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.

In 2008, the Cartographia blog started a post series called ‘Monday’s with Minard’. Some people consider Charles Joseph Minard the first to use arrow magnitude in his diagrams to represent quantities. (As a consequence, this means that Sankey diagrams would have to be renamed to Minard diagrams!).

What differentiates Minard maps from Sankey diagrams is that Minard’s fine works always have a geographical relation. The most famous one is his Map of Napoleon’s march to Moscow published in 1869. This “carte figurative des pertes successives en hommes de l’Armée Francaise dans la campagne de Russie 1812-13″ shows number of men (as width of arrows), geographic movement of the troops on the map both for invasion as well as for retreat, as well as time and temperature on a separate scale.

Cartographia blog has some other nice examples, two of which are shown here:

The first shows migration patterns across the globe. Arrows do not have an arrow head but the country of emigration is color coded. The outline of the countries is distorted to accomoadate large flows connected to them. For a detailed description please consider reading the original blog post. This map is similar to the one I showed in this post.

The other is a flow map for wool and cotton for the years 1858 and 1861. “Blue represents cotton and wool from the United States, the orange from British territories in South Asia … One millimeter represents 5,000 tons of cotton or wool.”. As one can see on the 1861 map, cotton imports from Asia have increased dramatically. See the description of the map in the blog post on Cartographia blog.

See all Monday’s with Minard posts here. There has been no activity on the blog since June 2008. I hope to see more of these posts some day.

After my posts on visualizing Rotterdam port’s imports/exports and on Internet traffic maps, I have started to experiment with showing the export quantities and destinations for a certain trade good.

I wanted to do a Saudia-Arabia or Irak oil export Sankey map, but couldn’t find good data. I finally came across this summary on Lybian oil exports, and converted the data from the pie chart Lybian Oil Exports, by Destination, 2006 to a Sankey style export flow diagram.

It was new to me that “Libya has the largest proven oil reserves in Africa” with 41.5 billion barrels, and estimated net exports of 1.525 million barrels per day in 2006.

The underlying map is a crop from a World map found on Wikicommons. I think it could be a little more transparent though…

Last weekend I had the possibility to visit a friend in the Netherlands, and we took a tour of Rotterdam Port. Despite the bad weather, I was fascinated by the huge container ships, the cranes, the noises….

Back home I did some research and came up with the cargo data for the year 2005 from the Port of Rotterdam website.

I did the following three Sankey diagrams. The first shows the inbound cargo quantities (in million tons gross weight of cargo) from the left, and the outbound quantities to the right, broken down to world regions. One can clearly see that Rotterdam handles mainly imports, with more than 281 million tons of cargo being unloaded, while only 88,2 million tons of cargo are being loaded onto ships.

Next I flipped inbound and outbound flows to the same side. However, I think that by this the diagram loses somehow, also because some purple flows (outbound to Africa and Oceania) are too thin.

In the third version, I added a shape for the balance difference between inbound and outgoing goods.

Tell me what you think about theses Sankey diagrams. It would be interesting to compare Rotterdam to other ports. Shanghai, for example, might have the opposite picture with much more exports, but I haven’t found any data yet to show this. And, if we are talking cargo traffic: how about doing a passenger Sankey diagram for one of the international airports in the U.S. (by origin/destination continent?, by airline?)