Application of the CEEMD algorithm in LIDAR for denoising and parameter detection

RISTORI P; BALI L; GOMEZ J; OTERO L; PAPANDREA S; SALVADOR J; QUEL E

Workshop; VIII Workshop on Lidar Measurements in Latin America; 2015

This work intends to adapt the Complete Ensemble Empirical Mode Decomposition (CEEMD) algorithm to analyze lidar profiles in presence of aerosol and clouds. This data-driven and noise-assisted decomposition is applicable to non-linear and non-stationary process and is able to decompose the lidar signal into a sequence of intrinsic mode functions (IMFs). The first ones captures the shortest signal oscillations of the lidar then the denoising process entails the removal of these IMFs from the original data. Clouds and aerosol detection is achieved by inspecting these higher order IMFs and involves a second methodology where this signal decomposition highlights relevant signal fluctuations associated to important atmospheric parameters. The results presented shows that the CEEMD method is a good candidate for lidar denoising as well as for atmospheric features detection.