Studies of polar mesosphere summer echoes with the EISCAT VHF and UHF radars: Information contained in the spectral shape

The nature of PMSE in the VHF and UHF frequency range is considered taking into account the shape of corresponding Doppler spectra. Assuming a turbulence-based model of PMSE it is argued that for cases where a VHF radar detects strong PMSE, the UHF radar could either detect enhanced coherent scattering caused by the same physical process as in the VHF (i.e., turbulence with large charged ice particles), there could be incoherent scattering modified by the charged ice particles, or there could be a mixture of both. In order to distinguish these cases a simple but robust method is introduced to characterize the shape of the Doppler spectra derived from observations at both frequencies. Spectral shapes arc quantified with one simple fitting parameter of a generalized fit to the auto-correlation function (=Fourier transform of the Doppler spectrum). This parameter takes a value of I for a Lorentzian spectrum indicative of pure incoherent scatter from the D-region, a value of 2 for coherent scatter owing to turbulence, and a value of less than I for incoherent scatter modified by the presence of charged aerosol particles. This method is applicable to observations at altitudes between similar to 70 and similar to 90 km. Simultaneous observations with the EISCAT VHF and UHF radar are presented in which all three cases mentioned above are identified. For the case of incoherent scatter modified by the presence of charged aerosol particles we quantify the radius of the involved ice particles to exceed similar to 5 nm. Most importantly, however, for the case where the UHF-signal exceeded the incoherent scatter signal significantly, the spectrum revealed a clear Gaussian shape indicative of a coherent scattering process with identical spectral width as for the VHF-observations. This finding gives strong support that both echoes are created by the same turbulence-based mechanism and not by different mechanisms as speculated by several previous authors. (C) 2009 COSPAR. Published by Elsevier Ltd. All rights reserved.