Time SeriesIntermediate

Fan Chart

A forecast visualization where expanding shaded bands radiate from a central line — the wider the fan, the greater the uncertainty about what lies ahead.

// 01 — The chart

What it looks like

Example — GDP growth forecast2020 – 2028
Today20202022202420262028

A fan chart showing GDP growth. The solid line is historical data; the dashed line is the central forecast; shaded bands widen to show increasing uncertainty.

// 02 — Definition

What is a fan chart?

A fan chart is a forecast visualization that shows a central prediction (usually a line) surrounded by symmetrically expanding bands. Each band represents a different probability interval — for example, 50%, 75%, and 90% confidence.

The shape resembles an opening hand fan: narrow at the present (where we have certainty) and wide in the future (where we don’t). This visual metaphor makes uncertainty intuitively understandable even to non-statisticians.

Fan charts were popularised by the Bank of England in 1996 for their Inflation Report. Today they are widely used in economics, climate science, and epidemiology.

Key insight: A fan chart is one of the few chart types specifically designed to communicate what we don’t know, not just what we do.

// 03 — Anatomy

Parts of a fan chart

ABC
A — Forecast origin: The point where historical data ends and the forecast begins
B — Confidence bands: Nested shaded areas representing probability intervals (e.g. 50%, 75%, 90%)
C — Central forecast: The dashed line showing the most likely (median) prediction

// 04 — Usage

When to use it — and when not to

✓Use a fan chart when…
  • Communicating forecasts with quantified uncertainty
  • You want decision-makers to see the range of possible outcomes
  • Showing how confidence decreases over longer forecast horizons
  • Comparing the width of uncertainty across different scenarios
  • Economic, climate, or epidemiological projections
×Avoid a fan chart when…
  • You don't have probability distributions — only a single point forecast
  • The audience is unfamiliar with probability intervals and would misinterpret the bands
  • You have multiple overlapping forecasts — the fans become an unreadable mess
  • Historical data is more important than the forecast — use a standard line chart
  • Uncertainty is asymmetric and you haven't modelled it properly

// 05 — Reading guide

How to read a fan chart

1

Find the dividing line

Identify where historical data (solid line) ends and the forecast (dashed line) begins.

2

Read the central forecast

The dashed line is the single most likely prediction — the median or mode of the distribution.

3

Interpret the bands

Each shaded band is a probability interval. The darkest band is the narrowest (e.g. 50%); lighter bands are wider (e.g. 90%).

4

Assess uncertainty growth

Watch how quickly the fan opens. A rapidly widening fan means high forecast uncertainty.

5

Check for asymmetry

If the bands are wider on one side, the forecast has skewed risk — more downside (or upside) than the central line suggests.

// 06 — Data format

What data you need

For each future time point: the central forecast and upper/lower bounds for each probability interval (e.g. 10th, 25th, 75th, 90th percentiles). Historical data is a simple time series.

// 07 — Construction

How to build one

Step 1: Plot the historical time series as a solid line.

Step 2: From the forecast origin, plot the central forecast as a dashed line.

Step 3: Draw shaded bands from the widest interval outward, layering narrower (darker) bands on top.

Step 4: Add a legend mapping each band to its probability interval. Mark the forecast origin clearly.

// 08 — Common mistakes

Mistakes to avoid

No legend for bands

Readers can't interpret the bands without knowing which probability interval each represents.

Too many bands

More than 3–4 nested bands become indistinguishable. Stick to 50%, 75%, and 90% intervals.

Hiding the historical line

Context matters. Always show the historical data so readers can judge forecast plausibility.

Implying false precision

Fan charts should convey ranges. If you also show a precise point forecast, readers focus on that and ignore the uncertainty.

// 09 — Real-world examples

Where you’ll see them

Bank of England Inflation Report

The original fan chart — showing CPI inflation forecasts with nested probability bands.

Climate projections (IPCC)

Temperature and sea-level rise projections use fan-style uncertainty bands across RCP scenarios.

Epidemiological forecasts

COVID-19 case projections from the CDC and academic groups used fan charts to show forecast ranges.

// 10 — Quick reference

Key facts

Also known asConfidence fan, uncertainty cone
CategoryTime series
DifficultyIntermediate
Popularised byBank of England (1996)
Best forSingle-variable forecasts with quantified uncertainty
Key trade-offShows uncertainty range but can overwhelm with overlapping bands

// 11 — Accessibility

Making it accessible

Fan charts use opacity gradients that can be hard for low-vision users. Use distinct opacity steps rather than continuous gradients. Provide ARIA labels summarising the central forecast and the range at key time points. Offer a data table with forecast intervals as a fallback.

// 12 — Variations

Common variations

Asymmetric fan chart

Bands are wider on one side, showing skewed forecast risk (e.g. more downside).

Spaghetti plot

Instead of bands, individual simulation runs are drawn as thin lines — giving a richer sense of path uncertainty.

Cone of uncertainty

Common in hurricane tracking — a single widening outline rather than nested bands.

Quantile dot plot

Dots at each forecast step show the distribution — more intuitive for lay audiences.

// 13 — FAQs

Frequently asked questions

What is a fan chart?+

A fan chart is a forecast visualization that shows a central prediction (usually a line) surrounded by symmetrically expanding bands. Each band represents a different probability interval — for example, 50%, 75%, and 90% confidence.

When should you use a fan chart?+

Use a fan chart when communicating forecasts with quantified uncertainty. It also works well when you want decision-makers to see the range of possible outcomes, and when showing how confidence decreases over longer forecast horizons.

When should you avoid a fan chart?+

Avoid a fan chart when you don't have probability distributions — only a single point forecast. It is also a poor fit when the audience is unfamiliar with probability intervals and would misinterpret the bands, or when you have multiple overlapping forecasts — the fans become an unreadable mess.

What data do you need to make a fan chart?+

For each future time point: the central forecast and upper/lower bounds for each probability interval (e.g. 10th, 25th, 75th, 90th percentiles). Historical data is a simple time series.

How is a fan chart different from an area chart?+

Both a fan chart and an area chart can look similar at first glance, but they answer different questions. Reach for a fan chart when the comparisons and patterns it was designed to reveal match what you need to communicate, and choose an area chart when its particular strengths better fit your data and audience.

What is another name for a fan chart?+

Fan Chart is also known as Confidence fan, uncertainty cone. The name varies between fields, but the visualisation technique is the same.

What size of dataset works best for a fan chart?+

Fan Chart works best for Single-variable forecasts with quantified uncertainty. Outside that range the chart either looks empty or becomes too cluttered to read clearly.

Are fan charts accessible to screen readers?+

Fan charts use opacity gradients that can be hard for low-vision users. Use distinct opacity steps rather than continuous gradients. Provide ARIA labels summarising the central forecast and the range at key time points. Offer a data table with forecast intervals as a fallback.