Climate change is a complex global issue, requiring simple communication about its effects at the local scale. This set of visualisations highlight how we have witnessed temperatures change across the globe over the past century or more. The colour of each stripe represents the temperature of a single year, ordered from the earliest available data at each location to now. All other superfluous information is removed so that the changes in temperature are seen simply and undeniably.
Annual global temperatures from 1850-2017
The colour scale represents the change in global temperatures covering 1.35°C [data]
In the UK, the Central England temperature timeseries stretches back to 1772 with observations every single day. For the wider UK, a set of stripes for different locations is also available, as used by the i newspaper.
Annual temperatures in central England from 1772-2017
The colour scale goes from 7.6°C (dark blue) to 10.8°C (dark red) [data]
Annual temperatures for the contiguous USA from 1895-2017
The colour scale goes from 50.2°F (dark blue) to 55.0°F (dark red) [data]
This stripe concept was adopted by CBC News to communicate climate change in several Canadian cities, including Toronto.
Annual temperatures in Toronto from 1841-2017
The colour scale goes from 5.5°C (dark blue) to 11.0°C (dark red) [data]
Annual temperatures in Germany from 1881-2017
The colour scale goes from 6.6°C (dark blue) to 10.3°C (dark red)
Annual temperatures in Vienna from 1775-2017 (appeared on front page of Weiner Zeitung)
The colour scale goes from 7.5°C (dark blue) to 12.0°C (dark red) [data]
Annual temperatures for Australia (1910-2017)
The colour scale goes from 20.7°C (dark blue) to 23.0°C (dark red) [data]
Annual temperatures for Switzerland (1864-2017)
The colour scale goes from 3.0°C (dark blue) to 6.5°C (dark red) [data]
Annual temperatures for New Zealand (1909-2017)
The colour scale goes from 11.3°C (dark blue) to 13.4°C (dark red) [data]
More locations coming soon…
124 thoughts on “Warming stripes”
This is really effective. I like the way it’s easy to see trends over all the noise of annual temperatures. It’s a very good way to explain why a cold year a couple of years ago, or a run of hot years 100 years ago is not particularly relevant to climate trends.
The chart seems to have only plotted since 1772-2017 for example. However, this is only the last little bit since the Roman Warm Period, then we have the Medieval Warming Period, then the Little Ice Age and then as partially plotted in the Warming Stripes. Seems to have missed the earlier periods that we kind of know of about 15,000 BC.
Dear Ed, I followed with great interest your publication of the warming strips. Do you have a version for France? If yes, I could broadcast it to French meteorological weather presenters. Could you send me the visual in high resolution so that the French media can use it with your coordinates? With My best regards Christian
Brilliant visualisation! Thank you for that. I would be interested in temperatures in non-built-up areas though. Simply because cities produce and keep the heat with all the concrete and no trees to compensate. This means, while the temperatures are correct, they don’t necessarily show the climate change correctly I think.
Great visualization. Something from Brazil? São Paulo maybe …
I am confused about what information is displayed as the verbal description varies from what I think I see.
Is the neutral (colour) point for each location the average annual temperature for that location over the entire time range for that location or for a subset?
Alternately is it as the verbal description seems to say, how much each year’s average temperature varies from the previous year’s average? Graphically it appears this is not what you are trying to display.
Looking at these bar graphs tells me that from the start to the current date the cities and countries depicted have seen population growth. City sizes have grown and with this growth the amount of heat generated by everything in these cities from industry to people to the parking lots, streets, houses sidewalks. Now if the populations of these cities and countries were to shrink and the infrastructure were also to decline with the population, one would see the temps decline as well.
Hello Ed, I love the warming stripes. Quick question though: the Germany one does not have its data listed. Where does it originate from?
great job and very useful for showing to normal people and it’s easy for them to get the feeling. Would love to see something from the most populated, polluted and burning cities in the world maybe New Delhi or Mumbai?
You should turn these into Ltd edition prints . I’d buy one for Aus and one for the UK.
Amazing job !! When the French warming strip will be available ?
this is a very nice illustration of an unfortunately severe problem…
I have the most minute of editorial comments for the “Annual temperatures in Vienna from 1775-2017” the newspaper should be “Wiener Zeitung” not “Weiner Zeitung”
Southern Africa soon…..esp South Africa / Cape Town?
Hey, great work!! Is there any chance to buy prints or to get print data? We are interested to display those graphs in the period of a thematic week
Also interested – would love to print one out that I could hang in my office 🙂
Here is an R package to make climate stripes with various options that may be interesting to some
The mainpage shows how to install and run the package with examples.
Hi there, that’s a very effective infographic. Apologies as I haven’t read all the comments – but I’d really like to see this as a classic graph with error bars, a line of best fit and some statistical testing to see if there is a significant difference between the temperatures at different periods of times. I assume you have done something of this nature – perhaps something more sophisticated and appropriate than my suggestion. I see from the data link that there is a paper about the uncertainty in these measurements, so it would be good to see uncertainties included. Admittedly it won’t be a very good infographic but will provide confidence to people with scientific training/interest.
Many thanks, Anita
Alexander Radtke takes these warming stripes one step further, and shows the stripes out to 2200 based on the choices we make today. To quote Dr. Brown from Back to the Future, “the future isn’t written yet, no ones’ is, so make sure it’s a good one!”
Since everyone seems to be so sure about their explanation for recent warming, where is the explanation for the consistent, much more dramatic declination of temperatures from 1850 to 1930?
I would like to use the warming stripes as an electronic signature. Is that allowed?
For the global warming stripes I superimposed the temperature values (Berkeleyear.org.data) .
It only gets really bad since the millenium when we stopped pumping soot into the atmosphere. Start pumping again to save the pacific islands
This is the best way to show show climate change I’ve ever seen.
I’m a school teacher and want to use the stripe image as a continuing art project; we start by painting thin pieces of wood according to your stripes and hang them on the wall. Every year we’ll place a new piece with the correct colour and then my question is:
How do we determin the colour based on the annual avg. temperature? I’ve tried to search for a “recipe” where a specific temperature corresponds to a certain colour but haven’t been able to find an answer…
Like all good ideas, and many of the greatest, the sheer simplicity and clarity of this graphic mean it’s a global winner. Where can I buy a badge like American senators are wearing?
Well done in developing your use of colours v temperature to communicate the escallation of global warming over the years.
May I ask to be able to illustrate a paper I am preparing, on the solution to global warming, with one of your illustrations? Due credit to yourself will be made clear.
This is a very effective graphic technique that can be used for other univariate time series. (I tried to do this for global mean temp with water colors a decade ago but have neither training nor talent so it neve went anywhere.) I appreciate the R code. I wonder if anyone has developed code in Stata or SPSS (which my students know more often than R).
It would be helpful to have a color legend to indicate what each color stripe represents.