Spatializing 6k years of global urbanization from 3700 BC to AD 2000 (opens in new tab)

This is one of those datasets that no one besides an academic with a very narrow research question is likely to find useful. It's not reflective of what we understand about the extent of historical urbanism today, it's just a synthesis from two earlier, systematically flawed datasets into a machine readable form. It misses a lot, like the entirety of ancient urbanism in North Mexico/the American southwest, Numidia, Axum, large urban centers in central Asia, etc. The authors were aware of many of these shortcomings when they published this, but didn't want to add additional shortcomings from omissions beyond what the original datasets had.