A local model for the prediction of rain-rate statistics for rain-attenuation models

A new local model for the prediction of rain-rate statistics is presented. Rain-rate statistics are needed for rain-attenuation and rain-interference prediction models. The new model uses 30-year or longer climate data sets to provide parameters for a closed-form rain-rate probability distribution model. The required climate data are available for the USA and its territories. The model provides for the prediction of annual and monthly distributions and for the expected year-to-year variations in these distributions. The model was tested against empirical rain-rate distributions obtained from long term (five or more year) observation programs in the USA and Canada. It provides the only model available for predicting monthly or seasonal probability distributions. It provides a solution to the worst month distribution estimation problem because distributions may be predicted for each of the months in a year and the worst one is then readily apparent. The model provides monthly and annual distribution predictions that match the observed distributions within the expected uncertainty produced by the intrinsic interannual variations in rain occurrence.