Rain attenuation measurements: Variability and data quality assessment

Publisher Website
Google Scholar
Abstract
Year‐to‐year variations in the cumulative distributions of rain rate or rain attenuation are evident in any of the published measurements for a single propagation path which span a period of several years of observation. These variations must be described by models for the prediction of rain attenuation statistics. Now that a large measurement data base has been assembled by the International Radio Consultative Committee (CCIR), the information needed to assess variability is available. On the basis of 317 sample cumulative distribution functions (cdf's) for the occurrence of attenuation by rain, the expected year‐to‐year variation in attenuation at a fixed probability level in the 0.001–1.0% of a year range is estimated to be 26%. The expected deviation from the predictions of any of the better attenuation models for a single year of observations is estimated to exceed 34% when a rain climate model is employed to estimate the rain rate cdf. The probability distribution for the variation in attenuation or rain rate at a fixed probability level is lognormal. The lognormal behavior of the deviations was used to compile statistics for variability, to establish hypothesis tests for identifying outliers—the observed sample cdf that deviates significantly from the expected (modeled) cdf, and to predict the attenuation level to be exceeded with a specified return period.