HF radars: Multifrequency study of refraction effects and localization of scattering

Monostatic HF radars take advantage of the refraction of radio waves in the ionosphere to achieve perpendicularity between the wave vector and the Earth's magnetic field at high latitude. The location of the backscattering region therefore depends on the horizontal and vertical ionospheric electron density profiles. Multifrequency radar data are used to infer estimates of differences in propagation path introduced by refraction without knowledge of the ionospheric plasma distribution. Because the plasma velocity due to the electric field mapped from the magnetosphere is a specific property of the ionosphere, the Doppler velocity measured on the same field line but at different probing frequencies should be identical. Differences in the radial velocity versus group delay profiles obtained at different frequencies can be attributed to differences in the propagation path. These effects are quantified by computing the correlation function between velocity profiles obtained at different frequencies. The shift corresponding to the maximum of the correlation function is a direct measure of the relative delay introduced by differences in ionospheric propagation between the two correlated frequencies. A statistical study has been conducted for several frequencies pairs using data from the Systeme HF d'Etude Radar Polaires et Aurorales radar. The statistical differences in group path are respectively 20 km for the 11/12-MHz pair and 60 km for the 11/14-MHz pair. The observed distribution of propagation path shifts have also been modeled with a ray-tracing program. It has been shown that the observed experimental values are overestimating the shift. Computed differences, for example, 10 km for the 11/14-Mhz pair, are lower than experimental ones. The discrepancy is shown to be a consequence of experimental conditions. Finally, errors on the location of scatters made by assuming straight-line propagation and fixed altitude are quantified. With a judicious choice of the scattering altitude, the maximum error is of the order of 20 km.