Prediction of surface refractivity gradient distributions, from weather station surface data

The availability of fixed radio links, especially at latitudes below 40 degrees, can be severely affected by sub-refractive events that may only be present for an hour or two each year, or around the 99.98% point of the refractivity gradient distribution. This tail point is beyond the highest percentile currently available from world-wide meteorological model predictions. In many countries radiosonde data is too sparsely distributed for useful accuracy, so a more densely distributed data source is required. This suggests the use of surface meteorological data, which is generally available from many automatic weather stations at good time resolution. This paper reviews empirical models for surface refractivity gradient cumulative distribution that use surface meteorological data, and presents a new model, which uses a new variable, surface refractivity anomaly.