Statistical evaluation of electric field and neutral wind results from beam-swing incoherent scatter measurements

When ionospheric electric field and neutral wind are determined from ion velocities given by monostatic beam swing incoherent scatter measurements, a stationary and horizontally homogeneous ionosphere is assumed. These assumptions are not necessarily valid during a single beam cycle or within the whole measurement region. Disturbances in the receiver may also cause errors. Thus the results may contain errors which are not of statistical nature. If most of the data come from regions where temporal and spatial variations are small, more reliable results are expected, if observations which violate the basic assumptions are rejected. This paper shows a means of finding such measurements. Electric fields and neutral winds with their standard deviations are first determined from measured ion velocities, and these results are used backwards to calculate reconstructed ion velocities and their standard deviations. Then the differences between the measured velocities and the reconstructed ones as well as the standard deviations of these differences are computed. If a given difference lies beyond a selected statistical limit, the corresponding measurement is regarded as unreliable and is rejected. New results are obtained by repeating the analysis with the cleaned data set. The process is repeated until either all remaining measurements are statistically reliable or the determination of the parameters fails. In the latter case, there will be a gap in the results. The paper presents the underlying statistical theory, the statistical evaluation of the results, and demonstrates the use of the method with incoherent scatter data. Key Points Improved reliability of electric fields and neutral winds from IS data Method for finding statistically unreliable velocities in IS radar data Application to EISCAT measurements