Predicting radio refractivity using a stochastic model

The radio refractivity of the atmosphere is a very important variable in radio communication. This paper attempts to model the radio refractivity over a tropical station (Akure, Southwestern Nigeria) with the aim of predicting future values. Atmospheric variables (temperature, pressure and relative humidity) were obtained over a period of eighteen months using a Davis Pro weather station. Differencing of the data was carried out to ensure stationarity. Suitable lags were obtained using the Box-Jenkins approach. Based on the model parameters, predictions were made. The predictions were found to be good for a period of about 11 days into the future.