MacEachren and Kraak and Simple Visualizations


            MacEachren and Kraak explain the importance of geovisualization as a way to merge human vision with domain expertise. Broad applicability in fields such as medical imaging awaits pending the solving of major issues in representation, integration, interface, and cognitive/usability issues. The authors round up their paper by pushing for practical solutions to increase research done on geovisualization.

            I would like to point out that improvements in geovisualization need not necessitate more realistic models. I undertook extensive fieldwork and research to present:

This is a screenshot from the simulation game Dwarf Fortress, whose graphics are entirely based on ASCII. The green triangles represent slope (upwards-pointing triangles represent an uphill slope, downward ones indicate a valley), while different elevation levels are conceived in stacked layer format which can be viewed at the press of a key. Depending on my purposes, this simple representation may be enough to inform my decision of uneven terrain ideal for defending my dwarves (don’t need exact elevation values). The graphics are certainly sufficient for representing how individuals interact and gather resources from the environment (e.g. shortest distance calculations by finding the nearest firewood). A bit contrived I know, but the argument holds for situations such as the Battle of the Boids Agent Based Model shown in class where ‘boids’ were simple triangles, yet were able to show movement patterns. I was also challenged in my raster GIS class where given a DEM of say, Mont Royal, what value would animating it in more realistic 3D have from a purely analytical perspective. I’d like to open this question to other readers (I only came up with being able to debug poor stitch jobs and mismatched elevations with other DEMs at the seams). I concede however that when exploring massive datasets with an abductive approach (no hypothesis in mind), realistic visualizations may offer more creative stimulation to the user.     

MacEachren and Kraak briefly touch on this point by noting a tension between realistic and abstract representations, saying some believe “abstraction is essential for achieving insight”. I feel that the reasons for abstract models tend to be more for practical reasons of limited time and resources than a belief that abstract models are more objective and thus insightful.



Comments are closed.