Lidar signal denoising methods- application to NARL Rayleigh lidar

Abstract

Lidar (Light Detection And Ranging) is a remote sensing tool of great practical importance in environmental monitoring sciences. As an active remote sensing instrument, lidar provides vertical profiles of aerosol layers and atmospheric temperatures. The effective range and data reliability of lidar is often limited by various noises. Signal processing for lidar applications involves highly nonlinear models and consequently nonlinear filtering as the backscattered signal follows log-linear form. Denoising of the signal is essential for reducing random unwanted variations of the signal, in order to get the significance of the signal as much as possible. In this work, various denoising methods are applied to the signal from Rayleigh receiver of lidar at National Atmospheric Research Laboratory (NARL), Gadanki (13.8°N, 79.2°E) near Tirupati, India. Denoising methods such as Moving Average method, Wavelet and Empirical Mode Decomposition (EMD) method are considered and compared in this work. The Signal to Noise Ratio (SNR), temperature profiles and statistical standard temperature errors are obtained using denoised signals. It is found that EMD gives better SNRs than other denoising methods. The statistical standard temperature error obtained using EMD denoised signal and the original signal are compared and the EMD method is found to reduce temperature errors at higher ranges better than conventional method.