Ensemble Prediction Systems (EPS) are numerical weather prediction (NWP) systems that allow us to estimate the uncertainty in a weather forecast as well as the most likely outcome. Instead of running the NWP model once (a deterministic forecast), the model is run many times from very slightly different initial conditions. Often the model physics is also slightly perturbed, and some ensembles use more than one model within the ensemble (multi-model EPS) or the same model but with different combinations of physical parameterization schemes (multi-physics EPS). Owing to the cost of running an NWP model many times, the EPS is normally run at around half the horizontal resolution of the equivalent deterministic NWP model. The EPS normally includes a control forecast that uses the ensemble resolution model but without any perturbations to the analysis or model. The individual NWP solutions that make up the ensemble are often referred to as the ensemble members. The range of different solutions in the forecast allows us to assess the uncertainty in the forecast, and how confident we should be in a deterministic forecast. The uncertainty in a weather forecast can vary widely from day to day according to the synoptic situation, and the EPS approach provides an estimate of this day-to-day uncertainty. The EPS is designed to sample the probability distribution function (pdf) of the forecast, and is often used to produce probability forecasts – to assess the probability that certain outcomes will occur.
The present guidelines are intended to provide some general advice to forecasters and forecast providers on the effective use of EPS, and on what EPS can and cannot be expected to provide. A general working knowledge of the principles and use of NWP is assumed.
Collection(s) and Series: WMO- No. 1091
Language(s): English; Other Languages: French, Russian, Spanish
Format: Digital (Free)