Performance of estimated Doppler velocity by maximum likelihood based on covariance matrix

Abstract

This paper investigates the efficient estimator of echo data processing to clean the spectrum through the denoising process. The maximum likelihood based on covariance matrix (MLCM) method without <italic<a priori</italic< knowledge of the spectral width is proposed for denoising the atmospheric signal. This method is applied to simulated and actual data to estimate the spectrum parameters. The probability density function of estimators as an empirical model is used to describe the performance of the estimators. The MLCM method is suggested to be an alternate estimator to precisely obtain the essential spectrum parameters with a lower standard deviation of good estimators and a larger detected range, which is improved by 20%, compared with the maximum likelihood method with <italic<a priori</italic< knowledge of the spectral width. Moreover, it can reduce the large velocity volatility and the uncertainties of the spectral width in the low signal-to-noise ratio regime. The MLCM method can be applied to obtain the whole wind profiling by the coherent Doppler lidar.