A bit of massage to NASA SEDAC dataset...

@rhermans As usual... code is a not exactly publication-grade and a bit interactive (if you mean finding the actual circle, not drawing it). Basically I took NASA SEDAC data (equirectangular population density data), compensated for latitude skew, and searched essentially by brute force for suitable GeoDisk, just using ImageMultiply between raster images to find covered populations...

Maybe some day I clean code to a level it'd be presentable... :)

Just figured out that technically I could have probably used ImageConvolve to simplify the process...

But yeah, I should take a look at it. It should not be too hard, just that striving too much on clean code in work in progress is a bit pointless... :)