(1964). Cicogna et al. (2000) studied the 1991–1997 hourly rainfall
measured by the OSMER–ARPA mesonet stations. More recently,
Manzato (2007) has studied the climatology of 6-hours (hereafter
6-h) accumulated rain in FVG (but only on the plains), in relation to
the values of sounding-derived instability indices. Hence, the first aim
of the present study is to provide an updated and spatially extended
climatology of the maximum 6-h rainfall in FVG.
The basic idea of the numerical model downscaling (for a review see
Kim et al., 1984 or Wilby and Wigley, 1997) is that the General Circulation
models (hereafter GC) have good skill when forecasting the synoptic
fields, while that skill is much lower when forecasting “local” fields
which vary greatly on small spatial scales. The aim of the model downscaling
is to fill this gap with a statistical relation between the well
forecasted general circulation pattern and the associated local meteorological
effects. The simplest approach is to fit a historical dataset with a
multilinear regression between some general circulation predictors and
the local predictand. When developing the statistical relation, if the historical
predictors are real observations (as, for example, sounding data),
then the downscaling is called “Perfect Prognosis” (e. g. Klein and Lewis,