As a test run, I'd like to visualize population on census blocks for a few states (New England for example). I assume that the kml to do this will be huge, when you factor in the census boundaries and population data.

How can I make my data efficient to be able to do such visualizations keeping in mind that I might extend this to having multiple census periods, and doing a time lapse combined with a flythrough?

Also note that usage of Google Earth isn't necessary. I'm open to using QGIS


I did this by converting shapefiles into rasters with a large gridcell size, and then converting these rasters back to polygons, and then to KML format. This is an easy way of simplifying the data, but I think it also looks better in a visualization.

(I'm happy to share the code, though it is for ArcGIS 10).

  • We're still using 9.3.1 – dassouki May 27 '11 at 19:02

You could generalize the boundaries and remove excess points. Fewer points = less data = more efficient. It would be less accurate, but then I get the feeling the visualization is more important than exact positional accuracy.

Or you could replace the census area polygons with a circle (max size that fits inside the boundary?). Again, if visualization is the key then it needn't be exact map boundaries that you are showing.

I like the idea of a network link, but I wonder how well it would work when doing a flythrough. Could it fetch data fast enough???

  • I'd like to get away from the idea of a network link, we're not setup to doing so, and I don't think there is interest to using geoserver at the moment – dassouki May 27 '11 at 19:03

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.