Stochastic Maximum Likelihood (SML) parametric estimation of overlapped Doppler echoes

摘要

This paper investigates the area of overlapped echo data processing. In suchcases, classical methods, such as Fourier-like techniques or pulse pairmethods, fail to estimate the first three spectral moments of the echoesbecause of their lack of resolution. A promising method, based on amodelization of the covariance matrix of the time series and on a StochasticMaximum Likelihood (SML) estimation of the parameters of interest, has beenrecently introduced in literature. This method has been tested onsimulations and on few spectra from actual data but no exhaustiveinvestigation of the SML algorithm has been conducted on actual data: this paperfills this gap. The radar data came from the thunderstorm campaign thattook place at the National Astronomy and Ionospheric Center (NAIC) inArecibo, Puerto Rico, in 1998.