The Global NWP Index is a measure of the forecasting skill of the global NWP model over persistence for up to 5 days ahead worldwide.
The Index is calculated in two ways: against observations and against model analyses. Observations are sparse in some parts of the world, and using full gridded analyses greatly increases the sample size.
It is based on global model forecasts of selected parameters out to 5 days ahead for regions covering the whole globe verified by comparison with observations and is based on 36 months of data. A skill score is calculated for each forecast included in the index by normalising the forecast root mean square (rms) error against the persistence rms error. These errors are computed in accordance with the method recommended by WMO's Commission for Basic Systems (CBS). The skill scores for each parameter are then combined to form a single value using weights reflecting the importance of each parameter to Met Office customers.
The index is compiled from the following parameters:
verified over the following areas:
at the following forecast ranges: T+24, 48, 72, 96 and 120 hours.
The forecasts that are used to compute the Index are taken from the normal operational run of the global model that provides products to customers.