READi Insights: Downscaling

Global climate models (GCMs) are an important tool for projecting future climate conditions under various greenhouse gas emissions scenarios. Although they are of key use for planners, their size and complexity often make them evaluate the climate at coarse spatial scales (generally around 100 km) and temporal scales (generally daily), whereas most weather and climate impact assessments require much more localized and granular information. Because of this, scientists and practitioners often convert the coarse data into finer scale through a set of methods called downscaling. This READi Insight briefly summarizes the main forms of downscaling—statistical, dynamical, and machine learning methods—and highlighting their relative advantages and disadvantages.