| Summary: | This work examines the application of selected estimation methods to path integrated direct detection CO2 lidar data, with the objective of improving the precision in the estimates of the log power, and log power ratios. Particular emphasis is given to the optimal estimation techniques of Kalman filtering theory, and to the consequent requirements for system and measurement model identification. A dual wavelength system was designed and constructed, employing two hybridised TEA lasers, a co-axial transceiver, and direct detection. Over a period of several months, a database of differential absorption measurements was accumulated, each consisting of 10,000 dual wavelength lidar returns. Various wavelength pairs were used, including those recommended for the monitoring of H2O, CO2, NH3 and C2H4. A subset of this database is used to evaluate the above mentioned estimation methods. The results are compared with simulated data files in which it was possible to control precisely process models which are believed to form an approximation to the real processes latent in the actual lidar data.
|
|---|
