Science Inventory

NONLINEAR-APPROXIMATION TECHNIQUE FOR DETERMINING VERTICAL OZONE-CONCENTRATION PROFILES WITH A DIFFERENTIAL-ABSORPTION LIDAR

Citation:

Kovalev, V., M. Bristow, AND J. Mcelroy. NONLINEAR-APPROXIMATION TECHNIQUE FOR DETERMINING VERTICAL OZONE-CONCENTRATION PROFILES WITH A DIFFERENTIAL-ABSORPTION LIDAR. Applied Optics 35(24):4803-4811, (1996).

Description:

A new technique is presented for the retrieval of ozone concentration profiles from backscattered signals obtained by a multi-wavelength differential-absorption lidar (DIAL). The technique makes it possible to reduce erroneous local fluctuations induced in the ozone-concentration profiles by signal noise and other phenomena such as aerosol inhomogeneity. Before the O3 profiles are derived, the dominant measurement errors are estimated and uncertainty boundaries for the measured profiles are established. The off-to online signal ratio is transformed into an intermediate function, and analysis approximations of the function are then determined. The separation of low- and high-frequency constituents of the measured ozone profile is made by the application of different approximation fits to appropriate intermediate functions. The low-frequency constituents are approximated with a low-order polynomial fit, whereas the high-frequency constituents are approximated with a trigonometric fit. The latter fit makes it possible to correct the measured 03 profiles in zones of large ozone concentration gradients where the low-order polynomial fit is found to be insufficient. Application of this technique to experimental data obtained in the lower troposphere shows that erroneous fluctuations induced in the ozone-concentration profile by signal noise and aerosol inhomogeneity undergo a significant reduction in comparison with the results from the convential technique based on straightforward numerical differentiation.

Record Details:

Record Type:DOCUMENT( JOURNAL/ PEER REVIEWED JOURNAL)
Product Published Date:01/01/1996
Record Last Revised:12/22/2005
Record ID: 9162