To develop new methods of producing climate change precipitation
scenarios and demonstrate their application to a range of
hydrological systems in order to provide impact assessments,
estimates of uncertainty and validation of the methodology.


There are 3 goals; 1) to build a framework for the production
of realistic, long duration daily precipitation scenarios; 2)
to downscale the scenarios to finer time resolution and
disaggregate spatially; and 3) to make assessments and
uncertainty estimates of the hydrological impacts of climate
change for four sensitive European river basins.

Two alternative methods of deriving future climate daily
precipitation from GCM variables other than rainfall will be
developed. A weather typing procedure will be used to derive
future rainfall scenarios, making use of present day
precipitation-climate statistics and GCM predictions of future
weather types. A second technique will use the
precipitation-temperature dependence relation and the GCM
prediction of temperature change. The variability of observed
and GCM climates over periods of 1-20 years and its effect on
precipitation will be analysed and used to condition the
scenarios in order to attempt the correct simulation of droughts
and extreme events.

Two techniques for downscaling to finer resolution will be used.
The first is the Neyman-Scott Rectangular Pulses model, which
will be developed using different types of rain cell to model
different weather types and extended spatially to produce
multi-site rainfall series. The second is the Modified Turning
Bands model, which will be developed for spatial and temporal
downscaling to any required resolution using model parameters
defined by reference to observed statistics associated with the
analysed daily weather type.

The impacts of climate change on water resources and hydrological
regimes will be assessed by applications of a number of
hydrological modelling systems to a range of European river
basins. The modelling systems will include both research and
engineering packages, applied both at fine and coarse spatial
resolution.