From David O’Sullivan1, self-described geospatial “problem solver, researcher, analyst, communicator, and trainer”:
In this post I look at a relatively simple example of [the problem of presenting complex multivariate data in single map views] (…) I’m deliberately not going to say what the data represent. Suffice to say we have three variables, shown below for a chunk of New Zealand. From left to right, these are a resource potentially at risk, and two risk factors.
David O’Sullivan walks us through various considerations and design alternatives to show the afore-mentioned data in a single map: tile maps using tilings of squares and other shapes, glyph maps, tile maps with diagrams, and a tile map combined with the value-by-alpha method popularized by Rob Roth2, for example.
In this article, I also learned about MapWeaver, a software that allows you to conduct your own experiments with various tiling schemes, and the underlying Python package weavingspace.
I love the out-of-the-box experimentation of this article and the more avant-garde approaches that challenge traditional visualization methods offered by your run-of-the-mill GIS software. More of this, please.
Footnotes
“O’Sullivan” in GIS… – Wait, what is this bell going off in my head? I was very pleased to realize that I was looking at the website of the O’Sullivan of “O’Sullivan and Unwin (2003): Geographic Information Analysis” (or 2010, 2nd ed.). This is my all-time favorite book on GIS and spatial analyis.↩︎