Probabilistic
The Operational Models Encyclopedia has moved to a NEW Google Site:
Please update any bookmarks to match the new site.
The modeler accepts the scientific limitations on NWP and creates an Ensemble Forecast System (EFS), using initial condition sensitivity, model estimates and errors, or both, to tease out potential forecast outcomes. Each member of the ensemble has a unique representation in the range or “envelope” of outcomes. Using the range of outcomes and after removing systematic bias, we can estimate the ensemble mean, spread, probability distributions, and extreme values. EFSs are especially useful in forecasting the probability of high-impact events, and clustering of members around two or more forecast outcomes.
This encyclopedia section provides the methods used to account for both initial condition sensitivity and NWP model approximations and errors for the EFSs produced by most of the world’s meteorological centers. Some of the important details for operational forecasters to consider are sampling of potential initial conditions and the accounting for model approximations.
Use the navigation bar at left to choose your probabilistic model.