Global Models

The largest scale for numerical weather prediction is the global or world scale, where the entire atmosphere around the Earth is simulated. There are two major forecast models using the global scale which are the Global Forecast System (or GFS) and the European Centre for Medium-Range Weather Forecasts (or ECMWF, sometimes just called the “Euro”). The GFS is produced by the United States National Weather Service, whereas the Euro model is produced by an international organization of European atmospheric scientists. Both the GFS and the Euro are typically run out to at least 10 days in the future. However, because of the chaotic nature of the atmosphere, model forecasts on the global scale become unreliable beyond seven days in the future.

The most useful aspect of global weather models is understanding large scale weather patterns and how they will evolve on the scale of 1-7 days. For example, a global model would be most useful for understanding the evolution of the strength and location of the upper-level jet stream, whereas it may not be as useful for understanding small details about where individual storms may occur. While model outputs from the GFS and the Euro include information about precipitation, this data is parameterized, meaning that the precipitation is not explicitly resolved by the model, but an algorithm guesses at what the precipitation amount would be. This is in contrast to convection-allowing models (or CAMs), which explicitly resolve individual thunderstorms and therefore have a more accurate representation of where and when those thunderstorms will form.