Radio meteorological analysis for forecasting microwave fading occurrence

Simplified regression models based on radio operators´ knowledge and statistical data are proposed for characterizing fading occurrences on microwave radio circuits in mid-northern Japan. The potential validity of the models is examined with the aim of developing fading occurrence forecasts useful for radio telecommunication network management. Typical phenomena experienced by radio operators were used for a preliminary selection of meteorological parameters and for a single-parameter analysis. A multivariable logistic regression model using several parameters such as temperature, vapor pressure, and atmospheric stability or implicit Pasquill index showed good regression performance