One of the hardest things to visualize is information that depends on time *and* space. These are known as “spatio-temporal” visualizations or “space-time cubes”, [several techniques](http://aviz.fr/~bbach/spacetimecubes/Bach2014spacetimecubes.pdf) have [been proposed](http://spacetimecubevis.com/) to deal with the complexity of seeing both information at the same time.

Here I propose a new\* and very simple visualization that shows *where* and *when*, I call it “Tree-Ring Graph”.

![Tree-Ring Graph of GDP over the years in Brazilian States (darker means increasing bigger, each color means 1/3 of GDP)](/media/01M43VVQK02MMVNS97B2NP9C24/dendrochronogram-or-tree-ring-graph-1.png "Tree-Ring Graph of GDP over the years in Brazilian States (darker means increasing bigger, each color means 1/3 of GDP)")

*Tree-Ring Graph of GDP over the years in Brazilian States (darker means increasing bigger, each color means 1/3 of GDP)*

The idea comes from choropleths and tree-ring analysis. Inner rings are “older” than outer ones, color intensity shows quantity. It’s easy to spot trends: if something is getting darker, it’s increasing; if something is getting lighter, it’s decreasing. Cycles are also easy to spot: any “zebra-like” shape is cyclic.

![Tree rings are used to find age of trees](/media/01M43VVRB25BY3GZ4BM9044MC8/dendrochronogram-or-tree-ring-graph-2.png "Tree rings are used to find age of trees")

*Tree rings are used to find age of trees*

I found the main limitation is the number of rings, fewer rings works better, some summarization over data may be needed. Also, it’s not good for precise locations.

Code for d3js can be found [here](http://bl.ocks.org/ruliana/04b0d7b1244ddff7cc9c7d3187358aa5).

*\*“new” as in “I couldn’t find anybody who had done it before”, which could mean I did a poor research. :)*