Estimation of the parameters of a turbulent drifting medium by a least squares analysis of spaced sensor measurements

A technique is outlined for determining the average structure and drift velocity of a turbulent medium by applying the least squares method to correlograms from two-dimensional sensor measurements. A general spatiotemporal correlation function is defined from the average temporal autocorrelogram for all sensors. The technique could easily be extended to three dimensions. This approach avoids errors introduced by the choice of an analytical function and would seem to have significant advantages over other correlation analysis techniques, because it easily handles the case of a large number of sensors, and the statistical nature of the method makes it possible to determine accurately the confidence that can be placed on the estimated parameters. An analytical correlation function (Gaussian) and the present method were applied to spaced radio frequency sensor measurements of ionospherically reflected radio waves and the results compared. In a few cases, close agreement was obtained between the two methods, showing that in these cases the analytical correlation function was a good fit to the data. This shows that the present method can be readily used to test the goodness of fit of selected analytical functions. In many cases, formally acceptable solutions were not obtained, indicating that the spatial and temporal correlation functions of the medium were dissimilar.