Drawing Sankey diagrams on a world map to show flows between different geographical location is always a challenge. One of the inherent problems is that large (=broad) arrows may run between two points on the map located very close to each other. Another problem is that one wishes to have the arrows more or less along the actual trade routes, which in many cases is close to impossible (take, for example, ships going through the Panama or the Suez canal).

I have shown quite a number of ‘Sankey diagram maps’ here on the blog, but most of them had shortcomings. Now here is an example that does extremely well in tackling the issue of Sankey diagram flows on a world map.

(CC licence – Carbon Brief)

This flow map of coal exports around the world shows the top exporters. It was crafted by Rosamund Pearce for the Carbon Brief article “Mapped: The global coal trade”. She decided to route the Sankey arrows nicely sorted, in parallel, and not along the actual shipping paths. See how much of the coal trade from Indonesia to China is led “virtually” south of Australia and New Zealand? Additionally the arrows are not led precisely to the actual port, but rather connect at a suitable place of each continent. With these simplifications the trade flow map is much clearer, understandable.

Another world coal flow diagram from 2012 can be found here.

The ‘Blog Audytorów Sledczych’ had two interesting Sankey-style diagrams.

The first is the trade balance of Poland for the month January 2014. Flows are in Euro (trade volume). There is a trade surplus of 176,4 million Euros. Unfortunately only the legend shows what the arrow colors mean, but it is not legible in this screenshot.

The second one focuses on the five main trade partners of Poland. Here the value of imported goods is higher (8,7 billion Euro) than the value of exported goods (8,5 billion Euro).

Again the legend is somewhat difficult to decipher, but we can see Germany, Czech Republic, UK and Russia. Largest trade partner (both regaring import and export) is Germany (large orange arrows).

Trade imbalance is visually barely noticeable, just by a slight step when looking at the incoming and outgoing arrows.

The 2012 GEA Global Energy Assessment report (GEA Global Energy Assessment – Toward a Sustainable Future, Cambridge University Press, Cambridge UK and New York, NY, USA and the International Institute for Applied Systems Analysis, Laxenburg, Austria) features five maps showing energy trade in the world on pages 128/129.

These can almost be considered Sankey diagrams, so I am featuring them here on the blog.

This one is for embodied energy in trade goods.

And this one is a classic oil and oil product trade flows map:

Flows lead from a specific color-coded region to another. The quantities are clustered in arrows with three different widths as shown in the legend (1, 5, 10 Exajoule)

Blog reader Panalion sent me a photo taken in Amsterdam’s Botanical Garden. It is of a map showing coffee and tea flows from producing countries to mainly Europe and North America. Panalion writes “I thought you might like this Sankey map I found attached to a cable between two palm trees. There were chairs set up to accomodate school classes”.

This map is for didactic purposes and features no absolute figures and no year. In addition to the export flows of coffee and tea shown as arrows the map also has circles of three different sizes representing percentage of world production of coffee, tea and cocoa in the originating country.

Infographers might have better ways of showing this information. But in this case I think it is sufficient to get the message across to the target audience, the school kids.

Blog reader Johannes send me a note and suggested to feature the below diagram. Thanks for that.

Tony Hirst from OUseful.info created it after seeing a map-based diagram for horse meat trade flows on the Guardian Data Blog. Tony used Mike Bostock’s D3s Sankey Plugin that allows creating this type of diagrams directly from data in Excel/CSV files. In his post he describes how he proceeded to build this Intra EU Horse Meat Trade diagram. Somewhat techie, but nevertheless makes an interesting read.

Overall trade quantity was more than 60.000 tonnes in 2012. Largest exporters are Belgium and Poland (left side), largest importers are Italy and France (right side). Data is from Eurostats.

The above is a only a static picture, but you can go here to play around with the interactive version. Data labels and quantities are available in the interactive version when you hover ths mouse over certain bands. You can also move the nodes up and down vertically and group the countries differently.

This special type of Sankey diagram is also refered to as distribution diagram and (…hate to say it in light of the current scandal) a Spaghetti diagram. Fineo and Parsets (see software list) can also be used for this type of diagrams where statistical data is grouped into categories (here: exporting and importing countries) and bands/streams/spaghettis are shown between the categories to represent the relationships between them.

Here is my May 2012 post on distribution diagrams with d3.js.

I am back after a few weeks of holiday. To get into posting again, here is a quick one I found at Graphic Design Forum in a discussion thread on software for creating trade flows (in this case oil flows) on a world map.

Flows in million tonnes (per year?). Scale element at the bottom left. No visible arrow direction, but instead a blue to green gradient on each band (blue for export, green for import). Middle East region being the largest exporter remains a problem with a very wide Sankey arrow leaving through the Indian Ocean.

A notice on scoop.it/visualdata led me to this fascinating video on visualnews. It shows the making of an infographic in two minutes or 3657 frames and is by Jess Bachmann for mint.com.

The central element of the infographic is a Sankey diagram on the trade flows between the United States and China (and to/from other countries).

it is interesting to see how Jess did every weighted arrow as a brush line with rounded head (the heads are neatly hidden behind the country maps, or capped at the other end). Each horizontal, vertical and curved segment is done individually.

In the YouTube comments of the long version of this video the author replied to one commenter: “After determining a metric, i.e 1 pixel width = $1M, I then stroked a line with the corresponding size brush. A $34M item would have a 34px width line. At one point you can even see a calculator popping up (0:55 into the video).

The long (7 minute) version has a lot more details on how the infographic comes to life. You can even see that Jess keeps saving his work from time to time…

Wow, what a hell lot of work – but the result sure looks gorgeous.

I calculated that Jess took more than 10h to complete this: 3657 frames, ten seconds between each frame = 36570 sec, 3600 seconds to an hour, makes 10.16 hours! I am just glad I have my Sankey diagramming software, so at least I don’t have to bother about brush sizes.

Just back from a holiday, and in order not to keep you waiting for new Sankey diagams … shuffle, shuffle, draw … here are two more one from the Mondays with Minard series at the Cartographia blog (see previous post).

The first one shows French wine exports in 1864. A lot of the good stuff goes to European neighbours, the U.S., to Brasil, Uruguay and Argentina [the latter today an exporter of great wines themself]. But there are also some gourmants that appreciate ‘un verre de bon rouge’ in remote places such as India, China, and of course the outer French territories (Mauritius and Reunion). Not sure what the unit of flow is, as the image is to small to read.

The second one is a map of France that shows the transport of wine and spirits in 1857.

Most of the good stuff is shipped to Paris along rivers (green) and major railway lines (pink). Road transport seems to have been excluded. You can also see the transports to the ports for exports (yellow). Flows are in tonnes, but the band itself shows transports in both directions. This national map could be seen as an inset into the above map.