pyart.util.estimate_noise_hs74#

pyart.util.estimate_noise_hs74(spectrum, navg=1, nnoise_min=1)[source]#

Estimate noise parameters of a Doppler spectrum.

Use the method of estimating the noise level in Doppler spectra outlined by Hildebrand and Sehkon, 1974.

Parameters:

  • spectrum (array like) – Doppler spectrum in linear units.

  • navg (int, optional) – The number of spectral bins over which a moving average has been taken. Corresponds to the p variable from equation 9 of the article. The default value of 1 is appropiate when no moving average has been applied to the spectrum.

  • nnoise_min (int) – Minimum number of noise samples to consider the estimation valid

Returns:

  • mean (float-like) – Mean of points in the spectrum identified as noise.

  • threshold (float-like) – Threshold separating noise from signal. The point in the spectrum with this value or below should be considered as noise, above this value signal. It is possible that all points in the spectrum are identified as noise. If a peak is required for moment calculation then the point with this value should be considered as signal.

  • var (float-like) – Variance of the points in the spectrum identified as noise.

  • nnoise (int) – Number of noise points in the spectrum.

References

P. H. Hildebrand and R. S. Sekhon, Objective Determination of the Noise Level in Doppler Spectra. Journal of Applied Meteorology, 1974, 13, 808-811.