Premium
Characterization of auroral radar power spectra and autocorrelation functions
Author(s) -
Jackel Brian J.
Publication year - 2000
Publication title -
radio science
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.371
H-Index - 84
eISSN - 1944-799X
pISSN - 0048-6604
DOI - 10.1029/1998rs002133
Subject(s) - autocorrelation , spectral density , maximum entropy spectral estimation , spectral line , radar , parametric statistics , doppler effect , computational physics , autocorrelation technique , backscatter (email) , physics , gaussian , power (physics) , statistical physics , mathematics , computer science , statistics , telecommunications , principle of maximum entropy , quantum mechanics , astronomy , wireless
Radar backscatter is a commonly used tool for studying plasma instabilities in the auroral E region. Analysis of the received signals typically involves moments of the scattered power spectrum such as total power, mean Doppler shift, and spectral width; in some cases the spectral asymmetry may also be of interest. This paper presents the steps required to estimate spectral moments directly from the autocorrelation function, and some advantages and limitations of working in the lag domain are discussed. Recent measurements of auroral spectra at UHF (440 and 933 MHz) are used to motivate the discussion and as test cases. The utility of parametric models is also studied with an emphasis on determining whether spectra are more nearly Gaussian or Lorentzian. A model autocorrelation function is introduced, with spectral characteristics similar to a Voigt distribution but a more convenient functional form.