With the popularity of data visualisations, cartographers are increasingly looking for aesthetically-appealing ways to visualise and maybe even ‘sex-up’ their geographic, and often thematic, data. Here is a brief low-down on some ‘different’ ways to show your data.
A more traditional cartographic method often deployed on World Maps. Equal-area projections are where the area on the map is directly proportional to area on the ground. Great for thematic maps as it allows different regions to be compared fairly. Generally a poor indicator of shape and location as positions on the map are warped but is often favoured over other projections when mapping the Polar Regions.
Area cartograms are a popular alternative to traditional choropleth maps due to their ability to grab the attention of an audience and to powerfully promote a message. The downside is that the underlying map is often dramatically distorted losing both geographic reference and any subtleties within the data.
A thematic mapping technique used to represent density for continuous data, usually using a diverging and often spectral colour scheme. Heat maps are often created from point data and so the values creating a continuous fill of colour are a process of algorithmic extrapolation and therefore one cannot assume the value at any given location to be reliable. Heat maps are however incredibly useful in identifying patterns and in particular, as the name might suggest, ‘hot spots’; because of this they are increasingly being used in sports analysis for example.
Like heat maps, cluster maps are generally created from point data. The purpose of a cluster map is to reduce the amount of clutter by representing dense pockets of often overlapping point data with a single point usually sized in relation to and/or labelled with the number of points that have been grouped together. Often used on interactive web maps, the clustering is likely to change with scale. The weakness is loss of positional accuracy and that it is difficult to show more than one class graphically.
A currently fashionable technique for visualising density when working with large point data sets, it uses hexagon shapes to create a grid and develop a spatial histogram. It then works like any other choropleth map: In any given hexagon, only the most frequently occuring value or class is shown and the colour scheme (sequential, diverging or qualitative) depends on the theme or data being shown.
3D geostatistical maps
These are where data values are shown as protruding from a surface. Not all of these techniques are as yet widely-adopted but all of the below have been popular with in university-based research for several years now. Examples include:
Spike or conal map
This works quite well for datasets that extreme values and enough geographical dispersion and if you wish to create a surface rather than points – think of it as a 3-dimensional heat map.
However in many cases the results look like a medical electroencephalogram, where the third dimension obscures much of the data which also makes it harder for the reader to reference or understand geographic location.
3D proportional symbol map
3D bubbles are a popular research alternative to spikes as they can be less obtrusive but that depends heavily on your dataset and styling. They are also often reduced further to domes. There were far better and more scientific examples presented at the International Cartographic Conference in Dresden last year but unfortunately I cannot seem to source any of these.
Proportional symbols can give a visual appeal but one must consider how to scale them, for example by one dimension or by volume. They don’t work so well in an immersive 3-dimensional environment; whilst they may look cool in Google Earth, they make it very hard for the viewer to understand and distinguish relative size between points due to both their shape and the spherical perspective of the globe.
Textures, patterns and images
The textures and patterns I refer to are not mere graphic effects aimed at ‘popping’ data. They are an alternative (or additional) method to colour in displaying polygonal data. I often use textures at work to add an extra dimension to woods or water but it has become increasingly popular to apply a texture to the background, for example all of the rough paper textures used on the ever popular pirate maps. Textures can help strengthen the relationship between a map and real world objects and surfaces so long as they don’t interfere too much with the map’s message.
Images are often used as a pattern fill too. Again not just as an effect but actually as an alternative to colour and or labelling in representing thematic data. These are simple but big hits on social media, for example mapping the foods of each region in Italy by filling each region polygon with an image of the dominant food associated to that particular area. In the example below-right I have shown the best selling newspaper for every country in Europe. This approach is a powerful visual for qualitative mapping where each value or polygon is unique.
Although not a data representation technique, the effect it has on the way in which data is represented is so profound that I sought to include it on the list. Data visualisation led to an explosion in mapping on a black or dark grey background in recent years. This trend was previously discussed by Charley Glynn on Cartoblography. Dark backgrounds are great for on-screen because in this environment it is a more natural background than white and allows more emphasis to be drawn from the colours of the map, for example the ‘bright light’ and trail effects of the time lapse mapping on themes such as world flights. The main downside of course is printing where far more skill is required to achieve appropriate colour balance.
With masses of data being geographically refenced, shared and visualised, I used the term ‘alternative’ at the start because there are now many ways in which one can show the same data and tell the same story. In terms of cartography, many of these techniques are actually quite traditional but with a modern flavour. And with the tools now at our disposal we can do these things more efficiently and often in a far more striking manner. Many alternative approaches to presenting geographic data lead into a number of cartographic pitfalls, so we should be aware of the cartographic rules even if we choose to flout them, however the world of data visualisation has shown that it sometimes pays to be slightly daring in one’s approach.
Some more techniques can be found on this well-presented webpage at Ubimix.