Noise Reduction in Lidar Signal Based on Sparse Difference Method

Abstract

Lidar is the only remote sensing device used to measure the dynamic properties of the atmosphere from stratosphere through the mesosphere. The range of Rayleigh lidar is affected due to different noises present in the atmosphere. In this paper, different types of noises in lidar signal are interpreted, and distinct denoising methods such as wavelets, Empirical Mode Decomposition (EMD) and Sparsity are tested on signal received from Rayleigh lidar receiver at National Atmospheric Research Laboratory (NARL), Gadanki. The proposed denoising using sparsity achieves better signal-to-noise ratio at higher altitudes, and the temperature profile also matches good with the SABER instrument in TIMED satellite and NRLMSISE-00 model data.