#Featured #Ideas

The Hidden Meanings of Maps: Between the Lines of Legend

35-40_fiordsMost cartographers neglect map legends. They put a lot of efforts in the projection and scale choice, data and pretty colors selection, but then they just want to add a powerful title and press the “share” button. That is  a mistake, as the cartographical legend is the key to understand the map. It is like building a beautiful door and forgetting to put a keyhole on it. It sounds obvious to say, but it is a major communication issue found in very prestigious media maps, still today. When maps were made by hand, making the legend was part of a very artistic process. This article puts side by side recent maps made with computers and ancient ones, drawn by hand, on purpose, to illustrate better the importance of skills and knowledge over technology when it comes to mapping.



One rule matters: every single element that is on the map must be in the legend. Every single symbol, color, shape, line styling that is not explicitly named on the map (so, that necessitates a reading key) has to be mentioned in the legend.

Other rules are secondary and have more to do with readability concerns. It is a lot clearer to group the legend elements, either by type (putting together lines, then points, then areas) or by theme – matching thematic layers of information GIS users know well (water: lines, areas and points; then vegetation: lines, areas, points; then human settlements, etc.). One last, but not least, rule that matters: unit of measurement must be visible (it is surprising to see how many mapper forget that one).


Here is an example (The Guns of Gettysburg) of a legend grouped by types of elements and theme. Elevation is represented as areas, as well as boulders, woods, orchards and swamps, all grouped together in the left part of the legend. Points and lines are mixed together in the middle and right part of the legend, as the author prefered to group the columns thematically. The map is about a historical battle, which justifies the choice of grouping together physical obstacles (4th column from the left), roads and streets (3rd column from the left) and battle strategic places (last column on the right).

On this other example (from a map illustrating the Jamaican slave revolt in the 1760’s), there are points and lines, grouped together that way. Note that the map legend was named “explanation” a few centuries ago, which is probably the best word to describe what it should really be!

The next legend (below) is from The Conventional Symbols for Mine Maps (by Lester and Uren, from the Mining and Scientific Press). It is a fantastic and complex mix of points and lines (and this is only one board, there are two more, with topographic symbols). Two types of elements take several shapes here, as such as shafts, rooms and pillars are points, tunnels and pipes are lines.




Google Chart offers a great tool to make a “gradient legend” for which you can set up as many colors and ranges as you want to illustrate your topic on a map. On top of numerical values, you can add qualitative description words to put the emphasis on the colors and disambiguate things. On this kind of legend, only areas are represented. The legend on the right is a geological grid of rock types with associated colors and patterns. Again, only areas are represented here, as visualizing the distribution of geological set is the main purpose of a geological map.

This legend (below, on the left) illustrates a map of the Wellington Flight Center. The map elements are points only, here sorted by type of services people can find in the airport. The legend on the right is the key to a suicide map of Seattle from 1914 to 1925. Again, all the elements of the map are points, here sorted by the type of place the people committed suicide (at home, other than residence in Seattle, out of Seattle or unknown location). As it is a black and white map, the cartographer had to be inventive and to play with the symbols.

This legend of the Clinton County railroad network map from 1892  has lines only. Each railway is associated with abbreviations of each town name crossed by the train.

The following map represents “a year of shipping routes mapped by GPS” published in 2010 in Wired Science, so seeing only lines make sense. Instead of showing shipping type, tons or passengers count, the color gradation illustrates the number of journeys per route. The legend is then a color gradation sliced into ranges, from the less to the most frequented. It is a little misleading that the less visited routes are in dark colors here, as deep red usually indicates high density, but it is a matter of color choice, not a legend one.


Let’s look a three examples from the media. The following maps have been published recently and illustrate very well the extreme simplification of some mapping practices since computers made mapping tools available for all. As extraordinary as it is to see more maps and more data visualization initiatives in the media and elsewhere, lack of attention to details can lead to over-simplifications of things.

  • The percentage of American who met the guidelines for aerobic exercise in 2013

That is one interesting map! The metric is still not clear to me after reading the article several times, but it seems to be based on the percentage of American people who reported meeting the guidelines for aerobic exercise for the past year. So we must assume the numbers on the legend are percentages. Basic rule for color gradation fitting percentages: always start at 0 and always end at 100%. The view here, excluding low and high values (probably because there are empty), is very misleading about the real distribution.

