On the derivation of rain anisotropy from Olympus beacon measurements

A method for the analysis of anisotropy and canting angle is applied to long-term dual-polarisation measurements using the Olympus beacons. Using this method, the following conclusions have been drawn: (a) for the calculation of anisotropy, it is preferred not to use the crosspolar phase, but instead make certain assumptions concerning the relation between the differential attenuation (A) and phase shift (β). This is because the crosspolar phase is mostly not measured accurately enough to make these calculations. (b) Transferring the canting angle, calculated from every data point at 20 GHz to the calculations for the two other frequencies (12.5 and 30 GHz), does not give good results, because of the addition of uncorrelated noise contributions. Assuming a constant canting angle appears to give more consistent results. (c) A contour plot is a more effective way to represent a joint distribution than the more often used scatterplot and 3D plot. Using this representation, separate clusters in the distribution of A and β versus copolar attenuation are clearly recognised