Techniques for the spatial downscaling of rain for application in SHF propagation modelling

There is current interest in the use of numerical weather prediction (NWP) models for propagation modelling for radio system design. However, NWP models currently operate with spatial and temporal resolutions much lower than those ideally required for modelling on typical terrestrial and satellite fixed links. There are fundamental problems when predicting the behaviour of links, possibly only a few hundred metres long, from parameters such as rain rate averaged over voxels with linear dimensions of ten kilometres or more. Although improvements in NWP models could reduce the grid size to the order of a kilometre, this is likely to be computationally very expensive. An alternative approach is to utilize spatial downscaling techniques to introduce small scale variations into the model fields based upon its larger scale behaviour. We describe the downscaling process and its integration into the network simulator.