The study shows that the downscaling and bias-correction of
the RCM is a crucial step when only one climate model is used to
study the impacts of climate change on small basins where many
threshold effects are present. Therefore, the selection of methods
and the treatment of uncertainties have important effects on the
conclusions drawn from the methodology applied, even on annual
or seasonal averages. It is expected that the results would be more
scattered for the extremes.