Nonparametric estimation of mean velocity and spectral width in weather radar

Proposes a new nonparametric approach to the estimation of the mean Doppler velocity (first spectral moment) and the spectral width (square root of the second spectral centered moment) of a zero-mean stationary complex Gaussian process immersed in independent additive white Gaussian noise. By assuming that the power spectral density of the underlying process is bandlimited, the exact maximum likelihood estimates of its spectral moments are derived. An estimate based on the sample covariances is also studied. Both methods are robust in the sense that they do not rely on any assumption concerning the power spectral density (besides being bandlimited). Under weak conditions, the estimates based on sample covariances are best asymptotically normal.