Weather Insider: Forecasting limitations

There are a number of tools and types of forecasts meteorologists use every day. However, forecasts are never perfect. In general, the further out a forecast goes, the more fine details are lost. This is because the complex equations and programs that create our weather models are time-dependent. The programs essentially act like a giant game of mathematical telephone.

There are two main types of forecast models we use, deterministic and ensemble. Deterministic forecasts use a single set of starting conditions and use all of the computing power to run that one outcome. This is what you'll see on most forecasting websites. Ensemble models simulate a bunch of different starting conditions to see a general trend of what will happen with a system. A good example of ensemble forecasts is the famous spaghetti plots used to help forecast hurricanes. An example can be seen above.