The Cone of Uncertainty

*Cone of Uncertainty (3-day)*

The forecast product most often used to convey the track forecast positions and the associated uncertainties is the aptly-named "cone of uncertainty."

Also known as the "error cone" and the "two-thirds probability cone," the cone of uncertainty was designed as a way to express the uncertainty visually. The uncertainty in the motion of storms is quantifiable: the uncertainty displayed in the cone is a statistical expression of the uncertainty in the forecast. The width of each section of the cone is scaled to represent two-thirds of the track forecast errors for each of the forecast periods from the past five hurricane seasons, hence "two-thirds probability cone." It means though that one-third of those past errors would lie beyond the boundaries of a present-day cone.

The cone of uncertainty comes in two varieties: the 3-day and the 5-day, belonging to the 3-day and 5-day track forecasts, respectively. It has two main purposes. The first is to communicate the spatial forecast coordinates (latitude and longitude) in a way that is user friendly to the average person. The second is to convey the scale of error that is currently taken to be inevitable given the current state of the science. Inherent in the cone's structure is the portrayal of uncertainty not only in a storm's future position but also in its (changing) forward speed.

All weather forecasts by necessity contain uncertainty. The more distant the forecast period (e.g., the 96h or 4-day forecast as compared to the 12h forecast), the larger the uncertainties generally are; this is why the cone is "cone-shaped." In portraying this uncertainty in a graphic and releasing it to the public, forecasters are implicitly advising people as to the present-day capabilities of the science. The cone itself contains information about where the science is in any given hurricane season, while the changes in the widths of its sections from one season to the next constitute a visual record of the rate at which it is progressing.

If the cone is an expression of where the science now stands, it is equally a portrait of how far it has to go. If it were possible to predict a storm's track exactly, the cone would disappear altogether; a single point on a map of some coastline indicating the exact spot of landfall would be all that might remain, with some precise hour and minute in the future time-stamped on it. The text advisory for this kind of forecast might contain but a single line, for instance: LANDFALL WILL OCCUR AT SOMECITY, GULF COAST TEXAS, FRI 2:03 P.M CDT. Such a forecast will never be possible, except in the realm of science fiction; the best that can be hoped for in a real-world future is an uncertainty cone that becomes progressively narrower with every passing hurricane season.

Next Page - Cone Construction
Previous Page - Track Forecast Interpretation
Back to Tutorial Home

Printer-friendly page