What is the difference between statistical and dynamical downscaling?
In short: Statistical downscaling derives local climate data from calibrated relationships between GCM output and observations, whereas dynamical downscaling runs a regional climate model using GCM boundary conditions. Statistical methods are generally faster and easier to apply to many models or stations. Dynamical methods produce physically consistent spatial fields but require substantially more computing power.
Statistical downscaling
Statistical approaches learn a transfer relationship during a historical overlap period. They are useful when the desired output is a local time series, but their reliability depends on representative observations and on the future stability of the calibrated relationship.
Dynamical downscaling
Dynamical approaches simulate regional atmospheric processes at finer resolution with a regional climate model. They retain a physics-based representation across a full grid, although they remain affected by model biases and normally need much more setup, storage, and computation.