Denoising of Spectral Data Using Complex Wavelets

摘要

Atmospheric signal processing is of interest to many scientists, where there is scope for the development of new and efficient tools for cleaning the spectrum, detection, and estimation of parameters like zonal (U), meridional (V ), wind speed (W), etc. This paper deals with a signal processing technique for the cleansing of spectrum, based on the complex wavelets with custom thresholding, by analyzing the Mesosphere¨CStratosphere¨CTroposphere radar data that are backscattered from the atmosphere at high altitudes and severe weather conditions with low signal-to-noise ratio. The proposed algorithm is self-consistent in detecting wind speeds up to a height of 18 km, in contrast to the existing method which estimates the Doppler manually and fails at higher altitudes. The results have been validated using the Global Positioning System sonde data