- The study presents the first spatially explicit dataset of urban settlements from 3700 BC to AD 2000, digitizing and geocoding historical urban population data.
- The dataset was developed using historical, archaeological, and census-based urban population data from Chandler and Modelski, with data cleaning and harmonization procedures applied.
- A reliability ranking system was created to assess the geographic uncertainty of each data point, with 75% of cities ranked as most accurate.
- The dataset aims to improve understanding of contemporary and historical urbanization trends, despite limitations such as temporal and spatial sparseness.
- Chandler's dataset includes cities with populations over 20,000 from AD 800 to AD 1850, and over 40,000 after AD 1850, using diverse sources like censuses, travelers’ diaries, and military records.
- Modelski's work extends Chandler's dataset by focusing on ancient urban settlements (3500 BC to 1000 BC) and using different population thresholds for defining cities.
- The dataset provides insights into the geographic distribution of urban populations over 6,000 years, though it lacks urban extent data and has gaps in certain regions and time periods.
- Future work could enhance the dataset by filling population gaps using updated sources like UN population datasets or archaeological data.