MeteoModels

Explainer

Spaghetti models, explained

Last updated: 16 November 2026

Every hurricane season, maps covered in coloured lines spread across social media. They are called spaghetti models, and they are one of the most shared and most misread products in weather. This guide explains what the lines are, which ones matter and what the plot can and cannot tell you.

What a spaghetti plot is

A spaghetti plot draws many forecasts of the same thing on one map. For a tropical storm, each line is one forecast of the path the storm’s centre will take over the coming days. Where the lines stay bundled together, the forecasts agree. Where they splay apart, they do not.

“Spaghetti models” is a nickname for the plot, not a type of model. The lines come from the same global and regional models used for everyday forecasting, plus a few built specifically for hurricanes.

Two kinds of spaghetti

Plots that look alike can be built in two different ways, and it is worth knowing which you are looking at.

A multi-model plot shows one line per model: the GFS, ECMWF, UKMET, the Canadian model, hurricane-specific models and several blends. There are usually ten to twenty lines, each labelled with a short code.

An ensemble plot shows the members of a single model’s ensemble: 31 lines from the American GEFS, or 51 from the ECMWF ensemble. Every line comes from the same model started from slightly different conditions. This version shows the range of outcomes that one modelling system considers possible.

The lines are not equal

This is the most important thing to know. A multi-model spaghetti plot draws every model with the same line width, but the models differ enormously in quality.

Early and late models

A full global model run takes hours to finish, so its output arrives after forecasters have to issue their advisory. To have something in hand on time, the previous run’s track is shifted to match the storm’s latest observed position. These adjusted tracks are called early or interpolated models, and they make up most of what appears on a public spaghetti plot.

The practical consequence: many of the lines you see are based on model runs that are six or more hours old. A fresh plot is not necessarily fresh guidance.

Spaghetti plot versus the forecast cone

The cone on an official forecast map is not a spaghetti plot and is not drawn from the model spread. The US National Hurricane Center makes one official track forecast, then draws circles around each forecast position sized so that they would have contained the storm’s centre in about two-thirds of its forecasts over the previous five years. The cone is the outline of those circles.

Two things follow. The cone is the same width for every storm in a given season, whether the models are in tight agreement or badly split. And it describes only the probable path of the centre. It says nothing about the size of the storm, and dangerous wind, surge and rain regularly extend far outside it.

The spaghetti plot is the opposite: it shows the disagreement for this particular storm, but with no official judgement about which lines to believe.

How to read one sensibly

A few habits make spaghetti plots useful instead of alarming.

Spaghetti plots beyond hurricanes

The same idea is used for everyday weather. Forecasters plot one pressure or upper-air contour from every ensemble member to see whether the jet stream pattern is settled, and plot every member’s temperature for one city as a plume. The principle never changes: agreement between forecasts is evidence, and disagreement is information.

That is also what the MeteoModels map does for rain, wind, temperature and pressure. Overlay the global models, switch on the certainty mask, and the bright areas are where the models bundle together, exactly as the tight part of a spaghetti plot does.

Frequently asked questions

What are spaghetti models?

Spaghetti models are maps that draw many forecasts of a storm’s track as separate lines. The nickname comes from how the tangled lines look. They show how much the forecasts agree or disagree.

Which spaghetti model is most accurate?

No single line wins every time. Over many seasons, consensus aids that average several good models have been the most accurate track guidance, followed by the leading global models such as ECMWF and the GFS.

What is the difference between spaghetti models and the cone of uncertainty?

Spaghetti models show individual model or ensemble tracks for one storm. The cone is drawn around the single official forecast using average forecast errors from the past five years, so it is the same size for every storm in a season.

Do spaghetti models show how strong a hurricane will be?

No. They show only the forecast path of the storm’s centre. Intensity, size, storm surge and rainfall are forecast separately.

Why do some spaghetti lines go in a completely different direction?

Outlier lines often come from simple statistical or steering models that are included as baselines, or from a single ensemble member exploring an unlikely scenario. They should carry little weight.

Keep reading

Compare the models on the live map All guides