In the absence of a full deterministic modelling of small-scalerainfall, it is common practice to resort to the use of stochasticdownscaling models to generate ensemble rainfall predictions to beused as inputs to rainfall-runoff models. Here we present anapplication of a novel spatial-temporal downscaling procedurebased on a non-linear transformation of a linearly correlated(gaussian) field. This procedure allows for reproducing thescaling properties (if any) of the rainfall pattern and it can beeasily linked with meteorological forecasts produced by limitedarea meteorological models.
展开▼