RCG

Topic: Downscaling...

Dynamic and statistical downscaling

General circulation models (GCMs) are powerful tools in modern climate research. They present the best possible way to simulate large-scale climate conditions and to project future climate changes due to known forcing effects such as the enhanced greenhouse effect. However, due to their coarse spatial resolution (typically around 300 km), their ability to capture smaller scale climatic features is limited. For practical applications and impact studies, it is local climate variables, such as temperature and precipitation that are of interest. Downscaling refers to a suite of methods by which local climate variables can be obtained from large-scale climate variables, which are usually provided by GCM.

Dynamic downscaling

Essentially there are two types of downscaling methods. The first is called dynamic downscaling which involves nesting a high-resolution limited area climate model to a GCM, while the other takes a statistical approach and is termed statistical downscaling.

Statistical downscaling

Statistical downscaling involves establishment of a connection between large scale climate variables (predictors) and local scale climate variables (predictands) via a statistical model. A schematic diagram showing the process of statistical downscaling is given here.