Effects of filtering on rate of change of rain-induced attenuation

Time series of rain attenuation on a slant path at 11.6 GHz are analyzed with two digital filters to assess their influence on the statistics of the rate of change. It is found that the statistics depend on the measuring apparatus and that a normalization to the median value reduces the dependence, efficiently permitting comparisons with statistics collected with different time resolutions. A cutoff in the rates is found at high attenuation level and the constancy of the log-normal variance found experimentally is discussed.