  •  The “revealing map of the world’s most and least ethnically diverse countries”

This very controversial map from the Washington Post illustrates the “ethnic diversity” of countries in the world. As if the topic was not touchy enough, the map below has a terrible legend, with no numbers whatsoever, very vague descriptive words (and a typo). Usually people put “more” and “less” at the two extremes, keeping the same adjective. Here we have to agree with “diverse” being the exact contrary of “homogenous”. Reading the article (several times as well), it seems that this map is based on “the odd that, on two random people called in a country, they would both say they have a different ethnical belonging than the other called person”… I am pretty sure you’ll get it better than I did.

diverity-map-harvard2Source: The Washington Post

  • The absolute Global Map: Facebook

This beautiful map is all over the place on the internet, as, of course, it has been massively shared on Facebook. The social network is so famous, so global, so universal that it does not need any key or explanation: everybody knows that this map showing connexions between people on Facebook. At least I hope so, because you won’t get any legend to help.


Source: Facebook, by Paul Butler

On this other map made from the first from fastcompany, the map maker had the kindness of writing the legend on the bottom right corner. The intention was probably to illustrate how Facebook leaves behind most of the dense areas in the world – so, most of the people.


Say thanks for this article (0)
The community is supported by:
Become a sponsor
#Business #Featured #Ideas #Science
Harnessing the Power of 30cm Satellite Data for Construction Mega Projects
Aleks Buczkowski 07.30.2023
#Featured #Fun #Ideas
The Geo Geek Gift Guide: 10 Christmas ideas for the geo geek in your life
Nikita Marwaha Kraetzig 12.14.2023
#Featured #GeoDev #People
MapAction looking for volunteers to unlock information management barriers in humanitarian sector
Alex Macbeth 10.31.2023
Next article

The Ocean GIS Initiative: Taking GIS underwater

source: Westfield State University

source: Westfield State University

As geographers and Geo-information technologists, we are well aware of the potential and applications of GIS. It is amazing how a technology that took shape in the 1960’s based on a simple concept of overlaying maps is now at the forefront of what we are defining as the “Geospatial Revolution“. I wonder if even Dr. Roger Tomlinson expected his idea and the project “Canada Geographic Information System” would one day lead to the explosion of GIS technology and its application in various domains ranging from agriculture to crime control and so on.  Dr. Tomlinson, the “father of GIS” deserves all the credit for his idea but personally I believe it was Dr. Dana Tomlin’s work “the Map algebra” that made GIS such an effective technology that it is today. But in spite of all the development over the last few decades, GIS technology was largely limited to being used on land.

A few years back, I was working on a GIS project that involved GIS research in the ocean environment. This was at a time when almost no GIS software, open-source or otherwise had any special toolbox for applications that involved the Oceans! The first question I had was:

How different is using the GIS software for land and ocean based projects?

bathyBased on my experience, I wouldn’t say it is totally different but surely was very different from using GIS for land-based applications and projects. Let me elaborate, consider a GIS project where height information plays a crucial role like in the case of Landslide Hazard mapping. The steeper the slope, the greater the hazard (disregarding all other factors for the sake of simplicity). Now calculating the slope on land where all height information is positive seems a relatively easy task, right? Performing the same operation for an Ocean GIS project would end up resulting in a million errors and roadblocks simply because all the height information would be negative numbers and inverting the numbers doesn’t help in the absolute sense.

It is simply not possible to assign the same dimension for the same problem on Land and Ocean. My team that was working on this Ocean GIS research a few years ago was facing issues like this and many others. We were not wanting to build an entirely new GIS software and so we had to circumvent the problem using smaller extensions. I’m sure many GIS applications for Oceans did not take off simply because there were no turn-key solutions available. Lack of commercial applications and interest was one of the main reasons why GIS softwares did not have any solutions for Ocean GIS.

But not anymore! With the boom in shipping industry, underwater internet cables, besides government regulations regarding oil spills and the potential of minerals underwater (which was the reasons why I started the research in the first place) have all made Ocean GIS an interesting requirement.

ArcGIS toolbox for Maritime Applications is probably the only complete solution that exists right now that would have made my research few years back much more simplier. It is interesting to note the number of projects that ESRI has taken up in its “Ocean GIS Initiative” that will not only aid in Ocean GIS application but also assist in understanding the complex world of water that surrounds us. Over 90% of cargo is shipped via Ships and at least a hundred million people world over depend on the oceans for their living. Oil spills and other disasters in the past could not be modelled using GIS but now we have the means to model them.

GIS is now scaling greater depths than ever before. What do you think?

Read on