Skip to Main Content
The development of a stochastic spatial downscaling algorithm, able to enhance the spatial resolution of rain or attenuation fields considered as input, is illustrated. The proposed algorithm takes advantages from the spatial characterization of the rain fields derived from numerical weather prediction models (NWPM) simulations at coarse spatial resolution. The associated attenuation field at frequencies higher than X band are derived from a numerical scattering simulator based on the T-matrix approach. The stochastic spatial downscaling technique will be described and used to provide an ensemble of possible realizations at a spatial resolution higher than that available at the output of the MM5 simulations which has been taken, for our purpose, as the reference NWPM. The developed methodology will be applied and discussed for some case studies where MM5 outputs, rain-gauge data and ground-based radar observations were available.