An optimal-inversion method to lidar data analysis
Wellington Carlos de Jesus1, Eduardo Landulfo1, Haroldo Fraga de Campos Velho2
1IPEN - Instituto de Pesquisas Energéticas e Nucleares
2INPE - Instituto Nacional de Pesquisas Espaciais
This work suggests a new Lidar algorithm to extract/estimate the aerosol backscatter and extinction coefficients and respectively, using the Lidar equation and an optimal-inversion method. In order to get satisfactory inversion data, the boundary conditions must be estimated when analyzing the measurements conditions and the recorded signals instead of being assumed a priori. Some of the most common inversion methods applied to lidar sensing studies are being improved by means of Bayesian inference together with a Gaussian statistic, creating a method where the most probable or the optimal solution corresponds to the maximum probability or pdf (probability density function). Then the measurement data linked to their covariances and the a priori considerations are used to obtain an atmospherical parameters vector. The new lidar algorithm presents some new advantages when compared to the other ones:
1) Possibility of incorporating multiple heterogeneous sources, an additional wavelength for instance. The algorithm uses aerosol optical thickness information from AERONET (NASA Aerosol Robotic Network) as additional information;
2) The analyzed data can vary over the irradiated region or time. For example, the atmosphere during the daytime presents different characteristics from the nighttime. The algorithm can process different kinds and amounts of information;
3) Make the error diagnosis possible, including analyzes and optimization of error components that are present directly on method. Yet, allows clearer and more confident measurements. The link used for optimization, as mentioned previously, is the AERONET aerosol optical thickness and more.


