Statistical downscaling: A powerful tool for estimating global change impacts on regional scale

Although General circulation models (GCMs) simulate climate change at large spatial scales very well, they perform poorly at the smaller space relevant to regional impact analyses. So-called `downscaling´ techniques have subsequently emerged as a means of translating information from coarse to fine resolution. This article reviews the most popular downscaling tool named statistics downscaling methods under three main headings: regression methods, artificial neutral networks (ANN) techniques and empirical orthogonal function (EOF) analysis. Statistics downscaling methods can provide local scale specific information in many global change impacts studies with less computer calculation and expenditure than other methods, which are very helpful for studies in regional scale.