This is carried over from something I mistakenly stuck on the to-do list.
I was looking at the way the database stores location information, and it seems to me that the way geographic information is represented could use some re-tooling. It’s currently just arbitrary human-readable zones. I think it might be more usefully represented in a more machine-readable format, (e.g. latitude range, bounding box, etc. - clearly this is tricky because not all the zones are strictly north-south, among other details). The human-readable part can always be generated on the human interface side as needed. I mention this because if data needs to be exported for incorporation into models, visualizations, etc., the more machine-readable it is, the more it will seamlessly integrate into those other things. Does this make sense or am I way off base?