How spaghetti models and cones help you refine your prediction

A spaghetti model plot maps a dozen or more weather model runs, showing the potential track and intensity of a given storm over the course of about a week. All the lines plotted together look like… well, a bunch of spaghetti thrown at the map, hence the name.

These graphs are very useful to meteorologists because they show a full range of possible outcomes. If the lines are close together and all bend in the same direction, this can give forecasters more confidence that the models are pointing them in the right direction.

On the other hand, if the lines are all mixed up and pointing in different directions, it means that there is serious disagreement between each individual model and that there is great uncertainty about a storm’s future track.

Forecast cones represent the margin of error

Another mainstay of hurricane forecasts is the cone of uncertainty. Think of these forecast cones as the margin of error in predicting a storm. The cone relates to the path of the storm’s center.

NHC cone of uncertainty

The USA National Hurricane Center (NHC) reviews their track predictions at the end of each season, calculating their mean error at each time step. A 12-hour forecast is more accurate than a 120-hour forecast, and therefore the cone grows larger with time.

Historically, the center of a tropical cyclone remains within the cone of uncertainty two-thirds of the time. This means that, based on the average accuracy of the NHC over the past few seasons, the storm will likely move in the general direction the forecast cone points.