Its main role is to produce forecast models by gathering all the information from weather satellites in space and observations on earth.

Sea ice began to be initialized in forecast models in 1971.

Three years later, the first global forecast model was introduced.

The activity of hindcasting involves running a forecast model after an event has happened in order to test whether the model's simulation is valid.

Obtaining integrated data on these relationships over a long-term period is crucial to improving forecast models and resource management for environmental change.

Use of forecast models can help determine the magnitude and pattern of the rainfall expected.

The low was poorly defined but global forecast models were anticipating cyclogenesis within the next two days.

Looking at a single forecast model gives no indication of how likely that forecast is to be correct.

Research and computer processing abilities increased over the years, which allowed for the first global forecast model to run by June 1966.

The process starts with weather data analysis input into a forecast model.