashlamp-pumped Nd:YAG
laser and a custom high-pressure methane gas cell for wavelength shifting via stimulated Raman scattering. The
atmospheric aerosol features in the images that REAL produces can be tracked to infer horizontal wind vectors.
The method of tracking macroscopic aerosol features has an advantage over Doppler lidars in that two components
of motion can be sensed. (Doppler lidars can sense only the radial component of flow.) Two-component
velocity estimation is done by computing two-dimensional cross-correlation functions (CCFs) and noting the
displacement of the peak of the CCF with respect to the origin. Motion vectors derived from this method are
compared with coincident sonic anemometer measurements at 1.6 km range. Preliminary results indicate the
method performs best when the atmosphere is stable with light winds.
|