"2007 is the first year in human history when most people on Earth live in cities"
United Nations' Global Environment Outlook 4 (2007)
Without any doubt human population growth can be regarded as a main driving force of global change during the most recent part of the Holocene. Therefore, it is crucial to get a good insight of the demographic developments which occurred that period. Although not many comprehensive global population studies exist which cover the Holocene, a few important sources are used in this study. Historical population numbers of McEvedy & Jones (1978), Livi-Bacci (2007), and Maddison (2003), Denevan (1992) form the basis of our national historical population estimates. Supplemented with the sub-national population numbers of Populstat (Lahmeyer, 2004, pers. comm.; who provides data for several time periods varying per country), time series were constructed for each province or state of every country of the world. A summary of the population numbers used for HYDE 3.1 is presented in Table 1, and an elaborate summary table can be found in the supplementary material. For simplicity reasons, current administrative units were kept constant over time, and every historical source was adjusted to match the current sub-national boundaries of HYDE 3 (e.g. by taking fractions of former larger empires). Country and regional totals were checked against other historical estimates (see supplementary material). Table 2 presents regional estimates for total population and population density, and Table 3 presents the resulting population growth rates per year over time for the different world regions.Urban fraction
Mankind started as a rural society. Just after the domestication of plants and animals people started to have sedentary agriculture and began building farms, houses and small towns/villages.There are quite few historical sources about urbanization numbers and fractions. Urban/rural fractions for all countries were derived from the U.N. after 1950 A.D. (UN, 2008). Earlier historical urbanization estimates for Europe were derived from (De Vries, 1984; Bairoch, 1988, Chandler (1987), Canada after 1890 A.D. from Urquhart and Buckley (1965), China from Rozman (1973) and Maddison (1995), Colombia from Etter et al (2006), USA from Dodd (1993), all other countries were estimated similar to De Vries (1984), which data yielded roughly a factor 10 lower in 1700 A.D. than the 1950 value of the U.N. Before 1700 A.D. the urban fraction approaches zero in 0 A.D. in most countries, with a few exceptions in the older early civilized regions, such as the Mediterranean, the Levant, Central/South America and parts of India and China. Global urbanization was estimated 1% in 0 A.D., 2.6% in 1000 A.D., 3.6% in 1500 A.D., 19% in 1900 A.D. and 29% in 1950 A.D. It reached almost 50% in 2000 A.D. and at present day more people on Earth live in cities than in rural areas. Although regional differences do exist, the overall urbanization levels remained very low for a long time. Europe became relatively more urbanized during the Middle Ages, with China and India as runner up. When the Industrial Revolution was in full swing, not only Europe but also other world regions began to urbanize. Especially after 1900 A.D. the urbanization fractions increased rapidly. Estimates of urbanization fractions for different world regions over time are presented in the supplementary material.Built-up area in 2000 A.D.
Infrastructure and urban development is increasing rapidly during the last decades (UN, 2008), but illustrative is the fact that even for the present day there is no clear picture of the built-up area in the world. Potere and Schneider (2007) compared six spatially explicit studies which reported global estimates for built-up area. Globally, estimates for the extent of built-up areas in 2000 range from 0.2% - 2.7% of the total land area, with 5 of the 7 estimates below the 0.5%. Most of the differences can be explained by various definitions of built-up area, and differences between satellite derived and inventory based data. All these percentages correspond to ca. 0.3 - 3.5 million km2 worldwide of land which is at first sight not available for producing food. For comparison, this study estimates a global built-up area percentage of ca.0.3% (0.5 million km2) in 2000 A.D.
The maps we use for current built-up areas were created by first combining ‘Urban and Built-up’ from the DISCover map (Loveland et al., 2000) with the areas of ‘Artificial surfaces and associated areas’ from GLC2000 (Bartholome et al., 2002). Subsequently, an overlay of this built-up map was made with the Landscan population counts map (Landscan, 2006) to calculate the average population density within built-up areas for each country for the year 2000. The LandScanTM Dataset comprises a worldwide population database compiled on a 30" X 30" latitude/longitude grid. Census counts (at sub-national level) were apportioned to each grid cell based on likelihood coefficients, which are based on proximity to roads, slope, land cover, nighttime lights, and other information. For some countries we took the urban densities from Demographia (2006) as a maximum, to avoid unrealistic (i.e. too high) values for national urban population densities (e.g. Macau, Hong Kong, due to a mismatch in our GIS system of area and total numbers). Since there are hardly any reliable estimates around for historical built-up area, we decided to use historical urban densities as a proxy instead.Urban density
Also here, few sources of historical urban densities can be found (De Vries, 1984, Baroich (1976), Demographia, 2008). Figure 1 depicts the urban density over time of some European and North American cities. The figure shows that for cities the urban density was not constant over time, but always seems to increase first rapidly to very high densities indeed (up to 40,000 inh/km2 !) and then, when the standard of living improves it decreases more slowly (compared to its original increase) over time. Income, cultural and human behavior aspects and planning policy play a role here. We therefore assume that the (historical) urban densities follow an asymmetric bell-shaped curve.
Since the shape of the curve is derived from very few available city data, a major assumption is that we believe that they are representative for a whole country and indeed for all countries in the world. Of course this introduces uncertainty, but in the light of this database (and its intended rather ‘crude’use in integrated models of global change) we think this is acceptable. Only the size and the shape of the curve differ between the countries, depending on their development stage in time, see Figure SF2.
We developed a method to determine the urban population density over time for each country. The asymptotes of the curve (L and K) are defined as a fraction of the maximum of the curve (Dmax). Also the parameter c which indicates the ratio between the decrease rate and the increase rate of the curve is estimated globally. To determine the final shape of the curve for each country, we need two points on the curve. The first point is derived from an overlay of the Landscan population density maps and satellite imagery for the year 2000. The second point needed is the time and magnitude where the curve reaches is maximum (Dmax and tmax). We assume that this maximum is reached when the increase in urban population is slowing down for the first time in history. The full methodology is described in the supplementary material.
We can now compute the built-up area for each country by dividing the total urban population numbers with the time dependent urban densities for each country. Figure 3 summarizes the methodology to calculate the population numbers, densities and urban areas on a spatially explicit way. These maps are used as input for the allocation of historical land use constructions for global change research (e.g. Klein Goldewijk et al. 2009).
The 2008 Revision of the United Nations World Population Prospects (UN, 2009) presents rural/urban population data on a country basis for the period 1950 - 2000. There are few other sources which presents data on historical urbanisation levels. Globally the precentage of humans living in cities increased from 3% in 1700 towards 47.4% in 2000.
You can download the summary tables and figures here.