Exploring Germany through Cartograms

My inten­tions for this blog are to spur intel­li­gent and thought pro­vok­ing dis­cus­sions and to pro­vide some insight and/or help to those who may want to delve into var­i­ous spa­tial analy­ses and car­to­graphic trends/methodologies. Car­tograms are def­i­nitely one of the most trend­ing top­ics in dis­cus­sions on how to alter­na­tively geo­vi­su­al­ize spa­tial phe­nom­ena. The idea roots back more than half a decade and has evolved from a ‘pen­cil and graph paper’ skill to a geo­graph­i­cal infor­ma­tion sys­tem (GIS) tool{box}. We can add… car­tograms along­side a plethora of geo­vi­su­al­iza­tion tech­niques, which include chloro­pleth map­ping, dot den­sity, and pie/bar charts to name a few.

Population Density Cartogram

I have been test­ing out a script that was posted on the Arc Resources page (here: http://arcscripts.esri.com/details.asp?dbid=15638), which lets you make your own fancy car­tograms. I was pulled towards map­ping Ger­many for a cou­ple of reasons.

1. Ger­many has an odd struc­tural lay­out of their “states”. Berlin, a free state (and capi­tol city) is located within another state, Bran­den­berg. This is also the case for Bre­men. Of course, we also have Ham­burg which is also some­what oddly placed, being tucked in between Nieder­sach­sen and Schleswig-Holstein. This would cer­tainly make for a great car­togram (I thought) because the sur­round­ing states are sparsely pop­u­lated com­pared to the 3 largest states (Berlin, Bre­men and Ham­burg) located within them.
2. Ger­many is aes­thet­i­cally pleas­ing to map and the data was made avail­able (for free) by GADM (here: http://gadm.org/)
3. Wikipedia (and by that I mean its con­tribut­ing community) offers pop­u­la­tion den­sity and raw counts for almost all denom­i­na­tions of car­to­graphic bound­aries for most places around the world (includ­ing Germany!).

The Map!

Web ver­sion here:

^ Click me to zoom! ^

Full size available here: Click me!

Some thoughts

The car­togram dis­torts (manip­u­lates) the land area in respect to the rel­e­vant data that you would like to explore. In this case, the land area is dis­torted based on pop­u­la­tion den­sity of each state. Berlin, Bre­men and Ham­burg become enor­mous, and the dark blue states shrink down in size while the yel­low and orange hue states change slightly in size. The car­togram pre­serves the rela­tion­ship between objects on the map while skew­ing the shape such that we can still rec­og­nize (although barely) that this coun­try is in fact Germany.

Although I am work­ing with the “Gastner-Newman” method for cre­at­ing a car­togram, the tool itself and inher­ently the method­ol­ogy for cre­at­ing car­tograms, is con­trolled by the user. The Gastner-Newman method is not the only method­ol­ogy, nor is it the best. I implore you to inves­ti­gate a wide vari­ety of con­tigu­ous, non-contigous and other abstract forms of geo­vi­su­al­iza­tion and map manip­u­la­tion. The tool I am using, allows the user to con­fig­ure many options, which cre­ates vary­ing results. Ide­ally, car­tograms should be visu­ally strik­ing but infor­ma­tive. The car­togram should pro­vide addi­tional infor­ma­tion to the user for use in inter­pre­ta­tion. I believe the car­togram is a great tool for cre­at­ing visu­ally impact­ing dis­plays. The chloro­pleth map of pop­u­la­tion den­sity in Ger­many (left) fails to rep­re­sent the ulti­mate goal of dis­cern­ing the vary­ing pop­u­la­tion den­si­ties in Ger­many. The car­togram (right) how­ever, por­trays a very strong visual stim­uli for the theme of pop­u­la­tion density.

Also note: I was really pleased to see that Berlin, Bre­men and Ham­burg showed up so promi­nently and in a really weird bug-eye/alien form. It really con­firmed what I thought it would look like from the begin­ning (which is what we always hope for in any type of analy­sis, right?).

Limitations?

I should prob­a­bly note that pop­u­la­tion den­sity was per­haps not the best method for cre­at­ing a car­togram. I may have to revisit the car­togram with raw pop­u­la­tion data. The rea­sons why we shouldn’t map rates using car­tograms are best cov­ered in this dis­cus­sion (here: http://www.worldmapper.org/mapping_rates.html) by the indi­vid­u­als who have really set a bench­mark for cre­at­ing car­tograms (here: http://www.worldmapper.org/ … 695 maps… wow!)

Cheers!